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).
128 JH/01 Avoid costly startup code when not strictly needed. This reduces time
129 for some exim process initialisations. It does mean that the logging
130 of TLS configuration problems is only done for the daemon startup.
132 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
134 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
135 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
138 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
139 DKIM signing is being done. Previously a transport_filter would always
140 disable CHUNKING, falling back to traditional DATA.
142 JH/05 Regard command-line receipients as tainted.
144 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
146 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
147 PAM library frees one of the arguments given to it, despite the
148 documentation. Therefore a plain malloc must be used.
150 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
151 on-stack buffers were used, resulting in a taint trap when DSN information
152 copied from a received message was written into the buffer.
154 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
155 the ordering of its ARC headers. This caused a crash.
157 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
158 a new record was being constructed with information from the peer, a trap
161 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
162 installation would get error messages from DMARC verify, when it hit the
163 nonexistent file indicated by the default. Distros wanting DMARC enabled
164 should both provide the file and set the option.
165 Also enforce no DMARC verification for command-line sourced messages.
167 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
168 could, depending on the platform, hang at the STARTTLS response.
170 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
171 message on a connection. Previously if one message had ARC headers and
172 the following one did not, a crash could result when adding an
173 Authentication-Results: header.
175 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
176 Exim main code and Exim-related utities. The introduction of taint
177 tracking also did many adjustments to string handling. Since then, eximon
178 frequently terminated with an assert failure.
180 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
181 check for 452 responses. This slightly helps the inefficieny of doing
182 a large alias-expansion into a recipient-limited target. The max_rcpt
183 transport option still applies (and at the current default, will override
184 the new feature). The check is done for either cause of synch, and forces
185 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
186 connection. The new facility is not tunable at this time.
188 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
189 library live data was being used, so the results became garbage. Make
190 copies while it is still usable.
192 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
193 on delivery deferred (==) and failed (**) lines (if a delivery was
194 attemtped). Previously it was only on completion (=>) lines.
196 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
197 for the expansion of the server_scram_iter and server_scram_salt options.
199 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
200 are now specifically given a NO_DATA response without hitting the system
201 resolver. The library goes on to do the now-standard TXT lookup.
202 Use of dnsdb lookups is not affected.
204 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
205 only retrieve the errormessage once. Previously two calls to dlerror()
206 were used, and the second one (for mainlog/paniclog) retrieved null
209 JH/20 Taint checking: disallow use of tainted data for
210 - the appendfile transport file and directory options
211 - the pipe transport command
212 - the autoreply transport file, log and once options
213 - file names used by the redirect router (including filter files)
215 - paths used by single-key lookups
216 Previously this was permitted.
218 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
219 adjusted the size of a major service buffer; this failed because the
220 buffer was in use at the time. Change to a compile-time increase in the
221 buffer size, when this authenticator is compiled into exim.
223 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
224 previous fast-mode was untenable in the face of glibs using mmap to
225 support larger malloc requests.
227 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
228 New values supported, if defined on system where compiled:
229 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
230 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
232 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
233 running a limited number of proceses in parallel, a benefit is gained. The
234 amount varies with the platform hardware and load. The use of the option
235 queue_run_in_order means we cannot do this, as ordering becomes
238 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
239 had introduced a string-copy (for ensuring NUL-termination) which was not
240 appropriate for that case, which can include embedded NUL bytes in the
241 block of data. Investigation showed the copy to actually be needless, the
242 data being length-specified.
244 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
245 done during a receiving connection, and both used TLS, global info was
246 used rather than per-connection info for tracking the state of data
247 queued for transmission. This could result in a connection hang.
249 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
250 Previously, when delivering serveral messages down a single connection
251 only the first would provide a SIZE. This was due to the size information
252 not being properly tracked.
254 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
255 TAI (at 37 seconds currently), pretend to be in UTC for time-related
256 expansion and logging. Previously, spurious values such as a future
257 minute could be seen.
259 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
260 it could crash from a null-deref. This could also affect the
261 ${addresses: } operator and ${readsock } item.
263 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
264 message following a mime one, the variable was not reset.
266 JH/30 When an pipelined-connect fails at the first response, assume incorrect
267 cached capability (perhaps the peer reneged?) and immediately retry in
270 JH/31 Fix spurious detection of timeout while writing to transport filter.
272 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
273 an attempt to copy the string was made before checking it.
275 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
276 taint of the lookup key was maintained; we now regard the presence in the
277 filesystem as sufficient validation.
279 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
280 argument is supplied.
282 JH/35 The dsearch lookup now requires that the directory is an absolute path.
283 Previously this was not checked, and nonempty relative paths made an
284 access under Exim's current working directory.
286 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
287 Previously no event was raised.
289 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
290 parameter supplied by the sender MAIL FROM command. Previously it was
291 ignored, and only the check_spool_space option value for the required
294 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
295 the size of the signing public-key. Previously it was instead giving
296 the size of the signature hash.
298 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
299 the default. See the (new) dkim_verify_min_keysizes option.
301 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
302 and an ACL use a lookup for checking either the local_part or domain,
303 stale data could be accessed. Ensure that variable references are
304 dropped between messages.
306 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
307 by the client was not checked as pointing within response data before
308 being used. A malicious client could thus cause an out-of-bounds read and
309 possibly gain authentication. Fix by adding the check.
311 JH/42 Internationalisation: change the default for downconversion in the smtp
312 transport to be "if needed". Previously it was "as previously set" for
313 the message, which usually meant "if needed" for message-submission but
314 "no" for everything else. However, MTAs have been seen using SMTPUTF8
315 even when the envelope addresses did not need it, resulting in forwarding
316 failures to non-supporting MTAs. A downconvert in such cases will be
317 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
318 transport. The change does mean that addresses needing conversion will
319 be converted when previously a delivery failure would occur.
321 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
322 response was received it would be used unchecked in a fail-DSN, violating
323 standards on line-length limits. Truncate if needed.
325 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
326 doesn't work. (Additionally add a new main config option to configure the
333 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
334 side, to decode a TLS 1.3 packet capture.
336 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
337 Previously the default library behaviour applied, sending two, each in
340 JH/03 Debug output for ACL now gives the config file name and line number for
343 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
345 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
347 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
348 buffer overrun for (non-chunking) other transports.
350 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
351 TLS1.3, means that a server rejecting a client certificate is not visible
352 to the client until the first read of encrypted data (typically the
353 response to EHLO). Add detection for that case and treat it as a failed
354 TLS connection attempt, so that the normal retry-in-clear can work (if
355 suitably configured).
357 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
358 and/or domain. Found and fixed by Jason Betts.
360 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
361 configuration). If a CNAME target was not a wellformed name pattern, a
364 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
365 the OS reports them interleaved with other addresses.
367 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
368 used both for input and for a verify callout, both encrypted, SMTP
369 responses being sent by the server could be lost. This resulted in
370 dropped connections and sometimes bounces generated by a peer sending
373 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
374 string. Previously it was possible to cause undefined behaviour in a
375 library routine (usually a crash). Found by "zerons".
377 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
380 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
381 API was removed, so update to use the newer ones.
383 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
384 any timeout set, is taking a long time. Previously we would hang on to a
385 rotated logfile "forever" if the input was arriving with long gaps
386 (a previous attempt to fix addressed lack, for a long time, of initial
389 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
390 shared (NFS) environment. The length of the tempfile name is now
391 4 + 16 ("hdr.$message_exim_id") which might break on file
392 systems which restrict the file name length to lower values.
395 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
396 shared (NFS) environment.
398 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
399 did for all versions <4.90). Notably -M, -m, --invert, -I may be
402 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
403 on some platforms for bit 31.
405 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
406 to changes apparently associated with TLS1.3 handling some of the APIs
407 previously used were either nonfunctional or inappropriate. Strings
408 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
409 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
410 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
411 This affects log line X= elements, the $tls_{in,out}_cipher variables,
412 and the use of specific cipher names in the encrypted= ACL condition.
414 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
416 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
417 verification result was not updated unless hosts_require_ocsp applied.
419 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
420 queue_list_requires_admin set to false, non-admin users were denied the
423 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
424 directory-of-certs mode. Previously they were advertised despite the
427 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
428 A single TCP connection by a client will now hold a TLS connection open
429 for multiple message deliveries, by default. Previously the default was to
432 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
433 default. If built with the facility, DANE will be used. The facility
434 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
436 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
437 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
438 must be defined and you must still, unless you define DISABLE_TLS, manage
439 the the include-dir and library-file requirements that go with that
440 choice. Non-TLS builds are still supported.
442 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
445 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
446 default. If the platform supports and has the facility enabled, it will
447 be requested on all coneections.
449 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
450 controlled by the build-time option SUPPORT_PIPE_CONNECT.
452 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
454 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
455 success-DSN messages. Previously the From: header was always the default
456 one for these; the option was ignored.
458 JH/28 Fix the timeout on smtp response to apply to the whole response.
459 Previously it was reset for every read, so a teergrubing peer sending
460 single bytes within the time limit could extend the connection for a
461 long time. Credit to Qualsys Security Advisory Team for the discovery.
463 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
464 delivery address, which leaked information of the results of local
465 forwarding. Change to the original envelope recipient address, per
468 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
469 requested. Previously not bounce was generated and a log entry of
470 error ignored was made.
472 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
474 JH/32 Introduce a general tainting mechanism for values read from the input
475 channel, and values derived from them. Refuse to expand any tainted
476 values, to catch one form of exploit.
478 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
479 was unused and the unexpanded text used for the test. Found and
480 fixed by Ruben Jenster.
482 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
483 an attempt to use a TLS library read routine dereffed a nul pointer,
486 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
487 them in our smtp response.
489 JH/36 Have the general router option retry_use_local_part default to true when
490 any of the restrictive preconditions are set (to anything). Previously it
491 was only for check_local user. The change removes one item of manual
492 configuration which is required for proper retries when a remote router
493 handles a subset of addresses for a domain.
495 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
496 link count into consideration.
498 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
499 caused the extension of big_buffer, the following lines were ignored.
501 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
502 accordance with RFC 2308. Previously there was no expiry, so a longlived
503 receive process (eg. due to ACL delays) versus a short SOA value could
506 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
508 JH/39 Promote DMARC support to mainline.
510 JH/40 Bug 2452: Add a References: header to DSNs.
512 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
513 parameters. The relevant library call is documented as "Deprecated: This
514 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
515 3.6.0, DH parameters are negotiated following RFC7919."
517 HS/06 Change the default of dnssec_request_domains to "*"
519 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
520 carried on and emitted a BDAT command, even when PIPELINING was not
523 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
524 buffer was used for the filename, resulting in a trap when tainted
525 arguments (eg. $domain) were used.
527 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
528 recommended to avoid a possible server-load attack. The feature can be
529 re-enabled via the openssl_options main cofiguration option.
531 JH/45 local_scan API: documented the current smtp_printf() call. This changed
532 for version 4.90 - adding a "more data" boolean to the arguments.
533 Bumped the ABI version number also, this having been missed previously;
534 release versions 4.90 to 4.92.3 inclusive were effectively broken in
535 respect of usage of smtp_printf() by either local_scan code or libraries
536 accessed via the ${dlfunc } expansion item. Both will need coding
537 adjustment for any calls to smtp_printf() to match the new function
538 signature; a FALSE value for the new argument is always safe.
540 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
541 the file-offset (which the Linux syscall does, and exim expects); this
542 resulted in an indefinite loop.
544 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
545 to do ARC verification. The Authentication-Results: header line added
546 by the configuration then had no ARC item.
552 JH/01 Remove code calling the customisable local_scan function, unless a new
553 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
555 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
556 non-signal-safe functions being used.
558 JH/03 Bug 2269: When presented with a received message having a stupidly large
559 number of DKIM-Signature headers, disable DKIM verification to avoid
560 a resource-consumption attack. The limit is set at twenty.
562 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
563 report of oldest_pass in ${authres } in consequence, and separate out
564 some descriptions of reasons for verification fail.
566 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
567 files in the spool were present and unlocked. A queue-runner could spot
568 them, resulting in a duplicate delivery. Fix that by doing the unlock
569 after the unlink. Investigation by Tim Stewart. Take the opportunity to
570 add more error-checking on spoolfile handling while that code is being
573 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
574 No known attacks, no CVE, this is defensive hardening.
576 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
577 a queue-runner could start a delivery while other operations were ongoing.
578 Cutthrough delivery was a common victim, resulting in duplicate delivery.
579 Found and investigated by Tim Stewart. Fix by using the open message data
580 file handle rather than opening another, and not locally closing it (which
581 releases a lock) for that case, while creating the temporary .eml format
582 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
584 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
585 $sender_verify_failure/$recipient_verify_failure to "random".
587 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
590 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
591 Previously this would segfault.
593 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
596 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
597 like zero, since the resolver should be doing this for us, But we need one
598 as a CNAME but no MX presence gets the CNAME returned; we need to check
599 that doesn't point to an MX to declare it "no MX returned" rather than
600 "error, loop". A new main option is added so the older capability of
601 following some limited number of chain links is maintained.
603 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
605 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
606 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
607 now-deprecated earlier definitions used only specified the range up to TLS
608 1.2 (in the older-version library docs).
610 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
612 JH/15 Rework TLS client-side context management. Stop using a global, and
613 explicitly pass a context around. This enables future use of TLS for
614 connections to service-daemons (eg. malware scanning) while a client smtp
615 connection is using TLS; with cutthrough connections this is quite likely.
617 JH/16 Fix ARC verification to do AS checks in reverse order.
619 JH/17 Support a "tls" option on the ${readsocket } expansion item.
621 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
622 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
623 Previously the "utf8" would be re-prepended for every additional message.
625 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
626 Previously thery were accepted, resulting in issues when attempting to
627 forward messages to a non-supporting MTA.
629 PP/02 Let -n work with printing macros too, not just options.
631 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
632 one parent address was copied, and bogus data was used at delivery-logging
633 time. Either a crash (after delivery) or bogus log data could result.
634 Discovery and analysis by Tim Stewart.
636 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
637 Previously if the string ended mid-character, we did not insert the
638 promised '?' replacement.
640 PP/04 Documentation: current string operators work on bytes, not codepoints.
642 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
643 should pack well giving a smaller memory footprint so better caching and
644 therefore performance. Group the declarations where this can't be done so
645 that the byte-sized flag variables are not interspersed among pointer
646 variables, giving a better chance of good packing by the compiler.
648 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
649 non-null, to avoid issues with sites running BATV. Previously reports were
650 sent with an empty envelope sender so looked like bounces.
652 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
653 The ignore_error flag wasn't being returned from the filter subprocess so
654 was not set for later routers. Investigation and fix by Matthias Kurz.
656 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
657 and a msg:complete for the whole, when a message is manually removed using
658 -Mrm. Developement by Matthias Kurz, hacked on by JH.
660 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
661 a "Gnu special" function, asprintf() in the DB utility binary builds; I
662 hope that is portable enough.
664 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
665 requiring a known-CA anchor certificate; make it now rely entirely on the
666 TLSA as an anchor. Checking the name on the leaf cert against the name
667 on the A-record for the host is still done for TA (but not for EE mode).
669 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
670 included in delivery lines for non-proxied connections, when compiled with
671 SUPPORT_SOCKS and running with proxy logging enabled.
673 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
674 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
675 move the existing event to fire before the normal logging of message
676 failure so that custom logging is bracketed by normal logging.
678 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
679 msg:fail:internal event. Developement by Matthias Kurz.
681 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
682 far too small for todays use of crypto signatures stored there. Go all
683 the way to the max DNS message size of 64kB, even though this might be
684 overmuch for IOT constrained device use.
686 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
687 copy a string over itself. The library routine is documented as not
688 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
690 JH/32 For main options check_spool_space and check_inode_space, where the
691 platform supports 64b integers, support more than the previous 2^31 kB
692 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
693 the previous G, M, k.
695 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
696 $authenticated_fail_id variable on authentication failure. Previously
699 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
700 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
701 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
702 GNUTLS_SEC_PARAM_MEDIUM.
704 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
705 side. Previously we would continue as if no SNI had been received.
707 JH/36 Harden the handling of string-lists. When a list consisted of a sole
708 "<" character, which should be a list-separator specification, we walked
709 off past the nul-terimation.
711 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
712 causes) even when the retry time is not yet met. Previously they were
713 not, meaning that when (say) an account was over-quota and temp-rejecting,
714 and multiple senders' messages were queued, only one sender would get
715 notified on each configured delay_warning cycle.
717 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
719 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
720 cipher-suites, an error can be left on the stack even for a succeeding
721 accept; this results in impossible error messages when a later operation
724 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
725 return error codes indicating retry. Under TLS1.3 this becomes required.
727 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
728 it only wrote the new authenticators, resulting in a lack of tracking of
729 peer changes of ESMTP extensions until the next cache flush.
731 JH/41 Fix the loop reading a message header line to check for integer overflow,
732 and more-often against header_maxsize. Previously a crafted message could
733 induce a crash of the recive process; now the message is cleanly rejected.
735 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
736 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
742 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
743 When redis_servers is set to a list of > 1 element, and the Redis servers
744 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
745 case of MOVED into a DEFER case instead, thus moving the query onto the
746 next server in the list. For a cluster of N elements, all N servers must
747 be defined in redis_servers.
749 GF/02 Catch and remove uninitialized value warning in exiqsumm
750 Check for existence of @ARGV before looking at $ARGV[0]
752 JH/01 Replace the store_release() internal interface with store_newblock(),
753 which internalises the check required to safely use the old one, plus
754 the allocate and data copy operations duplicated in both (!) of the
755 extant use locations.
757 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
758 modifier. This matches the restriction on the commandline.
760 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
761 Previously only the last row was returned.
763 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
764 we assumed that tags in the header were well-formed, and parsed the
765 element content after inspecting only the first char of the tag.
766 Assumptions at that stage could crash the receive process on malformed
769 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
770 While running the DKIM ACL we operate on the Permanent memory pool so that
771 variables created with "set" persist to the DATA ACL. Also (at any time)
772 DNS lookups that fail create cache records using the Permanent pool. But
773 expansions release any allocations made on the current pool - so a dnsdb
774 lookup expansion done in the DKIM ACL releases the memory used for the
775 DNS negative-cache, and bad things result. Solution is to switch to the
776 Main pool for expansions.
777 While we're in that code, add checks on the DNS cache during store_reset,
778 active in the testsuite.
779 Problem spotted, and debugging aided, by Wolfgang Breyha.
781 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
782 When none of the hosts presented to a transport match an already-open
783 connection, close it and proceed with the list. Previously we would
784 queue the message. Spotted by Lena with Yahoo, probably involving
787 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
788 Previously a spurious "250 OK id=" response was appended to the proper
791 JH/08 The "support for" informational output now, which built with Content
792 Scanning support, has a line for the malware scanner interfaces compiled
793 in. Interface can be individually included or not at build time.
795 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
796 by the template makefile "src/EDITME". The "STREAM" support for an older
797 ClamAV interface method is removed.
799 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
800 rows affected is given instead).
802 JH/11 The runtime Berkeley DB library version is now additionally output by
803 "exim -d -bV". Previously only the compile-time version was shown.
805 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
806 SMTP connection. Previously, when one had more receipients than the
807 first, an abortive onward connection was made. Move to full support for
808 multiple onward connections in sequence, handling cutthrough connection
809 for all multi-message initiating connections.
811 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
812 routers. Previously, a multi-recipient message would fail to match the
813 onward-connection opened for the first recipient, and cause its closure.
815 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
816 a timeout on read on a GnuTLS initiating connection, resulting in the
817 initiating connection being dropped. This mattered most when the callout
818 was marked defer_ok. Fix to keep the two timeout-detection methods
821 JH/15 Relax results from ACL control request to enable cutthrough, in
822 unsupported situations, from error to silently (except under debug)
823 ignoring. This covers use with PRDR, frozen messages, queue-only and
826 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
828 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
829 metadata, resulting in a crash in free().
831 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
832 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
833 Broken also in d185889f4, with init system revamp.
835 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
836 Previously we abruptly closed the connection after reading a malware-
837 found indication; now we go on to read the "scan ok" response line,
840 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
841 ACL. Previously, a crash would result.
843 JH/19 Speed up macro lookups during configuration file read, by skipping non-
844 macro text after a replacement (previously it was only once per line) and
845 by skipping builtin macros when searching for an uppercase lead character.
847 JH/20 DANE support moved from Experimental to mainline. The Makefile control
848 for the build is renamed.
850 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
851 was allocated for every new TLS startup, meaning one per message. Fix
852 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
854 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
855 reported the original. Fix to report (as far as possible) the ACL
856 result replacing the original.
858 JH/23 Fix memory leak during multi-message connections using STARTTLS under
859 OpenSSL. Certificate information is loaded for every new TLS startup,
860 and the resources needed to be freed.
862 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
864 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
867 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
868 DATA response info to the (existing) per-recipient response info for
869 the "C=" log element. It can have useful tracking info from the
870 destination system. Patch from Simon Arlott.
872 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
873 length value. Previously this would segfault.
875 HS/02 Support Avast multiline protoocol, this allows passing flags to
876 newer versions of the scanner.
878 JH/28 Ensure that variables possibly set during message acceptance are marked
879 dead before release of memory in the daemon loop. This stops complaints
880 about them when the debug_store option is enabled. Discovered specifically
881 for sender_rate_period, but applies to a whole set of variables.
882 Do the same for the queue-runner and queue-list loops, for variables set
883 from spool message files. Do the same for the SMTP per-message loop, for
884 certain variables indirectly set in ACL operations.
886 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
887 as a multi-recipient message from a mailinglist manager). The coding had
888 an arbitrary cutoff number of characters while checking for more input;
889 enforced by writing a NUL into the buffer. This corrupted long / fast
890 input. The problem was exposed more widely when more pipelineing of SMTP
891 responses was introduced, and one Exim system was feeding another.
892 The symptom is log complaints of SMTP syntax error (NUL chars) on the
893 receiving system, and refused recipients seen by the sending system
894 (propating to people being dropped from mailing lists).
895 Discovered and pinpointed by David Carter.
897 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
898 replaced by the ${authresults } expansion.
900 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
902 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
903 allows proper process termination in container environments.
905 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
906 Previously the "final dot" had a newline after it; ensure it is CR,LF.
908 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
909 and "err_perm", deprecated since 4.83 when the RFC-defined words
910 "temperror" and "permerror" were introduced.
912 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
913 transport-filters or DKIM-signing. The restriction was lost in the
914 consolidation of verify-callout and delivery SMTP handling.
915 Extend the restriction to also cover ARC-signing.
917 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
918 in defer=pass mode supply a 450 to the initiator. Previously the message
921 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
922 tls_require_ciphers is used as before.
924 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
925 "pass_unscanned". Only tmpfails from the scanner are written to
926 the paniclog, as they may require admin intervention (permission
927 denied, license issues). Other scanner errors (like decompression
928 bombs) do not cause a paniclog entry.
930 JH/36 Fix reinitialisation of DKIM logging variable between messages.
931 Previously it was possible to log spurious information in receive log
934 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
935 triggered odd behaviour from Outlook Express clients.
937 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
940 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
941 since the IETF WG has not yet settled on that versus the original
942 "bare" representation.
944 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
945 Previously the millisecond value corrupted the output.
946 Fix also for syslog_pid=no and log_selector +pid, for which the pid
947 corrupted the output.
953 JH/01 Rework error string handling in TLS interface so that the caller in
954 more cases is responsible for logging. This permits library-sourced
955 string to be attached to addresses during delivery, and collapses
956 pairs of long lines into single ones.
958 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
959 during configuration. Wildcards are allowed and expanded.
961 JH/02 Rework error string handling in DKIM to pass more info back to callers.
962 This permits better logging.
964 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
965 do not close it down and have the child transport start it up again on
966 the passed-on TCP connection. Instead, proxy the child (and any
967 subsequent ones) for TLS via a unix-domain socket channel. Logging is
968 affected: the continued delivery log lines do not have any DNSSEC, TLS
969 Certificate or OCSP information. TLS cipher information is still logged.
971 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
972 identical IP addresses on different listening ports. Will also affect
975 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
976 add noisy ifdef guards to special-case this sillyness.
977 Patch from Bernd Kuhls.
979 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
980 than 255 are no longer allowed.
982 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
983 Disable the session-cache too, which might reduce our load. Since we
984 currrectly use a new context for every connection, both as server and
985 client, there is no benefit for these.
986 GnuTLS appears to not support tickets server-side by default (we don't
987 call gnutls_session_ticket_enable_server()) but client side is enabled
988 by default on recent versions (3.1.3 +) unless the PFS priority string
991 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
992 <https://reproducible-builds.org/specs/source-date-epoch/>.
994 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
995 the check for any unsuccessful recipients did not notice the limit, and
996 erroneously found still-pending ones.
998 JH/08 Pipeline CHUNKING command and data together, on kernels that support
999 MSG_MORE. Only in-clear (not on TLS connections).
1001 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1002 transport-filter is involved we can buffer the headers in memory for
1003 creating the signature, and read the spool data file once for the
1004 signature and again for transmission.
1006 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1007 4.77 as the kernel support then wasn't solid, having issues in 64bit
1008 mode. Now, it's been long enough. Add support for FreeBSD also.
1010 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1011 case where the routing stage had gathered several addresses to send to
1012 a host before calling the transport for the first, we previously failed
1013 to close down TLS in the old transport process before passing the TCP
1014 connection to the new process. The new one sent a STARTTLS command
1015 which naturally failed, giving a failed delivery and bloating the retry
1016 database. Investigation and fix prototype from Wolfgang Breyha.
1018 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1019 false-negatives in the check that the sender had not preempted a response
1020 or prompt from Exim (running as a server), due to that code's lack of
1021 awareness of the SMTP input buffering.
1023 PP/04 Add commandline_checks_require_admin option.
1024 Exim drops privileges sanely, various checks such as -be aren't a
1025 security problem, as long as you trust local users with access to their
1026 own account. When invoked by services which pass untrusted data to
1027 Exim, this might be an issue. Set this option in main configuration
1028 AND make fixes to the calling application, such as using `--` to stop
1031 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1032 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1033 responses to those, into a single TLS record each way (this usually means
1034 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1037 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1038 affects you only if you're dancing at the edge of the param size limits.
1039 If you are, and this message makes sense to you, then: raise the
1040 configured limit or use OpenSSL 1.1. Nothing we can do for older
1043 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1044 cmdline element to get the documented default one. Previously it was
1047 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1048 are unsuable for non-message-specific reasons - eg. connection timeout,
1049 banner-time rejection.
1051 JH/16 Fix logging of delivery remote port, when specified by router, under
1054 PP/06 Repair manualroute's ability to take options in any order, even if one
1055 is the name of a transport.
1058 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1060 JH/17 Change the list-building routines interface to use the expanding-string
1061 triplet model, for better allocation and copying behaviour.
1063 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1064 Previously it was constructed the first time a possibly-matching string
1065 was met in the configuration file input during startup; now it is done
1068 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1069 compatible one, to avoid the (poorly documented) possibility of a config
1070 file in the working directory redirecting the DB files, possibly correpting
1071 some existing file. CVE-2017-10140 assigned for BDB.
1073 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1074 cache-hot. Previously, although the result was properly cached, the
1075 initial verify call returned a defer.
1077 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1078 the main verify for receipient in uncached-mode.
1080 JH/22 Retire historical build files to an "unsupported" subdir. These are
1081 defined as "ones for which we have no current evidence of testing".
1083 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1084 if present. Previously it was ignored.
1086 JH/24 Start using specified-initialisers in C structure init coding. This is
1087 a C99 feature (it's 2017, so now considered safe).
1089 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1090 if was a fixed-sized field and bitmask ops via macros; it is now more
1093 PP/07 GitHub PR 56: Apply MariaDB build fix.
1094 Patch provided by Jaroslav Škarvada.
1096 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1097 during Coverity cleanups [4.87 JH/47]
1098 Diagnosis and fix provided by Michael Fischer v. Mollard.
1100 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1101 the right size to place the terminating semicolon on its own folded
1102 line, the header hash was calculated to an incorrect value thanks to
1103 the (relaxed) space the fold became.
1105 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1106 and confused the parent.
1108 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1109 which could crash as a result. This could lead to undeliverable messages.
1111 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1114 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1115 matters on fast-turnover and PID-randomising systems, which were getting
1116 out-of-order delivery.
1118 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1119 a possibly-overlapping copy. The symptom was that "Remote host closed
1120 connection in response to HELO" was logged instead of the actual 4xx
1123 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1124 Previously only that bufferd was discarded, resulting in SYMTP command
1127 JH/32 DKIM: when a message has multiple signatures matching an identity given
1128 in dkim_verify_signers, run the dkim acl once for each. Previously only
1129 one run was done. Bug 2189.
1131 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1132 "panic the current process" to "deliberately defer". The panic log is
1133 still written with the problem list name; the mail and reject logs now
1134 get a temp-reject line for the message that was being handled, saying
1135 something like "domains check lookup or other defer". The SMTP 451
1136 message is still "Temporary local problem".
1138 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1139 A crafted sequence of BDAT commands could result in in-use memory beeing
1140 freed. CVE-2017-16943.
1142 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1143 from SMTP input. Previously it was always done; now only done for DATA
1144 and not BDAT commands. CVE-2017-16944.
1146 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1147 to the message (such as an overlong header line). Previously this was
1148 not done and we did not exit BDAT mode. Followon from the previous item
1149 though a different problem.
1155 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1156 than -2003 did; needs libidn2 in addition to libidn.
1158 JH/02 The path option on a pipe transport is now expanded before use.
1160 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1161 Patch provided by "Björn", documentation fix added too.
1163 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1164 missing a wire-to-host endian conversion.
1166 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1167 close after a BDAT command line could be taken as a following command,
1168 giving a synch failure. Fix by only checking for synch immediately
1169 before acknowledging the chunk.
1171 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1172 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1173 macro. Patches provided by Josh Soref.
1175 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1176 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1177 by IANA as a service extension. However, John Klensin suggests that we
1180 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1181 the dkim code may be unix-mode line endings rather than smtp wire-format
1182 CRLF, so prepend a CR to any bare LF.
1184 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1185 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1187 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1188 the first were themselves being wrongly included in the feed into dkim
1189 processing; with most chunk sizes in use this resulted in an incorrect
1190 body hash calculated value.
1192 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1193 DKIM signature block, for verification. Although advised against by
1194 standards it is specifically not ruled illegal.
1196 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1198 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1199 missing a body hash (the bh= tag).
1201 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1202 It seems that HAProxy sends the Proxy Protocol information in clear and
1203 only then does a TLS startup, so do the same.
1205 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1206 TCP connections (such as for Spamd) unless the daemon successfully set
1207 Fast Open mode on its listening sockets. This fixes breakage seen on
1208 too-old kernels or those not configured for Fast Open, at the cost of
1209 requiring both directions being enabled for TFO, and TFO never being used
1210 by non-daemon-related Exim processes.
1212 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1213 endings, at least on the first header line. Try to canonify any that get
1214 past that check, despite the cost.
1216 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1217 now limited to an arbitrary five deep, while parsing addresses with the
1218 strip_excess_angle_brackets option enabled.
1220 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1221 instead leave the unprompted TLS handshake in socket buffer for the
1222 TLS library to consume.
1224 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1226 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1228 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1229 region in one operation, for speed, and this leaves assigned pointers
1230 dangling. Add checks run only under the testsuite which checks all
1231 variables at a store-reset and panics on a dangling pointer; add code
1232 explicitly nulling out all the variables discovered. Fixes one known
1233 bug: a transport crash, where a dangling pointer for $sending_ip_address
1234 originally assigned in a verify callout, is re-used.
1236 PP/06 Drop '.' from @INC in various Perl scripts.
1238 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1240 PP/08 Reduce a number of compilation warnings under clang; building with
1241 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1242 should be warning-free.
1244 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1246 HS/01 Fix portability problems introduced by PP/08 for platforms where
1247 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1249 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1250 chunk. This allows us to accept broken chunked messages. We need a more
1251 general solution here.
1253 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1254 already-broken messages in the queue.
1256 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1258 JH/19 Fix reference counting bug in routing-generated-address tracking.
1264 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1265 supports it and a size is available (ie. the sending peer gave us one).
1267 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1268 years of being deprecated). The replacements are the ACLs
1269 acl_smtp_mime and acl_not_smtp_mime.
1271 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1272 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1273 or even an in-clear connection were permitted. Now, if the host lookup
1274 was dnssec and dane was requested then the host is only used if the
1275 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1276 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1277 if one fails this test.
1278 This means that a poorly-configured remote DNS will make it incommunicado;
1279 but it protects against a DNS-interception attack on it.
1281 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1282 non-noisy when a race steals the message being considered.
1284 JH/05 If main configuration option tls_certificate is unset, generate a
1285 self-signed certificate for inbound TLS connections.
1287 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1288 in rewrites and routers.
1290 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1291 and logged a warning sing 4.83; now they are a configuration file error.
1293 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1294 (lacking @domain). Apply the same qualification processing as RCPT.
1296 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1298 JH/10 Support ${sha256:} applied to a string (as well as the previous
1301 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1302 a cutthrough deliver is pending, as we always want to make a connection.
1303 This also avoids re-routing the message when later placing the cutthrough
1304 connection after a verify cache hit.
1305 Do not update it with the verify result either.
1307 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1308 when routing results in more than one destination address.
1310 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1311 signing (which inhibits the cutthrough capability). Previously only
1312 the presence of an option was tested; now an expansion evaluating as
1313 empty is permissible (obviously it should depend only on data available
1314 when the cutthrough connection is made).
1316 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1317 the relevant preceding SMTP command did not note the pipelining mode.
1319 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1320 Previously they were not counted.
1322 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1323 as one having no matching records. Previously we deferred the message
1324 that needed the lookup.
1326 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1327 distinguished as "(=".
1329 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1330 for missing MX records. Previously it only worked for missing A records.
1332 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1334 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1335 after the data-go-ahead and data-ack. Patch from Jason Betts.
1337 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1338 even for a "none" policy. Patch from Tony Meyer.
1340 JH/22 Fix continued use of a connection for further deliveries. If a port was
1341 specified by a router, it must also match for the delivery to be
1344 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1345 When one of the recipients of a message was unsuitable for the connection
1346 (has no matching addresses), we lost track of needing to mark it
1347 deferred. As a result mail would be lost.
1349 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1351 JH/25 Decoding ACL controls is now done using a binary search; the source code
1352 takes up less space and should be simpler to maintain. Merge the ACL
1353 condition decode tables also, with similar effect.
1355 JH/26 Fix problem with one_time used on a redirect router which returned the
1356 parent address unchanged. A retry would see the parent address marked as
1357 delivered, so not attempt the (identical) child. As a result mail would
1360 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1361 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1362 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1365 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1366 Main config options check_{log,spool}_{inodes,space} are now
1367 100 inodes, 10MB unless set otherwise in the configuration.
1369 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1370 Previously it only applied to the main-section connection policy
1373 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1375 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1376 by me. Added RFC7919 DH primes as an alternative.
1378 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1379 are not in the system include path.
1381 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1382 GnuTLS, when a session startup failed (eg because the client disconnected)
1383 Exim did stdio operations after fclose. This was exposed by a recent
1384 change which nulled out the file handle after the fclose.
1386 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1387 signed directly by the cert-signing cert, rather than an intermediate
1388 OCSP-signing cert. This is the model used by LetsEncrypt.
1390 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1392 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1393 an incoming connection.
1395 HS/02 Bug 1802: Do not half-close the connection after sending a request
1398 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1399 fallback to "prime256v1".
1401 JH/34 SECURITY: Use proper copy of DATA command in error message.
1402 Could leak key material. Remotely exploitable. CVE-2016-9963.
1408 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1409 and 3.4.4 - once the server is enabled to respond to an OCSP request
1410 it does even when not requested, resulting in a stapling non-aware
1411 client dropping the TLS connection.
1413 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1414 support variable-length bit vectors. No functional change.
1416 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1417 The I= interface field on outgoing lines is now after the H= remote
1418 host field, same as incoming lines. There is a separate
1419 outgoing_interface log selector which allows you to disable the
1422 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1423 If not running log_selector +smtp_connection the mainlog would be held
1424 open indefinitely after a "too many connections" event, including to a
1425 deleted file after a log rotate. Leave the per net connection logging
1426 leaving it open for efficiency as that will be quickly detected by the
1427 check on the next write.
1429 HS/01 Bug 1671: Fix post transport crash.
1430 Processing the wait-<transport> messages could crash the delivery
1431 process if the message IDs didn't exist for some reason. When
1432 using 'split_spool_directory=yes' the construction of the spool
1433 file name failed already, exposing the same netto behaviour.
1435 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1436 mime_regex ACL conditions.
1438 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1439 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1440 to HELO, local diagnostic string.
1442 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1443 log line to debug. Even when configured with a tls authenticator many
1444 client connections are expected to not authenticate in this way, so
1445 an authenticate fail is not an error.
1447 HS/02 Add the Exim version string to the process info. This way exiwhat
1448 gives some more detail about the running daemon.
1450 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1451 matter for fast-change records such as DNSBLs.
1453 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1454 retry record, even if constant. There may be multiple transports with
1455 different interface settings and the retry behaviour needs to be kept
1458 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1460 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1462 JH/10 Bug 840: fix log_defer_output option of pipe transport
1464 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1465 hosts_randomize. This matters a lot when combined with mua_wrapper.
1467 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1468 ${quote_pgsql:<string>} operator.
1470 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1472 JH/14 Change header file name for freeradius-client. Relevant if compiling
1473 with Radius support; from the Gentoo tree and checked under Fedora.
1475 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1477 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1478 option transport_filter, meaning no filtering.
1480 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1482 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1483 defaults to "*" (all hosts). The variable is now available when not built
1484 with TLS, default unset, mainly to enable keeping the testsuite sane.
1485 If a server certificate is not supplied (via tls_certificate) an error is
1486 logged, and clients will find TLS connections fail on startup. Presumably
1487 they will retry in-clear.
1488 Packagers of Exim are strongly encouraged to create a server certificate
1489 at installation time.
1491 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1492 with the $config_file variable.
1494 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1495 in transport context, after the attempt, and per-recipient. The latter type
1496 is per host attempted. The event data is the error message, and the errno
1497 information encodes the lookup type (A vs. MX) used for the (first) host,
1498 and the trailing two digits of the smtp 4xx response.
1500 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1501 to write to mainlog (or rejectlog, paniclog) in the window between file
1502 creation and permissions/ownership being changed. Particularly affects
1503 installations where exicyclog is run as root, rather than exim user;
1504 result is that the running daemon panics and dies.
1506 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1508 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1509 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1510 "pri" and "weight". Note that the previous implicit priority given by the
1511 list order is no longer honoured.
1513 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1514 for DKIM processing.
1516 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1517 by defining SUPPORT_SOCKS.
1519 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1520 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1521 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1522 variables are renamed to proxy_{local,external}_{address,port}.
1524 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1525 for a build by defining SUPPORT_I18N
1527 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1528 of the query string, and make ${quote_redis:} do that quoting.
1530 JH/29 Move Events support from Experimental to mainline, enabled by default
1531 and removable for a build by defining DISABLE_EVENT.
1533 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1535 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1536 cached by the daemon.
1538 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1539 by defining LOOKUP_REDIS. The libhiredis library is required.
1541 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1542 keys are given for lookup.
1544 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1545 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1546 only supported when built with TLS support. The PolarSSL SHA routines
1547 are still used when the TLS library is too old for convenient support.
1549 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1550 openssl_options), for security. OpenSSL forces this from version 1.1.0
1551 server-side so match that on older versions.
1553 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1554 allocation for $value could be released as the expansion processing
1555 concluded, but leaving the global pointer active for it.
1557 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1558 and to use the domains and local_parts ACL conditions.
1560 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1561 incorrectly not doubled on cutthrough transmission, hence seen as a
1562 body-termination at the receiving system - resulting in truncated mails.
1563 Commonly the sender saw a TCP-level error, and retransmitted the message
1564 via the normal store-and-forward channel. This could result in duplicates
1565 received - but deduplicating mailstores were liable to retain only the
1566 initial truncated version.
1568 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1570 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1572 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1573 we're in there, support oversigning also; bug 1309.
1575 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1577 HS/04 Add support for keep_environment and add_environment options.
1579 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1580 either intentional arithmetic overflow during PRNG, or testing config-
1583 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1584 delivery resulted in actual delivery. Cancel cutthrough before DATA
1587 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1588 rejects a recipient: pass the reject to the originator.
1590 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1591 Many were false-positives and ignorable, but it's worth fixing the
1594 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1595 for the new environment-manipulation done at startup. Move the routines
1596 from being local to tls.c to being global via the os.c file.
1598 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1599 an extract embedded as result-arg for a map, the first arg for extract
1600 is unavailable so we cannot tell if this is a numbered or keyed
1601 extraction. Accept either.
1607 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1610 JH/02 The smtp transport option "multi_domain" is now expanded.
1612 JH/03 The smtp transport now requests PRDR by default, if the server offers
1615 JH/04 Certificate name checking on server certificates, when exim is a client,
1616 is now done by default. The transport option tls_verify_cert_hostnames
1617 can be used to disable this per-host. The build option
1618 EXPERIMENTAL_CERTNAMES is withdrawn.
1620 JH/05 The value of the tls_verify_certificates smtp transport and main options
1621 default to the word "system" to access the system default CA bundle.
1622 For GnuTLS, only version 3.0.20 or later.
1624 JH/06 Verification of the server certificate for a TLS connection is now tried
1625 (but not required) by default. The verification status is now logged by
1626 default, for both outbound TLS and client-certificate supplying inbound
1629 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1632 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1633 Status Notification (bounce) messages are now MIME format per RFC 3464.
1634 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1635 under the control of the dsn_advertise_hosts option, and routers may
1636 have a dsn_lasthop option.
1638 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1639 default, modifiable by a malware= option. The list separator for
1640 the options can now be changed in the usual way. Bug 68.
1642 JH/10 The smtp_receive_timeout main option is now expanded before use.
1644 JH/11 The incoming_interface log option now also enables logging of the
1645 local interface on delivery outgoing connections.
1647 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1648 if the interface and destination host and port all match.
1650 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1653 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1654 Patch from Andrew Lewis.
1656 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1657 now supports optional time-restrictions, weighting, and priority
1658 modifiers per server. Patch originally by <rommer@active.by>.
1660 JH/16 The spamd_address main option now supports a mixed list of local
1661 and remote servers. Remote servers can be IPv6 addresses, and
1662 specify a port-range.
1664 JH/17 Bug 68: The spamd_address main option now supports an optional
1665 timeout value per server.
1667 JH/18 Bug 1581: Router and transport options headers_add/remove can
1668 now have the list separator specified.
1670 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1673 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1676 JH/21 Support for the A6 type of dns record is withdrawn.
1678 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1679 rather than the verbs used.
1681 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1682 from 255 to 1024 chars.
1684 JH/24 Verification callouts now attempt to use TLS by default.
1686 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1687 are generic router options now. The defaults didn't change.
1689 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1690 Original patch from Alexander Shikoff, worked over by JH.
1692 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1693 files if ZCAT_COMMAND is not executable.
1695 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1697 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1699 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1700 Normally benign, it bites when the pair was led to by a CNAME;
1701 modern usage is to not canonicalize the domain to a CNAME target
1702 (and we were inconsistent anyway for A-only vs AAAA+A).
1704 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1706 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1707 when evaluating $sender_host_dnssec.
1709 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1710 $sender_helo_dnssec variable.
1712 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1714 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1716 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1718 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1719 documented as working, but never had. Support all but $spam_report.
1721 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1722 added for tls authenticator.
1724 HS/03 Add perl_taintmode main config option
1730 TL/01 When running the test suite, the README says that variables such as
1731 no_msglog_check are global and can be placed anywhere in a specific
1732 test's script, however it was observed that placement needed to be near
1733 the beginning for it to behave that way. Changed the runtest perl
1734 script to read through the entire script once to detect and set these
1735 variables, reset to the beginning of the script, and then run through
1736 the script parsing/test process like normal.
1738 TL/02 The BSD's have an arc4random API. One of the functions to induce
1739 adding randomness was arc4random_stir(), but it has been removed in
1740 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1741 function when detected.
1743 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1744 cause callback expansion.
1746 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1747 syntax errors in an expansion can be treated as a string instead of
1748 logging or causing an error, due to the internal use of bool_lax
1749 instead of bool when processing it.
1751 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1752 server certificates when making smtp deliveries.
1754 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1756 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1758 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1760 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1761 Merged patch from Sebastian Wiedenroth.
1763 JH/05 Fix results-pipe from transport process. Several recipients, combined
1764 with certificate use, exposed issues where response data items split
1765 over buffer boundaries were not parsed properly. This eventually
1766 resulted in duplicates being sent. This issue only became common enough
1767 to notice due to the introduction of connection certificate information,
1768 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1770 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1771 size buffer was used, resulting in syntax errors when an expansion
1774 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1775 version 3.3.6 or later.
1777 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1778 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1779 both become "event_action", the variables become $event_name, $event_data
1780 and $event_defer_errno. There is a new variable $verify_mode, usable in
1781 routers, transports and related events. The tls:cert event is now also
1782 raised for inbound connections, if the main configuration event_action
1785 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1786 early OCSP support, but no stapling (appears to be less than 1.0.0).
1788 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1789 server certificate names available under the smtp transport option
1790 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1793 JH/10 Time-related extraction expansions from certificates now use the main
1794 option "timezone" setting for output formatting, and are consistent
1795 between OpenSSL and GnuTLS compilations. Bug 1541.
1797 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1798 encoded parameter in the incoming message. Bug 1558.
1800 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1801 include certificate info, eximon was claiming there were spoolfile
1804 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1806 JH/14 Log delivery-related information more consistently, using the sequence
1807 "H=<name> [<ip>]" wherever possible.
1809 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1810 are problematic for Debian distribution, omit them from the release
1813 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1815 JH/16 Fix string representation of time values on 64bit time_t architectures.
1818 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1819 output list separator was used.
1824 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1825 checkers that were complaining about end of non-void function with no
1828 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1829 This was a regression introduced in 4.83 by another bugfix.
1831 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1833 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1834 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1840 TF/01 Correctly close the server side of TLS when forking for delivery.
1842 When a message was received over SMTP with TLS, Exim failed to clear up
1843 the incoming connection properly after forking off the child process to
1844 deliver the message. In some situations the subsequent outgoing
1845 delivery connection happened to have the same fd number as the incoming
1846 connection previously had. Exim would try to use TLS and fail, logging
1847 a "Bad file descriptor" error.
1849 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1850 utilities have not been installed.
1852 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1853 temporary space as the ACL may create new global variables.
1855 TL/01 LDAP support uses per connection or global context settings, depending
1856 upon the detected version of the libraries at build time.
1858 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1859 to extract and use the src ip:port in logging and expansions as if it
1860 were a direct connection from the outside internet. PPv2 support was
1861 updated based on HAProxy spec change in May 2014.
1863 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1865 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1866 Properly escape header and check for NULL return.
1868 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1871 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1873 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1874 characters in header names, implemented as a verify condition.
1875 Contributed by Michael Fischer v. Mollard.
1877 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1878 results permerror and temperror. Previous values are deprecated but
1879 still accepted. In a future release, err_perm and err_temp will be
1880 completely removed, which will be a backward incompatibility if the
1881 ACL tests for either of these two old results. Patch contributed by
1882 user bes-internal on the mailing list.
1884 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1886 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1887 selectors, in both main and reject logs.
1889 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1892 JH/07 Add malware type "sock" for talking to simple daemon.
1894 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1896 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1897 routers/transports under cutthrough routing.
1899 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1900 numbers. Touch up "bool" conditional to keep the same definition.
1902 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1904 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1906 JH/12 Expand items in router/transport headers_add or headers_remove lists
1907 individually rather than the list as a whole. Bug 1452.
1909 Required for reasonable handling of multiple headers_ options when
1910 they may be empty; requires that headers_remove items with embedded
1911 colons must have them doubled (or the list-separator changed).
1913 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1914 view the policy declared in the DMARC record. Currently, $dmarc_status
1915 is a combined value of both the record presence and the result of the
1918 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1920 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1921 dnslookup router and the smtp transport (applying to the forward
1924 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1925 of ldap servers used for a specific lookup. Patch provided by Heiko
1928 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1929 New variable $lookup_dnssec_authenticated for observability.
1931 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1932 Patch submitted by Lars Timman.
1934 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1936 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1937 Requires trusted mode and valid format message id, aborts otherwise.
1938 Patch contributed by Heiko Schlichting.
1940 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1941 certextract with support for various fields. Bug 1358.
1943 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1944 is requested by default, modifiable by smtp transport option
1947 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1948 operate on certificate variables to give certificate fingerprints
1949 Also new ${sha256:cert_variable}.
1951 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1953 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1954 Christian Aistleitner.
1956 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1958 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1959 file. Patch from Wolfgang Breyha.
1961 JH/25 Expand the coverage of the delivery $host and $host_address to
1962 client authenticators run in verify callout. Bug 1476.
1964 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1965 align with daemon_smtp_ports. Bug 72.
1967 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1968 support and error reporting did not work properly.
1970 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1971 and is readable. Patch from Andrew Colin Kissa.
1973 TL/14 Enhance documentation of ${run expansion and how it parses the
1974 commandline after expansion, particularly in the case when an
1975 unquoted variable expansion results in an empty value.
1977 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1979 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1980 by renaming away from it. Addresses GH issue 15, reported by
1983 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1985 TL/15 SECURITY: prevent double expansion in math comparison functions
1986 (can expand unsanitized data). Not remotely exploitable.
1993 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1995 PP/02 Make -n do something, by making it not do something.
1996 When combined with -bP, the name of an option is not output.
1998 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2001 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2002 $sender_host_name and config options to manage this, and basic check
2005 PP/05 DSCP support for outbound connections and control modifier for inbound.
2007 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2008 (Only plugin which currently uses this is kerberos4, which nobody should
2009 be using, but we should make it available and other future plugins might
2010 conceivably use it, even though it would break NAT; stuff *should* be
2011 using channel bindings instead).
2013 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2014 name; added for Sendmail compatibility; requires admin caller.
2015 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2016 just ignore it); requires trusted caller.
2017 Also parse but ignore: -Ac -Am -X<logfile>
2020 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2022 TL/02 Add +smtp_confirmation as a default logging option.
2024 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2025 Patch by Magnus Holmgren from 2007-02-20.
2027 TL/04 Bugzilla 1281 - Spec typo.
2028 Bugzilla 1283 - Spec typo.
2029 Bugzilla 1290 - Spec grammar fixes.
2031 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2033 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2035 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2036 mailing list by by Dmitry Isaikin.
2038 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2040 JH/02 Support "G" suffix to numbers in ${if comparisons.
2042 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2044 NM/01 Bugzilla 1197 - Spec typo
2045 Bugzilla 1196 - Spec examples corrections
2047 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2049 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2050 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2053 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2054 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2056 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2057 "acl {{name}{arg}...}", and optional args on acl condition
2060 JH/05 Permit multiple router/transport headers_add/remove lines.
2062 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2064 JH/07 Avoid using a waiting database for a single-message-only transport.
2065 Performance patch from Paul Fisher. Bugzilla 1262.
2067 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2070 JH/09 Add $headers_added variable, with content from use of ACL modifier
2071 add_header (but not yet added to the message). Bugzilla 199.
2073 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2074 Pulled from Bugzilla 817 by Wolfgang Breyha.
2076 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2078 (nb: this is the same fix as in Exim 4.80.1)
2080 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2083 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2086 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2088 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2089 advertise SMTP AUTH mechanism to us, instead of a generic
2090 protocol violation error. Also, make Exim more robust to bad
2091 data from the Dovecot auth socket.
2093 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2095 When a queue runner is handling a message, Exim first routes the
2096 recipient addresses, during which it prunes them based on the retry
2097 hints database. After that it attempts to deliver the message to
2098 any remaining recipients. It then updates the hints database using
2101 So if a recipient address works intermittently, it can get repeatedly
2102 deferred at routing time. The retry hints record remains fresh so the
2103 address never reaches the final cutoff time.
2105 This is a fairly common occurrence when a user is bumping up against
2106 their storage quota. Exim had some logic in its local delivery code
2107 to deal with this. However it did not apply to per-recipient defers
2108 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2110 This change adds a proper retry rule check during routing so that the
2111 final cutoff time is checked against the message's age. We only do
2112 this check if there is an address retry record and there is not a
2113 domain retry record; this implies that previous attempts to handle
2114 the address had the retry_use_local_parts option turned on. We use
2115 this as an approximation for the destination being like a local
2116 delivery, as in LMTP.
2118 I suspect this new check makes the old local delivery cutoff check
2119 redundant, but I have not verified this so I left the code in place.
2121 TF/02 Correct gecos expansion when From: is a prefix of the username.
2123 Test 0254 submits a message to Exim with the header
2127 When I ran the test suite under the user fanf2, Exim expanded
2128 the header to contain my full name, whereas it should have added
2129 a Resent-Sender: header. It erroneously treats any prefix of the
2130 username as equal to the username.
2132 This change corrects that bug.
2134 GF/01 DCC debug and logging tidyup
2135 Error conditions log to paniclog rather than rejectlog.
2136 Debug lines prefixed by "DCC: " to remove any ambiguity.
2138 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2140 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2141 Bug spotted by Jeremy Harris; was flawed since initial commit.
2142 Would have resulted in OCSP responses post-SNI triggering an Exim
2143 NULL dereference and crash.
2145 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2147 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2148 Bug detection, analysis and fix by Samuel Thibault.
2149 Bugzilla 1331, Debian bug #698092.
2151 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2153 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2154 Server implementation by Todd Lyons, client by JH.
2155 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2156 config variable "prdr_enable" controls whether the server
2157 advertises the facility. If the client requests PRDR a new
2158 acl_data_smtp_prdr ACL is called once for each recipient, after
2159 the body content is received and before the acl_smtp_data ACL.
2160 The client is controlled by both of: a hosts_try_prdr option
2161 on the smtp transport, and the server advertisement.
2162 Default client logging of deliveries and rejections involving
2163 PRDR are flagged with the string "PRDR".
2165 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2166 fclose(). Diagnosis by Todd Lyons.
2168 PP/17 Update configure.default to handle IPv6 localhost better.
2169 Patch by Alain Williams (plus minor tweaks).
2172 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2173 This is now consistent with GnuTLS, and is now documented: the
2174 previous undocumented portable approach to treating the option as
2175 unset was to force an expansion failure. That still works, and
2176 an empty string is now equivalent.
2178 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2179 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2180 not performing validation itself.
2182 PP/20 Added force_command boolean option to pipe transport.
2183 Patch from Nick Koston, of cPanel Inc.
2185 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2188 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2190 PP/21 Fix eximon continuous updating with timestamped log-files.
2191 Broken in a format-string cleanup in 4.80, missed when I repaired the
2192 other false fix of the same issue.
2193 Report and fix from Heiko Schlichting.
2196 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2197 Report from Prashanth Katuri.
2199 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2200 It's SecureTransport, so affects any MacOS clients which use the
2201 system-integrated TLS libraries, including email clients.
2203 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2204 using a MIME ACL for non-SMTP local injection.
2205 Report and assistance in diagnosis by Warren Baker.
2207 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2209 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2211 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2212 last id that failed so it may be referenced in subsequent ACL's.
2214 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2217 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2218 ldap library initialization, allowing self-signed CA's to be
2219 used. Also properly sets require_cert option later in code by
2220 using NULL (global ldap config) instead of ldap handle (per
2221 session). Bug diagnosis and testing by alxgomz.
2223 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2224 the src/util/ subdirectory.
2226 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2227 renamed to Transport Post Delivery Action by Jeremy Harris, as
2230 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2231 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2232 redis_servers = needs to be configured which will be used by the redis
2233 lookup. Patch from Warren Baker, of The Packet Hub.
2235 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2237 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2238 hostname or reverse DNS when processing a host list. Used suggestions
2239 from multiple comments on this bug.
2241 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2243 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2244 Missed a few lines, added it to make the runtest require no keyboard
2247 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2248 contains upper case chars. Make router use caseful_local_part.
2250 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2251 support when GnuTLS has been built with p11-kit.
2257 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2259 This, or similar/improved, will also be change PP/11 of 4.82.
2265 PP/01 Handle short writes when writing local log-files.
2266 In practice, only affects FreeBSD (8 onwards).
2267 Bugzilla 1053, with thanks to Dmitry Isaikin.
2269 NM/01 Bugzilla 949 - Documentation tweak
2271 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2274 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2276 PP/02 Implemented gsasl authenticator.
2278 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2280 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2281 `pkg-config foo` for cflags/libs.
2283 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2284 with rest of GSASL and with heimdal_gssapi.
2286 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2287 `pkg-config foo` for cflags/libs for the TLS implementation.
2289 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2290 properties get this fed in as external SSF. A number of robustness
2291 and debugging improvements to the cyrus_sasl authenticator.
2293 PP/08 cyrus_sasl server now expands the server_realm option.
2295 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2296 Patch by Jeremy Harris.
2298 PP/10 Added dbmjz lookup type.
2300 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2302 PP/12 MAIL args handles TAB as well as SP, for better interop with
2303 non-compliant senders.
2304 Analysis and variant patch by Todd Lyons.
2306 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2307 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2308 Patch from Dirk Mueller <dmueller@suse.com>
2310 PP/13 tls_peerdn now print-escaped for spool files.
2311 Observed some $tls_peerdn in wild which contained \n, which resulted
2312 in spool file corruption.
2314 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2315 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2316 or write after TLS renegotiation, which otherwise led to messages
2319 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2320 as a tracking header (ie: a signed header comes before the signature).
2321 Patch from Wolfgang Breyha.
2323 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2324 comma-sep list; embedded commas doubled.
2326 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2328 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2330 Report and patch from Dmitry Banschikov.
2332 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2333 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2334 failures. We appear to now support TLS1.1+ with Exim.
2336 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2337 lets Exim select keys and certificates based upon TLS SNI from client.
2338 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2339 before an outbound SMTP session. New log_selector, +tls_sni.
2341 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2342 NULL dereference. Report and patch from Alun Jones.
2344 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2345 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2346 Not seeing resolver debug output on NetBSD, but suspect this is a
2347 resolver implementation change.
2349 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2350 Left warnings. Added "eximon gdb" invocation mode.
2352 PP/21 Defaulting "accept_8bitmime" to true, not false.
2354 PP/22 Added -bw for inetd wait mode support.
2356 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2357 locate the relevant includes and libraries. Made this the default.
2359 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2360 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2362 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2363 This may cause build issues on older platforms.
2365 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2366 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2367 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2368 Added SNI support via GnuTLS too.
2369 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2371 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2373 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2374 Applied second patch from Janne, implementing suggestion to default
2375 multiple-strings-in-record handling to match SPF spec.
2377 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2379 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2380 read-only, out of scope).
2381 Patch from Wolfgang Breyha, report from Stuart Northfield.
2383 PP/29 Fix three issues highlighted by clang analyser static analysis.
2384 Only crash-plausible issue would require the Cambridge-specific
2385 iplookup router and a misconfiguration.
2386 Report from Marcin Mirosław.
2388 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2390 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2391 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2392 As part of this, removing so much warning spew let me fix some minor
2393 real issues in debug logging.
2395 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2396 assignment on my part. Fixed.
2398 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2399 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2400 Janne Snabb (who went above and beyond: thank you).
2402 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2403 string otherwise requires a connection and a bunch more work and it's
2404 relatively easy to get wrong. Should also expose TLS library linkage
2407 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2408 64-bit ${eval} (JH/03).
2410 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2411 GNU libc to support some of the 64-bit stuff, should not lead to
2412 conflicts. Defined before os.h is pulled in, so if a given platform
2413 needs to override this, it can.
2415 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2416 protection layer was required, which is not implemented.
2417 Bugzilla 1254, patch from Wolfgang Breyha.
2419 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2420 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2421 tls_dhparam take prime identifiers. Also unbreak combination of
2422 OpenSSL+DH_params+TLSSNI.
2424 PP/39 Disable SSLv2 by default in OpenSSL support.
2430 PP/01 Solaris build fix for Oracle's LDAP libraries.
2431 Bugzilla 1109, patch from Stephen Usher.
2433 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2435 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2438 TF/02 Fix a couple more cases where we did not log the error message
2439 when unlink() failed. See also change 4.74-TF/03.
2441 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2442 lock up or crash if it happened to be inside a call to libc when it
2443 got a SIGUSR1 from exiwhat.
2445 The SIGUSR1 handler appends the current process status to the process
2446 log which is later printed by exiwhat. It used to use the general
2447 purpose logging code to do this, but several functions it calls are
2448 not safe for signals.
2450 The new output code in the SIGUSR1 handler is specific to the process
2451 log, and simple enough that it's easy to inspect for signal safety.
2452 Removing some special cases also simplifies the general logging code.
2453 Removing the spurious timestamps from the process log simplifies
2456 TF/04 Improved ratelimit ACL condition.
2458 The /noupdate option has been deprecated in favour of /readonly which
2459 has clearer semantics. The /leaky, /strict, and /readonly update modes
2460 are mutually exclusive. The update mode is no longer included in the
2461 database key; it just determines when the database is updated. (This
2462 means that when you upgrade Exim will forget old rate measurements.)
2464 Exim now checks that the per_* options are used with an update mode that
2465 makes sense for the current ACL. For example, when Exim is processing a
2466 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2467 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2468 must specify per_mail/readonly. If you omit the update mode it defaults to
2469 /leaky where that makes sense (as before) or /readonly where required.
2471 The /noupdate option is now undocumented but still supported for
2472 backwards compatibility. It is equivalent to /readonly except that in
2473 ACLs where /readonly is required you may specify /leaky/noupdate or
2474 /strict/noupdate which are treated the same as /readonly.
2476 A useful new feature is the /count= option. This is a generalization
2477 of the per_byte option, so that you can measure the throughput of other
2478 aggregate values. For example, the per_byte option is now equivalent
2479 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2481 The per_rcpt option has been generalized using the /count= mechanism
2482 (though it's more complicated than the per_byte equivalence). When it is
2483 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2484 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2485 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2486 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2487 Note that using per_rcpt with a non-readonly update mode in more than
2488 one ACL will cause the recipients to be double-counted. (The per_mail
2489 and per_byte options don't have this problem.)
2491 The handling of very low rates has changed slightly. If the computed rate
2492 is less than the event's count (usually one) then this event is the first
2493 after a long gap. In this case the rate is set to the same as this event's
2494 count, so that the first message of a spam run is counted properly.
2496 The major new feature is a mechanism for counting the rate of unique
2497 events. The new per_addr option counts the number of different
2498 recipients that someone has sent messages to in the last time period. It
2499 behaves like per_rcpt if all the recipient addresses are different, but
2500 duplicate recipient addresses do not increase the measured rate. Like
2501 the /count= option this is a general mechanism, so the per_addr option
2502 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2503 example, measure the rate that a client uses different sender addresses
2504 with the options per_mail/unique=$sender_address. There are further
2505 details in the main documentation.
2507 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2509 TF/06 Removed a few PCRE remnants.
2511 TF/07 Automatically extract Exim's version number from tags in the git
2512 repository when doing development or release builds.
2514 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2515 Bugzilla 879. Patch from Paul Fisher.
2517 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2518 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2521 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2523 PP/05 Variable $av_failed, true if the AV scanner deferred.
2524 Bugzilla 1078. Patch from John Horne.
2526 PP/06 Stop make process more reliably on build failure.
2527 Bugzilla 1087. Patch from Heiko Schlittermann.
2529 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2530 Bugzilla 1089. Patch from Heiko Schlittermann.
2532 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2533 Bugzilla 1131. Patch from Holger Weiß.
2535 PP/09 Handle IPv6 addresses with SPF.
2536 Bugzilla 860. Patch from Wolfgang Breyha.
2538 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2540 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2543 PP/11 match_* no longer expand right-hand-side by default.
2544 New compile-time build option, EXPAND_LISTMATCH_RHS.
2545 New expansion conditions, "inlist", "inlisti".
2547 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2549 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2551 PP/14 fix log_write() format string regression from TF/03.
2552 Bugzilla 1152. Patch from Dmitry Isaikin.
2558 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2560 PP/02 Harmonised TLS library version reporting; only show if debugging.
2561 Layout now matches that introduced for other libraries in 4.74 PP/03.
2563 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2565 PP/04 New "dns_use_edns0" global option.
2567 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2570 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2571 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2573 TK/01 Updated PolarSSL code to 0.14.2.
2574 Bugzilla 1097. Patch from Andreas Metzler.
2576 PP/07 Catch divide-by-zero in ${eval:...}.
2577 Fixes bugzilla 1102.
2579 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2582 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2583 format-string attack -- SECURITY: remote arbitrary code execution.
2585 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2586 time unintentionally subject to list matching rules, letting the header
2587 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2588 arbitrary string expansion). This allowed for information disclosure.
2590 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2591 INT_MIN/-1 -- value coerced to INT_MAX.
2597 NM/01 Workaround for PCRE version dependency in version reporting
2600 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2601 This fixes portability to compilers other than gcc, notably
2602 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2604 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2605 makefiles for portability to HP-UX and POSIX correctness.
2607 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2608 Also via indented variable definition in the Makefile.
2609 (Debugging by Oliver Heesakkers).
2611 PP/02 Restore caching of spamd results with expanded spamd_address.
2612 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2614 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2615 Improves build reliability. Fix from: Frank Elsner
2617 NM/02 Fix wide character breakage in the rfc2047 coding
2618 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2620 NM/03 Allow underscore in dnslist lookups
2621 Fixes bug 1026. Patch from Graeme Fowler
2623 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2624 Code patches from Adam Ciarcinski of NetBSD.
2626 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2629 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2630 is logged, to avoid truncation. Patch from John Horne.
2632 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2633 Patch from Jakob Hirsch.
2635 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2636 SQL string expansion failure details.
2637 Patch from Andrey Oktyabrski.
2639 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2640 Patch from Simon Arlott.
2642 PP/09 New lookups functionality failed to compile on old gcc which rejects
2643 extern declarations in function scope.
2644 Patch from Oliver Fleischmann
2646 PP/10 Use sig_atomic_t for flags set from signal handlers.
2647 Check getgroups() return and improve debugging.
2648 Fixed developed for diagnosis in bug 927 (which turned out to be
2651 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2652 Patch from Mark Zealey.
2654 PP/12 Bugzilla 1056: Improved spamd server selection.
2655 Patch from Mark Zealey.
2657 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2658 Based on patch from Heiko Schlittermann.
2660 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2661 Patch from Uwe Doering, sign-off by Michael Haardt.
2663 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2664 variable declaration deep within a block. Bug and patch from
2667 PP/15 lookups-Makefile IRIX compatibility coercion.
2669 PP/16 Make DISABLE_DKIM build knob functional.
2671 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2672 Patch by Simon Arlott
2674 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2675 variable argument macros. Our copy now differs from upstream.
2681 TF/01 Failure to get a lock on a hints database can have serious
2682 consequences so log it to the panic log.
2684 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2685 controlled using the smtp_confirmation log selector.
2687 TF/03 Include the error message when we fail to unlink a spool file.
2689 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2690 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2691 for maintaining out-of-tree patches for some time.
2693 PP/01 Bugzilla 139: Documentation and portability issues.
2694 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2695 Handle per-OS dynamic-module compilation flags.
2697 PP/02 Let /dev/null have normal permissions.
2698 The 4.73 fixes were a little too stringent and complained about the
2699 permissions on /dev/null. Exempt it from some checks.
2700 Reported by Andreas M. Kirchwitz.
2702 PP/03 Report version information for many libraries, including
2703 Exim version information for dynamically loaded libraries. Created
2704 version.h, now support a version extension string for distributors
2705 who patch heavily. Dynamic module ABI change.
2707 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2708 privilege escalation vulnerability whereby the Exim run-time user
2709 can cause root to append content of the attacker's choosing to
2712 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2715 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2716 If dropping privileges for untrusted macros, we disabled normal logging
2717 on the basis that it would fail; for the Exim run-time user, this is not
2718 the case, and it resulted in successful deliveries going unlogged.
2719 Fixed. Reported by Andreas Metzler.
2725 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2726 only prepend for the Resent-* case. Fixes regression introduced in
2727 Exim 4.70 by NM/22 for Bugzilla 607.
2729 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2730 increasing numbers of administrators be bitten by this.
2732 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2734 PP/03 Bugzilla 994: added openssl_options main configuration option.
2736 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2738 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2740 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2742 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2743 without a peer certificate, leading to a segfault because of an
2744 assumption that peers always have certificates. Be a little more
2745 paranoid. Problem reported by Martin Tscholak.
2747 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2748 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2749 NB: ClamAV planning to remove STREAM in "middle of 2010".
2750 CL also introduces -bmalware, various -d+acl logging additions and
2751 more caution in buffer sizes.
2753 PP/09 Implemented reverse_ip expansion operator.
2755 PP/10 Bugzilla 937: provide a "debug" ACL control.
2757 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2759 PP/12 Bugzilla 973: Implement --version.
2761 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2763 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2765 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2767 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2768 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2769 ignore trailing whitespace.
2771 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2773 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2776 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2777 Notification from Dr Andrew Aitchison.
2779 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2780 ExtendedDetectionInfo response format.
2781 Notification from John Horne.
2783 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2786 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2787 XSL and documented dependency on system catalogs, with examples of how
2790 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2793 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2794 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2795 configuration file which is writeable by the Exim user or group.
2797 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2798 of configuration files to cover files specified with the -C option if
2799 they are going to be used with root privileges, not just the default
2802 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2803 option (effectively making it always true).
2805 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2806 files to be used while preserving root privileges.
2808 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2809 that rogue child processes cannot use them.
2811 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2812 run-time user, instead of root.
2814 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2815 Exim run-time user without dropping privileges.
2817 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2818 result string, instead of calling string_vformat() twice with the same
2821 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2822 for other users. Others should always drop root privileges if they use
2823 -C on the command line, even for a whitelisted configure file.
2825 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2827 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2833 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2834 $data_path, and $header_path variables; fixed documentation bugs and
2837 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2838 exipick to access non-standard spools, including the "frozen" queue
2841 NM/01 Bugzilla 965: Support mysql stored procedures.
2842 Patch from Alain Williams
2844 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2846 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2847 Patch from Andreas Metzler
2849 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2850 Patch from Kirill Miazine
2852 NM/05 Bugzilla 671: Added umask to procmail example.
2854 JJ/03 installed exipick 20100323.0, fixing doc bug
2856 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2857 directory. Notification and patch from Dan Rosenberg.
2859 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2861 TK/02 Improve log output when DKIM signing operation fails.
2863 MH/01 Treat the transport option dkim_domain as a colon separated
2864 list, not as a single string, and sign the message with each element,
2865 omitting multiple occurences of the same signer.
2867 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2868 Bugzilla 985, 986. Patch by Simon Arlott
2870 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2871 Patch by Simon Arlott
2873 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2874 MBX locking. Notification from Dan Rosenberg.
2880 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2882 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2884 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2886 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2888 NM/04 Bugzilla 915: Flag broken perl installation during build.
2894 TK/01 Added patch by Johannes Berg that expands the main option
2895 "spamd_address" if it starts with a dollar sign.
2897 TK/02 Write list of recipients to X-Envelope-Sender header when building
2898 the mbox-format spool file for content scanning (suggested by Jakob
2901 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2902 (http://www.dcc-servers.net/) support via dccifd. Activated by
2903 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2905 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2906 by Mark Daniel Reidel <mr@df.eu>.
2908 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2909 When building exim an external PCRE library is now needed -
2910 PCRE is a system library on the majority of modern systems.
2911 See entry on PCRE_LIBS in EDITME file.
2913 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2914 conversation. Added nologin parameter to request.
2915 Patch contributed by Kirill Miazine.
2917 TF/01 Do not log submission mode rewrites if they do not change the address.
2919 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2921 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2922 log files in place. Contributed by Roberto Lima.
2924 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2926 TF/03 Bugzilla 615: When checking the local_parts router precondition
2927 after a local_part_suffix or local_part_prefix option, Exim now
2928 does not use the address's named list lookup cache, since this
2929 contains cached lookups for the whole local part.
2931 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2932 Robert Millan. Documentation is in experimental-spec.txt.
2934 TF/04 Bugzilla 668: Fix parallel build (make -j).
2936 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2938 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2939 Patch provided by Jan Srzednicki.
2941 TF/05 Leading white space used to be stripped from $spam_report which
2942 wrecked the formatting. Now it is preserved.
2944 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2945 that they are available at delivery time.
2947 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2949 TF/08 TLS error reporting now respects the incoming_interface and
2950 incoming_port log selectors.
2952 TF/09 Produce a more useful error message if an SMTP transport's hosts
2953 setting expands to an empty string.
2955 NM/06 Bugzilla 744: EXPN did not work under TLS.
2956 Patch provided by Phil Pennock.
2958 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2959 Patch provided by Richard Godbee.
2961 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2962 acl_smtp_notquit, added index entry.
2964 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2965 Patch provided by Eugene Bujak.
2967 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2968 accept(). Patch provided by Maxim Dounin.
2970 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2971 Patch provided by Phil Pennock.
2973 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2975 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2976 Patch provided by Brad "anomie" Jorsch.
2978 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2979 Patch provided by Dean Brooks.
2981 TK/05 Add native DKIM support (does not depend on external libraries).
2983 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2984 Patch provided by Graeme Fowler.
2986 NM/16 Bugzilla 851: Documentation example syntax fix.
2988 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2990 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2993 NM/19 Bugzilla 745: TLS version reporting.
2994 Patch provided by Phil Pennock.
2996 NM/20 Bugzilla 167: bool: condition support.
2997 Patch provided by Phil Pennock.
2999 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3000 clients. Patch provided by Phil Pennock.
3002 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3003 Patch provided by Brad "anomie" Jorsch.
3005 NM/23 Bugzilla 687: Fix misparses in eximstats.
3006 Patch provided by Heiko Schlittermann.
3008 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3009 Patch provided by Heiko Schlittermann.
3011 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3012 plus update to original patch.
3014 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3016 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3017 Patch provided by David Brownlee.
3019 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3021 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3023 NM/30 Bugzilla 888: TLS documentation bugfixes.
3025 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3027 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3028 Unlike the original bugzilla I have changed all shell scripts in src tree.
3030 NM/33 Bugzilla 898: Transport filter timeout fix.
3031 Patch by Todd Rinaldo.
3033 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3034 Patch by Serge Demonchaux.
3036 NM/35 Bugzilla 39: Base64 decode bug fixes.
3037 Patch by Jakob Hirsch.
3039 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3041 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3043 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3045 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3051 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3052 ALT-N's libdkim that I have put here:
3053 http://duncanthrax.net/exim-experimental/
3055 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3056 were called 'true' and it seems that C99 defines that as a reserved
3057 keyword to be used with 'bool' variable types. That means you could
3058 not include C99-style headers which use bools without triggering
3059 build errors in sieve.c.
3061 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3062 as mailq or other aliases. Changed the --help handling significantly
3063 to do whats expected. exim_usage() emits usage/help information.
3065 SC/01 Added the -bylocaldomain option to eximstats.
3067 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3069 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3071 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3077 PH/01 Another patch from the Sieve maintainer.
3079 PH/02 When an IPv6 address is converted to a string for single-key lookup
3080 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3081 dots are used instead of colons so that keys in lsearch files need not
3082 contain colons. This was done some time before quoting was made available
3083 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3084 (notated using the quote facility) so as to distinguish them from IPv4
3085 keys. This meant that lookups for IP addresses in host lists did not work
3086 for iplsearch lookups.
3088 This has been fixed by arranging for IPv6 addresses to be expressed with
3089 colons if the lookup type is iplsearch. This is not incompatible, because
3090 previously such lookups could never work.
3092 The situation is now rather anomalous, since one *can* have colons in
3093 ordinary lsearch keys. However, making the change in all cases is
3094 incompatible and would probably break a number of configurations.
3096 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3099 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3100 conversion specification without a maximum field width, thereby enabling
3101 a rogue spamd server to cause a buffer overflow. While nobody in their
3102 right mind would setup Exim to query an untrusted spamd server, an
3103 attacker that gains access to a server running spamd could potentially
3104 exploit this vulnerability to run arbitrary code as the Exim user.
3106 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3107 $primary_hostname instead of what libspf2 thinks the hosts name is.
3109 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3110 a directory entry by the name of the lookup key. Previously, if a
3111 symlink pointed to a non-existing file or a file in a directory that
3112 Exim lacked permissions to read, a lookup for a key matching that
3113 symlink would fail. Now it is enough that a matching directory entry
3114 exists, symlink or not. (Bugzilla 503.)
3116 PH/03 The body_linecount and body_zerocount variables are now exported in the
3119 PH/04 Added the $dnslist_matched variable.
3121 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3122 This means they are set thereafter only if the connection becomes
3125 PH/06 Added the client_condition to authenticators so that some can be skipped
3126 by clients under certain conditions.
3128 PH/07 The error message for a badly-placed control=no_multiline_responses left
3129 "_responses" off the end of the name.
3131 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3133 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3134 (without spaces) instead of just copying the configuration text.
3136 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3138 PH/11 Added $max_received_linelength.
3140 PH/12 Added +ignore_defer and +include_defer to host lists.
3142 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3143 way in which PCRE > 7.0 is built.
3145 PH/14 Implemented queue_only_load_latch.
3147 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3148 MAIL command. The effect was to mangle the value on 64-bit systems.
3150 PH/16 Another patch from the Sieve maintainer.
3152 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3154 PH/18 If a system quota error occurred while trying to create the file for
3155 a maildir delivery, the message "Mailbox is full" was not appended to the
3156 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3157 only to a quota excession during the actual writing of the file.
3159 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3160 characters?) which causes problems in log lines. The DN values are now
3161 passed through string_printing() before being added to log lines.
3163 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3164 and InterBase are left for another time.)
3166 PH/21 Added message_body_newlines option.
3168 PH/22 Guard against possible overflow in moan_check_errorcopy().
3170 PH/23 POSIX allows open() to be a macro; guard against that.
3172 PH/24 If the recipient of an error message contained an @ in the local part
3173 (suitably quoted, of course), incorrect values were put in $domain and
3174 $local_part during the evaluation of errors_copy.
3180 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3181 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3184 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3185 issue a MAIL command.
3187 PH/02 In an ACL statement such as
3189 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3191 if a client was not listed at all, or was listed with a value other than
3192 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3193 the condition was not true (as it should be), so access was not denied.
3194 The bug was that the ! inversion was incorrectly passed on to the second
3195 item. This has been fixed.
3197 PH/03 Added additional dnslists conditions == and =& which are different from
3198 = and & when the dns lookup returns more than one IP address.
3200 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3201 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3203 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3204 FSYNC, which compiles an option called disable_fsync that allows for
3205 bypassing fsync(). The documentation is heavily laced with warnings.
3207 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3209 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3210 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3211 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3212 including adding "make clean"; (3) Added -fPIC when compiling the test
3213 dynamically loaded module, to get rid of a warning.
3215 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3216 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3217 = 0s. The bug is otherwise harmless.
3219 PH/07 There was a bug in the dovecot authenticator such that the value of
3220 $auth1 could be overwritten, and so not correctly preserved, after a
3221 successful authentication. This usually meant that the value preserved by
3222 the server_setid option was incorrect.
3224 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3226 PH/09 Installed PCRE release 7.0.
3228 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3229 run for batched SMTP input. It is now run at the start of every message
3230 in the batch. While fixing this I discovered that the process information
3231 (output by running exiwhat) was not always getting set for -bs and -bS
3232 input. This is fixed, and it now also says "batched" for BSMTP.
3234 PH/11 Added control=no_pipelining.
3236 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3237 patch, slightly modified), and move the expansion of helo_data till after
3238 the connection is made in the smtp transport (so it can use these
3241 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3243 PH/14 Added log_selector = +pid.
3245 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3247 PH/16 Add ${if forany and ${if forall.
3249 PH/17 Added dsn_from option to vary the From: line in DSNs.
3251 PH/18 Flush SMTP output before performing a callout, unless control =
3252 no_callout_flush is set.
3254 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3255 was true (the default) a successful delivery failed to delete the retry
3256 item, thus causing premature timeout of the address. The bug is now
3259 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3261 PH/21 Long custom messages for fakedefer and fakereject are now split up
3262 into multiline responses in the same way that messages for "deny" and
3263 other ACL rejections are.
3265 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3266 with slight modification.
3268 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3269 draft, changing the syntax and factoring some duplicate code".
3271 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3272 for deliveries of the second and subsequent messages over the same SMTP
3275 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3276 ${reduce, with only minor "tidies".
3278 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3280 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3281 expansion side effects.
3283 PH/27 When a message times out after an over-quota error from an Exim-imposed
3284 quota, the bounce message says "mailbox is full". This message was not
3285 being given when it was a system quota that was exceeded. It now should
3288 MH/03 Made $recipients available in local_scan(). local_scan() already has
3289 better access to the recipient list through recipients_list[], but
3290 $recipients can be useful in postmaster-provided expansion strings.
3292 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3293 in the case of a MAIL command with additional options following the
3294 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3295 were accidentally chopped off.
3297 PH/29 SMTP synchronization checks are implemented when a command is read -
3298 there is a check that no more input is waiting when there shouldn't be
3299 any. However, for some commands, a delay in an ACL can mean that it is
3300 some time before the response is written. In this time, more input might
3301 arrive, invalidly. So now there are extra checks after an ACL has run for
3302 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3303 pipelining has not been advertised.
3305 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3307 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3308 returned in the SMTP response when smtp_return_error_details was false.
3309 This has been fixed.
3311 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3312 instead of the C I/O that was originally supplied, because problems were
3313 reported on Solaris.
3315 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3316 Exim which did not show up earlier: it was assuming that a call to
3317 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3318 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3319 was a macro that became an assignment, so it seemed to work. This has
3320 changed to a proper function call with a void return, hence the compile
3321 error. Exim's code has been fixed.
3323 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3326 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3328 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3330 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3331 because it turns out to be too big or there is a timeout) there is no
3332 "Completed" line in the log. When some messages of this type were
3333 selected by exigrep, they were listed as "not completed". Others were
3334 picked up by some special patterns. I have improved the selection
3335 criteria to be more general.
3337 PH/38 The host_find_failed option in the manualroute router can now be set
3338 to "ignore", to completely ignore a host whose IP address cannot be
3339 found. If all hosts are ignored, the behaviour is controlled by the new
3340 host_all_ignored option.
3342 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3343 homing or because of multiple MX records with /mx) generated more than
3344 one IP address, and the following item turned out to be the local host,
3345 all the secondary addresses of the first item were incorrectly removed
3346 from the list, along with the local host and any following hosts (which
3347 is what is supposed to happen).
3349 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3350 whoever called Exim into the -H file. In the case of the daemon it was
3351 behaving confusingly. When first started, it used values for whoever
3352 started the daemon, but after a SIGHUP it used the Exim user (because it
3353 calls itself on a restart). I have changed the code so that it now always
3356 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3357 message are rejected with the same error (e.g. no authentication or bad
3358 sender address), and a DATA command is nevertheless sent (as can happen
3359 with PIPELINING or a stupid MUA), the error message that was given to the
3360 RCPT commands is included in the rejection of the DATA command. This is
3361 intended to be helpful for MUAs that show only the final error to their
3364 PH/42 Another patch from the Sieve maintainer.
3366 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3367 Eximstats - Fixed some broken HTML links and added missing column headers
3369 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3370 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3372 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3374 PH/43 Yet another patch from the Sieve maintainer.
3376 PH/44 I found a way to check for a TCP/IP connection going away before sending
3377 the response to the final '.' that terminates a message, but only in the
3378 case where the client has not sent further data following the '.'
3379 (unfortunately, this is allowed). However, in many cases there won't be
3380 any further data because there won't be any more messages to send. A call
3381 to select() can be used: if it shows that the input is "ready", there is
3382 either input waiting, or the socket has been closed. An attempt to read
3383 the next input character can distinguish the two cases. Previously, Exim
3384 would have sent an OK response which the client would never have see.
3385 This could lead to message repetition. This fix should cure that, at
3386 least in a lot of common cases.
3388 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3389 advertised in response to EHLO.
3395 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3396 fixed by 4.65/MH/01 (is this a record?) are fixed:
3398 (i) An empty string was always treated as zero by the numeric comparison
3399 operators. This behaviour has been restored.
3401 (ii) It is documented that the numeric comparison operators always treat
3402 their arguments as decimal numbers. This was broken in that numbers
3403 starting with 0 were being interpreted as octal.
3405 While fixing these problems I realized that there was another issue that
3406 hadn't been noticed. Values of message_size_limit (both the global option
3407 and the transport option) were treated as octal if they started with 0.
3408 The documentation was vague. These values are now always treated as
3409 decimal, and I will make that clear in the documentation.
3415 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3416 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3419 MH/01 Don't check that the operands of numeric comparison operators are
3420 integers when their expansion is in "skipping" mode (fixes bug
3421 introduced by 4.64-PH/07).
3423 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3424 child addresses, Exim now panics and dies. Previously, because the count
3425 is held in a short int, deliveries were likely to be lost. As such a
3426 large number of recipients for a single message is ridiculous
3427 (performance will be very, very poor), I have chosen to impose a limit
3428 rather than extend the field.
3434 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3435 leftover -K file (the existence of which was triggered by #402).
3436 While we were at it, introduced process PID as part of the -K
3437 filename. This should rule out race conditions when creating
3440 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3441 processing considerably. Previous code took too long for large mails,
3442 triggering a timeout which in turn triggers #401.
3444 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3445 in the DK code in transports.c. sendfile() is not really portable,
3446 hence the _LINUX specificness.
3448 TF/01 In the add_headers option to the mail command in an Exim filter,
3449 there was a bug that Exim would claim a syntax error in any
3450 header after the first one which had an odd number of characters
3453 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3454 callout verification, Exim cached a "reject" for the entire domain. This
3455 is correct for most verifications, but it is not correct for a recipient
3456 verification with use_sender or use_postmaster set, because in that case
3457 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3458 case of MAIL FROM:<> rejection from other early rejections (e.g.
3459 rejection of HELO). When verifying a recipient using a non-null MAIL
3460 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3461 Whatever the result of the callout, the value of the domain cache is
3462 left unchanged (for any other kind of callout, getting as far as trying
3463 RCPT means that the domain itself is ok).
3465 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3468 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3469 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3470 session, the daemon ignores ECONNECT errors and logs others; it now
3471 ignores EPIPE as well.
3473 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3474 (quoted-printable decoding).
3476 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3477 later the small subsequent patch to fix an introduced bug.
3479 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3481 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3483 PH/08 An error is now given if message_size_limit is specified negative.
3485 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3486 to be given (somewhat) arbitrary names.
3488 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3491 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3492 miscellaneous code fixes
3494 PH/10 Added the log_reject_target ACL modifier to specify where to log
3497 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3498 hostname. This is wrong, because it relates to the incoming message (and
3499 probably the interface on which it is arriving) and not to the outgoing
3500 callout (which could be using a different interface). This has been
3501 changed to use the value of the helo_data option from the smtp transport
3502 instead - this is what is used when a message is actually being sent. If
3503 there is no remote transport (possible with a router that sets up host
3504 addresses), $smtp_active_hostname is used.
3506 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3507 tweaks were necessary in order to get it to work (see also 21 below):
3508 (a) The code assumed that strncpy() returns a negative number on buffer
3509 overflow, which isn't the case. Replaced with Exim's string_format()
3511 (b) There were several signed/unsigned issues. I just did the minimum
3512 hacking in of casts. There is scope for a larger refactoring.
3513 (c) The code used strcasecmp() which is not a standard C function.
3514 Replaced with Exim's strcmpic() function.
3515 (d) The code set only $1; it now sets $auth1 as well.
3516 (e) A simple test gave the error "authentication client didn't specify
3517 service in request". It would seem that Dovecot has changed its
3518 interface. Fortunately there's a specification; I followed it and
3519 changed what the client sends and it appears to be working now.
3521 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3524 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3525 address A is aliased to B and C, where B exists and C does not. Without
3526 -v the output is "A verified" because verification stops after a
3527 successful redirection if more than one address is generated. However,
3528 with -v the child addresses are also verified. Exim was outputting "A
3529 failed to verify" and then showing the successful verification for C,
3530 with its parentage. It now outputs "B failed to verify", showing B's
3531 parentage before showing the successful verification of C.
3533 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3534 look up a TXT record in a specific list after matching in a combined
3537 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3538 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3539 they consult the DNS. I had assumed they would set it the way they
3540 wanted; and indeed my experiments on Linux seem to show that in some
3541 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3542 To be on the safe side, however, I have now made the interface to
3543 host_find_byname() similar to host_find_bydns(), with an argument
3544 containing the DNS resolver options. The host_find_byname() function now
3545 sets these options at its start, just as host_find_bydns() does. The smtp
3546 transport options dns_qualify_single and dns_search_parents are passed to
3547 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3548 of host_find_byname() use the default settings of RES_DEFNAMES
3549 (qualify_single) but not RES_DNSRCH (search_parents).
3551 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3552 spool_read_header() do less string testing, by means of a preliminary
3553 switch on the second character of optional "-foo" lines. (This is
3554 overdue, caused by the large number of possibilities that now exist.
3555 Originally there were few.) While I was there, I also converted the
3556 str(n)cmp tests so they don't re-test the leading "-" and the first
3557 character, in the hope this might squeeze out yet more improvement.
3559 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3560 flag allowing group syntax was set by the header_syntax check but not
3561 turned off, possible causing trouble later; (2) The flag was not being
3562 set at all for the header_verify test, causing "group"-style headers to
3563 be rejected. I have now set it in this case, and also caused header_
3564 verify to ignore an empty address taken from a group. While doing this, I
3565 came across some other cases where the code for allowing group syntax
3566 while scanning a header line wasn't quite right (mostly, not resetting
3567 the flag correctly in the right place). These bugs could have caused
3568 trouble for malformed header lines. I hope it is now all correct.
3570 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3571 with the "reply" argument non-NULL. The code, however (which originally
3572 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3573 but it didn't always do it. This confused somebody who was copying the
3574 code for some other use. I have removed all the tests.
3576 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3577 feature that was used to support insecure browsers during the U.S. crypto
3578 embargo. It requires special client support, and Exim is probably the
3579 only MTA that supported it -- and would never use it because real RSA is
3580 always available. This code has been removed, because it had the bad
3581 effect of slowing Exim down by computing (never used) parameters for the
3582 RSA_EXPORT functionality.
3584 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3585 authenticator to fail if there's a tab character in the incoming data
3586 (there should never be unless someone is messing about, as it's supposed
3587 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3588 if the connection is using TLS or if the remote IP is the same as the
3589 local IP, and the "valid-client-cert option" if a client certificate has
3592 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3593 authenticators. This can be used for authorization after authentication
3594 succeeds. (In the case of plaintext, it servers for both authentication
3597 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3598 if any retry times were supplied.
3600 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3601 connection, where there is no client IP address to check. In this
3602 situation, the verify now always succeeds.
3604 PH/25 Applied John Jetmore's -Mset patch.
3606 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3608 PH/27 In a string expansion for a processed (not raw) header when multiple
3609 headers of the same name were present, leading whitespace was being
3610 removed from all of them, but trailing whitespace was being removed only
3611 from the last one. Now trailing whitespace is removed from each header
3612 before concatenation. Completely empty headers in a concatenation (as
3613 before) are ignored.
3615 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3616 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3618 PH/29 [Removed. This was a change that I later backed out, and forgot to
3619 correct the ChangeLog entry (that I had efficiently created) before
3620 committing the later change.]
3622 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3623 address errors (4xx response to RCPT) over the same connection as other
3624 messages routed to the same hosts. Such deliveries are always "forced",
3625 so retry times are not inspected. This resulted in far too many retries
3626 for the affected addresses. The effect occurred only when there were more
3627 hosts than the hosts_max_try setting in the smtp transport when it had
3628 the 4xx errors. Those hosts that it had tried were not added to the list
3629 of hosts for which the message was waiting, so if all were tried, there
3630 was no problem. Two fixes have been applied:
3632 (i) If there are any address or message errors in an SMTP delivery, none
3633 of the hosts (tried or untried) are now added to the list of hosts
3634 for which the message is waiting, so the message should not be a
3635 candidate for sending over the same connection that was used for a
3636 successful delivery of some other message. This seems entirely
3637 reasonable: after all the message is NOT "waiting for some host".
3638 This is so "obvious" that I'm not sure why it wasn't done
3639 previously. Hope I haven't missed anything, but it can't do any
3640 harm, as the worst effect is to miss an optimization.
3642 (ii) If, despite (i), such a delivery is accidentally attempted, the
3643 routing retry time is respected, so at least it doesn't keep
3644 hammering the server.
3646 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3647 in ${readsocket because some servers need this prod.
3649 PH/32 Added some extra debug output when updating a wait-xxx database.
3651 PH/33 The hint "could be header name not terminated by colon", which has been
3652 given for certain expansion errors for a long time, was not being given
3653 for the ${if def:h_colon_omitted{... case.
3655 PH/34 The spec says: "With one important exception, whenever a domain list is
3656 being scanned, $domain contains the subject domain." There was at least
3657 one case where this was not true.
3659 PH/35 The error "getsockname() failed: connection reset by peer" was being
3660 written to the panic log as well as the main log, but it isn't really
3661 panic-worthy as it just means the connection died rather early on. I have
3662 removed the panic log writing for the ECONNRESET error when getsockname()
3665 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3666 runs only) independently of the message's sender address. This meant
3667 that, if the 4xx error was in fact related to the sender, a different
3668 message to the same recipient with a different sender could confuse
3669 things. In particular, this can happen when sending to a greylisting
3670 server, but other circumstances could also provoke similar problems.
3671 I have changed the default so that the retry time for these errors is now
3672 based a combination of the sender and recipient addresses. This change
3673 can be overridden by setting address_retry_include_sender=false in the
3676 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3677 remote server are returned as part of bounce messages. This was not
3678 happening for LMTP over a pipe (the lmtp transport), but now it is the
3679 same for both kinds of LMTP.
3681 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3682 in header lines that were in fact CNAMEs. This is no longer the case.
3684 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3685 and queue runs started by the daemon processed all messages. This has
3686 been fixed so that -R and -S can now usefully be given with -q<time>.
3688 PH/40 Import PCRE release 6.7 (fixes some bugs).
3690 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3692 PH/42 Give an error if -q is specified more than once.
3694 PH/43 Renamed the variables $interface_address and $interface_port as
3695 $received_ip_address and $received_port, to make it clear that these
3696 values apply to message reception, and not to the outgoing interface when
3697 a message is delivered. (The old names remain recognized, of course.)
3699 PH/44 There was no timeout on the connect() call when using a Unix domain
3700 socket in the ${readsocket expansion. There now is.
3702 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3703 be meaningful with "accept".
3705 SC/01 Eximstats V1.43
3706 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3708 SC/02 Eximstats V1.44
3709 Use a glob alias rather than an array ref in the generated
3710 parser. This improves both readability and performance.
3712 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3713 Collect SpamAssassin and rejection statistics.
3714 Don't display local sender or destination tables unless
3715 there is data to show.
3716 Added average volumes into the top table text output.
3718 SC/04 Eximstats V1.46
3719 Collect data on the number of addresses (recipients)
3720 as well as the number of messages.
3722 SC/05 Eximstats V1.47
3723 Added 'Message too big' to the list of mail rejection
3724 reasons (thanks to Marco Gaiarin).
3726 SC/06 Eximstats V1.48
3727 Mainlog lines which have GMT offsets and are too short to
3728 have a flag are now skipped.
3730 SC/07 Eximstats V1.49 (Alain Williams)
3731 Added the -emptyok flag.
3733 SC/08 Eximstats V1.50
3734 Fixes for obtaining the IP address from reject messages.
3736 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3737 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3738 whitespace changes from 4.64-PH/27
3740 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3743 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3744 are found, allow negative numbers in numeric criteria)
3746 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3748 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3749 to match changes made in 4.64-PH/43
3751 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3753 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3754 "domain" is actually an IP address, and force a failure. This locks out
3755 those revolvers/nameservers that support "A-for-A" lookups, in
3756 contravention of the specifications.
3758 PH/48 When a host name was looked up from an IP address, and the subsequent
3759 forward lookup of the name timed out, the host name was left in
3760 $sender_host_name, contrary to the specification.
3762 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3763 restricted to single-key lookups, Exim was not diagnosing an error if
3764 * or *@ was used with a query-style lookup.
3766 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3768 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3769 long ago, but noone interested enough thought of it. Let's just say that
3770 the "1.1" means that there are some new functions that weren't there at
3771 some point in the past.
3773 PH/51 Error processing for expansion failure of helo_data from an smtp
3774 transport during callout processing was broken.
3776 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3777 tested/used via the -bh/-bhc/-bs options.
3779 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3780 bug, fixed in subsequent PCRE releases).
3782 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3783 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3785 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3791 SC/01 Use a glob alias rather than an array ref in eximstats generated
3792 parser. This improves both readability and performance.
3794 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3795 Don't display local sender or destination tables in eximstats unless
3796 there is data to show.
3797 Added average volumes into the eximstats top table text output.
3799 SC/03 Collect data on the number of addresses (recipients) as well
3800 as the number of messages in eximstats.
3802 TF/01 Correct an error in the documentation for the redirect router. Exim
3803 does (usually) call initgroups() when daemonizing.
3805 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3806 with consistent privilege compared to when running as a daemon.
3808 TF/03 Note in the spec that $authenticated_id is not set for local
3809 submissions from trusted users.
3811 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3812 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3814 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3815 by adding some example configuration directives to the default
3816 configuration file. A little bit of work is required to uncomment the
3817 directives and define how usernames and passwords are checked, but
3818 there is now a framework to start from.
3820 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3821 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3822 without this. I don't know how relevant this is to other LDAP libraries.
3824 PH/02 Add the verb name to the "unknown ACL verb" error.
3826 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3828 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3830 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3831 directory not expanded when it should be if an expanded home directory
3832 was set for the address (which is overridden by the transport).
3834 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3837 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3838 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3839 because it is too late at that time, and has no effect.
3841 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3842 security issue with \' (bugzilla #107). I could not use the
3843 PQescapeStringConn() function, because it needs a PGconn value as one of
3846 PH/08 When testing addresses using -bt, indicate those final addresses that
3847 are duplicates that would not cause an additional delivery. At least one
3848 person was confused, thinking that -bt output corresponded to deliveries.
3849 (Suppressing duplicates isn't a good idea as you lose the information
3850 about possibly different redirections that led to the duplicates.)
3852 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3853 systems where poll() doesn't work, in particular OS X.
3855 PH/10 Added more information to debugging output for retry time not reached.
3857 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3858 operations in malware.c.
3860 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3863 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3864 syslog_duplication set false, log lines that would normally be written
3865 both the the main log and to the reject log were not written to syslog at
3868 PH/14 In the default configuration, change the use of "message" in ACL warn
3869 statements to "add_header".
3871 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3872 not followed by a command (e.g. "seen endif").
3874 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3875 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3878 PH/17 Added extra conditions to the default value of delay_warning_condition
3882 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3883 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3884 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3887 The Auto-Submitted: and various List- headers are standardised, whereas I
3888 don't think Precedence: ever was.
3890 PH/18 Refactored debugging code in route_finduser() to show more information,
3891 in particular, the error code if getpwnam() issues one.
3893 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3894 This is apparently needed in addition to the PH/07 change above to avoid
3895 any possible encoding problems.
3897 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3898 but not after initializing Perl.
3900 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3901 output them only if debugging. By default they are written stderr,
3902 apparently, which is not desirable.
3904 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3907 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3910 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3912 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3913 authenticated or an ident call has been made. Suppress the default
3914 values for $authenticated_id and $authenticated_sender (but permit -oMai
3915 and -oMas) when testing with -bh.
3917 PH/24 Re-jigged the order of the tests in the default configuration so that the
3918 tests for valid domains and recipients precede the DNS black list and CSA
3919 tests, on the grounds that those ones are more expensive.
3921 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3922 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3923 This bug exists in every version of Exim that I still have, right back to
3926 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3927 However, an attempt to turn on case-sensitivity in a regex key by
3928 including (?-i) didn't work because the subject string was already
3929 lowercased, and the effects were non-intuitive. It turns out that a
3930 one-line patch can be used to allow (?-i) to work as expected.
3936 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3937 other effects) broke the use of negated acl sub-conditions.
3939 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3942 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3943 "Deny" causes Exim to reject the incoming connection with a 554 error.
3944 Unfortunately, if there is a major crisis, such as a disk failure,
3945 tcp-wrappers gives "deny", whereas what one would like would be some
3946 kind of temporary error. A kludge has been added to help with this.
3947 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3948 554 error is used if errno is still zero or contains ENOENT (which occurs
3949 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3952 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3954 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3955 errors. Otherwise a message that provokes a temporary error (when other
3956 messages do not) can cause a whole host to time out.
3958 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3959 addresses were routed directly to files or pipes from a redirect router.
3960 File deliveries just didn't batch; pipe deliveries might have suffered
3963 PH/06 A failure to get a lock for a hints database would erroneously always say
3964 "Failed to get write lock", even when it was really a read lock.
3966 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3967 of 0600. This has been changed to use the value of the lockfile_mode
3968 option (which defaults to 0600).
3970 PH/08 Applied small patch from the Sieve maintainer.
3972 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3973 folder from quota calculations, a direct delivery into this folder messed
3974 up the contents of the maildirsize file. This was because the regex was
3975 used only to exclude .Trash (or whatever) when the size of the mailbox
3976 was calculated. There was no check that a delivery was happening into an
3977 excluded directory. This bug has been fixed by ignoring all quota
3978 processing for deliveries into excluded directories.
3980 PH/10 Added the maildirfolder_create_regex option to appendfile.
3986 PH/01 The code for finding all the local interface addresses on a FreeBSD
3987 system running IPv6 was broken. This may well have applied to all BSD
3988 systems, as well as to others that have similar system calls. The broken
3989 code found IPv4 interfaces correctly, but gave incorrect values for the
3990 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3991 that it would not match correctly against @[] and not recognize the IPv6
3994 PH/02 The ipliteral router was not recognizing addresses of the form user@
3995 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3997 PH/03 Added disable_ipv6.
3999 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4000 decoded form, because it is most often used to construct To: headers
4001 lines in autoreplies, and the decoded form may well be syntactically
4002 invalid. However, $reply_address has leading white space removed, and all
4003 newlines turned into spaces so that the autoreply transport does not
4006 PH/05 If group was specified without a user on a router, and no group or user
4007 was specified on a transport, the group from the router was ignored.
4009 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4010 for visible compile-time settings that can be used to change these
4011 numbers, for those that want even more. Backwards compatibility with old
4012 spool files has been maintained. However, going back to a previous Exim
4013 release will lost any variables that are in spool files.
4015 PH/07 Two small changes when running in the test harness: increase delay when
4016 passing a TCP/IP connection to a new process, in case the original
4017 process has to generate a bounce, and remove special handling of
4018 127.0.0.2 (sic), which is no longer necessary.
4020 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4021 be the same on different OS.
4023 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4026 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4027 whether --show-vars was specified or not
4029 JJ/02 exipick: Added support for new ACL variable spool format introduced
4032 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4033 syntactically invalid From: or Reply-to: line, and a filter used this to
4034 generate an autoreply, and therefore failed to obtain an address for the
4035 autoreply, Exim could try to deliver to a non-existent relative file
4036 name, causing unrelated and misleading errors. What now happens is that
4037 it logs this as a hard delivery error, but does not attempt to create a
4040 PH/11 The exinext utility has a -C option for testing purposes, but although
4041 the given file was scanned by exinext itself; it wasn't being passed on
4042 when Exim was called.
4044 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4045 an end-of-file indication when reading a command response.
4047 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4048 compiled. In many other places in Exim, IPv6 addresses are always
4049 recognized, so I have changed this. It also means that IPv4 domain
4050 literals of the form [IPV4:n.n.n.n] are now always recognized.
4052 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4053 used if the router is not running as root, for example, when verifying at
4054 ACL time, or when using -bh. The debugging output from this situation was
4055 non-existent - all you got was a failure to exec. I have made two
4058 (a) Failures to set uid/gid, the current directory, or a process leader
4059 in a subprocess such as that created by queryprogram now generate
4060 suitable debugging output when -d is set.
4062 (b) The queryprogram router detects when it is not running as root,
4063 outputs suitable debugging information if -d is set, and then runs
4064 the subprocess without attempting to change uid/gid.
4066 PH/15 Minor change to Makefile for building test_host (undocumented testing
4069 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4070 additional section of a DNS packet that returns MX or SRV records.
4071 Instead, it always explicitly searches for A/AAAA records. This avoids
4072 major problems that occur when a DNS server includes only records of one
4073 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4074 fixed another bug: if SRV records were looked up and the corresponding
4075 address records were *not* found in the additional section, the port
4076 values from the SRV records were lost.
4078 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4079 using the correct key (the original address) when searching the retry
4080 rules in order to find which one to use for generating the retry hint.
4082 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4083 adding the default one. Similarly, if it contains a Reply-To: header, the
4084 errors_reply_to option, if set, is not used.
4086 PH/19 When calculating a retry time, Exim used to measure the "time since
4087 failure" by looking at the "first failed" field in the retry record. Now
4088 it does not use this if it is later than than the arrival time of the
4089 message. Instead it uses the arrival time. This makes for better
4090 behaviour in cases where some deliveries succeed, thus re-setting the
4091 "first failed" field. An example is a quota failure for a huge message
4092 when small messages continue to be delivered. Without this change, the
4093 "time since failure" will always be short, possible causing more frequent
4094 delivery attempts for the huge message than are intended.
4095 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4097 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4098 $1, $2, $3) because the numerical variables can be reset during some
4099 expansion items (e.g. "match"), thereby losing the authentication data.
4101 PH/21 Make -bV show the size of off_t variables so that the test suite can
4102 decide whether to run tests for quotas > 2G.
4104 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4105 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4106 filecount value is greater than 2G or if a quota value is greater than 2G
4107 on a system where the size of off_t is not greater than 4, a panic error
4110 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4111 never match. The debug and -bh output now contains an explicit error
4112 message indicating a malformed IPv4 address or mask.
4114 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4115 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4116 PH/23 above applies.
4118 PH/25 Do not write to syslog when running in the test harness. The only
4119 occasion when this arises is a failure to open the main or panic logs
4120 (for which there is an explicit test).
4122 PH/26 Added the /no_tell option to "control=freeze".
4124 PH/27 If a host name lookup failed very early in a connection, for example, if
4125 the IP address matched host_lookup and the reverse lookup yielded a name
4126 that did not have a forward lookup, an error message of the form "no IP
4127 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4128 could be logged. Now it outputs the IP address instead of "NULL".
4130 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4131 allows the sender and the authenticated sender to be set when
4132 submitting a message from within Exim. Since child_open_exim() is
4133 documented for local_scan(), the new function should be too.
4135 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4136 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4137 results in an empty string is now treated as unset.
4139 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4141 PH/31 Added sender_verify_fail logging option.
4143 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4144 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4145 tidied the source and removed it altogether.
4147 PH/33 When a queue run was abandoned because the load average was too high, a
4148 log line was always written; now it is written only if the queue_run log
4149 selector is set. In addition, the log line for abandonment now contains
4150 information about the queue run such as the pid. This is always present
4151 in "start" and "stop" lines but was omitted from the "abandon" line.
4153 PH/34 Omit spaces between a header name and the colon in the error message that
4154 is given when verify = headers_syntax fails (if there are lots of them,
4155 the message gets confusing).
4157 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4158 names, as there are now some PTR records that contain slashes. This check
4159 is only to protect against broken name servers that fall over on strange
4160 characters, so the fact that it applies to all lookups doesn't matter.
4162 PH/36 Now that the new test suite is complete, we can remove some of the
4163 special code in Exim that was needed for the old test suite. For example,
4164 sorting DNS records because real resolvers return them in an arbitrary
4165 order. The new test suite's fake resolver always returns records in the
4168 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4169 bounces) except when queue_only is set, to avoid logging races between
4170 the different processes.
4172 PH/38 Panic-die if .include specifies a non-absolute path.
4174 PH/39 A tweak to the "H" retry rule from its user.
4176 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4177 a label. They prevented compilation on older perls.
4179 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4180 a warning to be raised on newish perls.
4182 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4183 on queue. Changes to match documented behaviour of showing count of
4184 messages matching specified criteria.
4186 PH/40 Changed the default ident timeout from 30s to 5s.
4188 PH/41 Added support for the use of login_cap features, on those BSD systems
4189 that have them, for controlling the resources used by pipe deliveries.
4191 PH/42 The content-scanning code uses fopen() to create files in which to put
4192 message data. Previously it was not paying any attention to the mode of
4193 the files. Exim runs with umask(0) because the rest of the code creates
4194 files with open(), and sets the required mode explicitly. Thus, these
4195 files were ending up world-writeable. This was not a big issue, because,
4196 being within the spool directory, they were not world-accessible. I have
4197 created a function called modefopen, which takes an additional mode
4198 argument. It sets umask(777), creates the file, chmods it to the required
4199 mode, then resets the umask. All the relevant calls to fopen() in the
4200 content scanning code have been changed to use this function.
4202 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4203 to 24 hours. This avoids potential overflow problems when processing G
4204 and H retry rules. I suspect nobody ever tinkers with this value.
4206 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4208 PH/45 When the plaintext authenticator is running as a client, the server's
4209 challenges are checked to ensure they are valid base64 strings. By
4210 default, the authentication attempt is cancelled if an invalid string is
4211 received. Setting client_ignore_invalid_base64 true ignores these errors.
4212 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4213 they are received. Thus, the responses can be made to depend on the
4214 challenges. If an invalid string is ignored, an empty string is placed in
4217 PH/46 Messages that are created by the autoreply transport now contains a
4218 References: header, in accordance with RFCs 2822 and 3834.
4220 PH/47 Added authenticated_sender_force to the smtp transport.
4222 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4224 PH/49 Installed latest patch from the Sieve maintainer.
4226 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4227 was also set, the appendfile transport was unnecessarily scanning a
4228 directory of message files (e.g. for maildir delivery) to find the count
4229 of files (along with the size), even though it did not need this
4230 information. It now does the scan only if it needs to find either the
4231 size of the count of files.
4233 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4235 PH/52 Two bugs concerned with error handling when the smtp transport is
4238 (i) Exim was not creating retry information for temporary errors given
4239 for individual recipients after the DATA command when the smtp transport
4240 was used in LMTP mode. This meant that they could be retried too
4241 frequently, and not timed out correctly.
4243 (ii) Exim was setting the flag that allows error details to be returned
4244 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4245 recipients that were returned after the DATA command.
4247 PH/53 This is related to PH/52, but is more general: for any failing address,
4248 when detailed error information was permitted to be returned to the
4249 sender, but the error was temporary, then after the final timeout, only
4250 "retry timeout exceeded" was returned. Now it returns the full error as
4251 well as "retry timeout exceeded".
4253 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4254 do this, and (what is worse) MTAs that accept it.
4256 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4257 will now be deprecated.
4259 PH/56 New os.c-cygwin from the Cygwin maintainer.
4261 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4262 formats (previously only available in exim formats via -bpr, -bpru,
4263 and -bpra. Now also available in native and exiqgrep formats)
4265 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4266 with very large, slow to parse queues
4268 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4270 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4272 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4273 responses to authentication challenges, though it was showing the
4274 challenges; (ii) I've removed the CR characters from the debug output for
4277 PH/58 Allow for the insertion of a newline as well as a space when a string
4278 is turned into more than one encoded-word during RFC 2047 encoding. The
4279 Sieve code now uses this.
4281 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4282 data_4xx, lost_connection, tls_required.
4284 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4285 message was being sent as an SMTP response.
4287 PH/61 Add -l and -k options to exicyclog.
4289 PH/62 When verifying, if an address was redirected to one new address, so that
4290 verification continued, and the new address failed or deferred after
4291 having set something in $address_data, the value of $address_data was not
4292 passed back to the ACL. This was different to the case when no
4293 redirection occurred. The value is now passed back in both cases.
4295 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4296 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4297 use login_cap.h, so on its own it isn't the distinguishing feature. The
4298 new name refers directly to the setclassresources() function.
4300 PH/65 Added configuration files for NetBSD3.
4302 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4304 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4305 is preferred over IPv4.
4307 PH/68 The bounce_return_message and bounce_return_body options were not being
4308 honoured for bounces generated during the reception of non-SMTP messages.
4309 In particular, this applied to messages rejected by the ACL. This bug has
4310 been fixed. However, if bounce_return_message is true and bounce_return_
4311 body is false, the headers that are returned for a non-SMTP message
4312 include only those that have been read before the error was detected.
4313 (In the case of an ACL rejection, they have all been read.)
4315 PH/69 The HTML version of the specification is now built in a directory called
4316 spec_html instead of spec.html, because the latter looks like a path with
4317 a MIME-type, and this confuses some software.
4319 PH/70 Catch two compiler warnings in sieve.c.
4321 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4322 function verify_get_ident() calls ip_connect() to connect a socket, but
4323 if the "connect()" function timed out, ip_connect() used to close the
4324 socket. However, verify_get_ident() also closes the socket later, and in
4325 between Exim writes to the log, which may get opened at this point. When
4326 the socket was closed in ip_connect(), the log could get the same file
4327 descriptor number as the socket. This naturally causes chaos. The fix is
4328 not to close the socket in ip_connect(); the socket should be closed by
4329 the function that creates it. There was only one place in the code where
4330 this was missing, in the iplookup router, which I don't think anybody now
4331 uses, but I've fixed it anyway.
4333 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4334 well as to direct DNS lookups. Otherwise the handling of names in host
4335 lists is inconsistent and therefore confusing.
4341 PH/01 Two changes to the default runtime configuration:
4343 (1) Move the checks for relay_from_hosts and authenticated clients from
4344 after to before the (commented out) DNS black list checks.
4346 (2) Add control=submission to the relay_from_hosts and authenticated
4347 clients checks, on the grounds that messages accepted by these
4348 statements are most likely to be submissions.
4350 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4352 (1) Generate an error if the third argument for the ${prvs expansion is
4355 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4358 (3) Reset the variables that are obtained from the first argument of
4359 ${prvscheck and used in the second argument before leaving the code,
4360 because their memory is reclaimed, so using them afterwards may do
4363 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4364 one (it's much easier than Tom thought :-).
4366 (5) Because of (4), we can now allow for the use of $prvscheck_result
4367 inside the third argument.
4369 PH/03 For some reason, the default setting of PATH when running a command from
4370 a pipe transport was just "/usr/bin". I have changed it to
4373 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4374 anything to be listed in the output from -bV.
4376 PH/05 When a filter generated an autoreply, the entire To: header line was
4377 quoted in the delivery log line, like this:
4379 => >A.N.Other <ano@some.domain> <original@ddress> ...
4381 This has been changed so that it extracts the operative address. There
4382 may be more than one such address. If so, they are comma-separated, like
4385 => >ano@some.domain,ona@other.domain <original@ddress> ...
4387 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4388 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4389 not being looked up in the rDNS to get a host name, Exim was showing the
4390 IP address twice in Received: lines, even though the IP addresses were
4391 identical. For example:
4393 Received: from [1.2.3.4] (helo=[1.2.3.4])
4395 However, if the real host name was known, it was omitting the HELO data
4396 if it matched the actual IP address. This has been tidied up so that it
4397 doesn't show the same IP address twice.
4399 PH/07 When both +timestamp and +memory debugging was on, the value given by
4400 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4401 called by the debug printing, thereby overwriting the timestamp buffer.
4402 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4404 PH/08 When the original message was included in an autoreply transport, it
4405 always said "this is a copy of the message, including all the headers",
4406 even if body_only or headers_only was set. It now gives an appropriate
4409 PH/09 Applied a patch from the Sieve maintainer which:
4411 o fixes some comments
4412 o adds the (disabled) notify extension core
4413 o adds some debug output for the result of if/elsif tests
4414 o points to the current vacation draft in the documentation
4415 and documents the missing references header update
4419 o fixes a bug in processing the envelope test (when testing
4420 multiple envelope elements, the last element determined the
4423 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4424 Electronic Mail") by including:
4426 Auto-submitted: auto-generated
4428 in the messages that it generates (bounce messages and others, such as
4429 warnings). In the case of bounce messages for non-SMTP messages, there was
4430 also a typo: it was using "Auto_submitted" (underscore instead of
4431 hyphen). Since every message generated by Exim is necessarily in response
4432 to another message, thes have all been changed to:
4434 Auto-Submitted: auto-replied
4436 in accordance with these statements in the RFC:
4438 The auto-replied keyword:
4440 - SHOULD be used on messages sent in direct response to another
4441 message by an automatic process,
4443 - MUST NOT be used on manually-generated messages,
4445 - MAY be used on Delivery Status Notifications (DSNs) and Message
4446 Disposition Notifications (MDNs),
4448 - MUST NOT be used on messages generated by automatic or periodic
4449 processes, except for messages which are automatic responses to
4452 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4453 to the default Received: header definition.
4455 PH/12 Added log selector acl_warn_skipped (default on).
4457 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4458 variables because (a) they are in the wrong storage pool and (b) even if
4459 they were copied, it wouldn't work properly because of the caching.
4461 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4462 checking when decoding. Apparently there are clients that generate
4463 overlong encoded strings. Why am I not surprised?
4465 PH/15 If the first argument of "${if match_address" was not empty, but did not
4466 contain an "@" character, Exim crashed. Now it writes a panic log message
4467 and treats the condition as false.
4469 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4471 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4472 extension "envelope-auth". The code is finished and in agreement with
4473 other implementations, but there is no documentation so far and in fact,
4474 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4475 not changing the active code.
4477 Print executed "if" and "elsif" statements when debugging is used. This
4478 helps a great deal to understand what a filter does.
4480 Document more things not specified clearly in RFC3028. I had all this
4481 sorted out, when out of a sudden new issues came to my mind. Oops."
4483 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4486 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4487 canonical form (as documented). However, after a host name lookup from
4488 the IP address, check_host() was doing a simple string comparison with
4489 addresses acquired from the DNS when checking that the found name did
4490 have the original IP as one of its addresses. Since any found IPv6
4491 addresses are likely to be in abbreviated form, the comparison could
4492 fail. Luckily, there already exists a function for doing the comparison
4493 by converting both addresses to binary, so now that is used instead of
4494 the text comparison.
4496 PH/20 There was another similar case to PH/19, when a complete host name was
4497 given in a host list; looking up its IP address could give an abbreviated
4498 form, whereas the current host's name might or might not be abbreviated.
4499 The same fix has been applied.
4505 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4506 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4509 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4510 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4512 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4514 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4515 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4516 Also, for "Auto-Submitted", treat anything other than "no" as
4517 non-personal, in accordance with RFC 3834. (Previously it treated
4518 anything starting "auto-" as non-personal.)
4520 TF/01 The control=submission/name=... option had a problem with syntax
4521 errors if the name included a slash character. The /name= option
4522 now slurps the rest of the string, so it can include any characters
4523 but it must come last in the list of options (after /sender_retain
4526 PH/05 Some modifications to the interface to the fake nameserver for the new
4534 TK/01 Added the "success_on_redirect" address verification option. See
4535 NewStuff for rationale and an example.
4537 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4539 PH/02 Patch to exigrep to allow it to work on syslog lines.
4541 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4542 fread() to skip over the body file's header line, because in Cygwin the
4543 header line is locked and is inaccessible.
4545 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4546 co-exist for some time) to make it clear that it is the Exim ID that is
4547 referenced, not the Message-ID: header line.
4549 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4550 string_format() function, because snprintf() does not exist on all
4553 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4554 ${stat: expansion item.
4556 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4557 protocol synchronization error", to keep the pedants happy.
4559 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4560 well as for IRIX systems, when gcc is being used. See the host.c source
4563 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4565 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4568 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4569 transport and to the smtp transport in LMTP mode.
4571 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4573 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4574 run a filter in a subprocess. This could lead to confusion in subsequent
4575 lookups in the parent process. There should also be a search_tidyup() at
4576 the end of the subprocess.
4578 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4579 only if the host matched helo_try_verify_hosts, which caused the
4580 verification to occur when the EHLO/HELO command was issued. The ACL just
4581 tested the remembered result. Now, if a previous verification attempt has
4582 not happened, "verify = helo" does it there and then.
4584 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4586 TK/03 Fix log output including CR from clamd.
4588 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4589 exist provoked a memory error which could cause a segfault.
4591 PH/15 Installed PCRE 6.2
4593 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4595 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4596 of the problem. Specifically, suggested +O2 rather than +O1 for the
4599 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4601 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4602 record, the port was not correctly logged when the outgoing_port log
4603 selector was set (it logged the transort's default port).
4605 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4606 fallback_hosts, and "hosts" in the smtp transport.
4608 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4609 host errors such as "Connection refused".
4611 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4612 authentication with radiusclient 0.4.9:
4614 - Error returned from rc_read_config was caught wrongly
4615 - Username/password not passed on to radius server due to wrong length.
4617 The presumption is that some radiusclient API changes for 4.51/PH/17
4618 were not taken care of correctly. The code is still untested by me (my
4619 Linux distribution still has 0.3.2 of radiusclient), but it was
4620 contributed by a Radius user.
4622 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4623 expanding the "port" option of the smtp transport.
4625 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4626 while reading a MIME header. Thanks to Tom Hughes for a patch.
4628 PH/24 Include config.h inside local_scan.h so that configuration settings are
4631 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4632 that in an ACL for RCPT (for example), you can examine exactly what was
4635 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4636 commands, but it was not correctly comparing the address with the actual
4637 client host address. Thus, it would show the EHLO address in Received:
4638 header lines when this was not necessary.
4640 PH/27 Added the % operator to ${eval:}.
4642 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4643 it should be ignoring failures (because with -C, for example, it has lost
4644 privilege). It wasn't ignoring creation failures other than "already
4647 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4650 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4651 because an input line was too long, either on its own, or by virtue of
4652 too many continuations, the temporary file was not being removed, and the
4653 return code was incorrect.
4655 PH/31 Missing "BOOL" in function definition in filtertest.c.
4657 PH/32 Applied Sieve patches from the maintainer.
4659 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4661 PH/33 Added "verify = not_blind".
4663 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4664 Local/Makefile (with some defaults set). These are used in built scripts
4665 such as exicyclog, but they have never been used in the exim_install
4666 script (though there are many overriding facilities there). I have
4667 arranged that the exim_install script now takes note of these two
4670 PH/35 Installed configuration files for Dragonfly.
4672 PH/36 When a locally submitted message by a trusted user did not contain a
4673 From: header, and the sender address was obtained from -f or from an SMTP
4674 MAIL command, and the trusted user did not use -F to supply a sender
4675 name, $originator_name was incorrectly used when constructing a From:
4676 header. Furthermore, $originator_name was used for submission mode
4677 messages from external hosts without From: headers in a similar way,
4678 which is clearly wrong.
4680 PH/37 Added control=suppress_local_fixups.
4682 PH/38 When log_selector = +received_sender was set, and the addition of the
4683 sender made the log line's construction buffer exactly full, or one byte
4684 less than full, an overflow happened when the terminating "\n" was
4687 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4688 when the result of a list match is failure because a DNS lookup failed.
4690 PH/40 RM_COMMAND is now used in the building process.
4692 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4693 the "build-* directories that it finds.
4695 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4696 address in a domain literal was a prefix of an interface address.
4698 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4699 when verifying a sender address, unless rewrite_headers is false.
4701 PH/44 Wrote a long comment about why errors_to addresses are verified as
4702 recipients, not senders.
4704 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4705 the ratelimit ACL was added.
4707 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4709 PH/46 Added extra information about PostgreSQL errors to the error string.
4711 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4712 avoiding the need to install special zones in a real server. This is
4713 backwards compatible; if it can't find the fake resolver, it drops back.
4714 Thus, both old and new test suites can be run.
4716 TF/02 Added util/ratelimit.pl
4718 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4719 clock is set back in time.
4721 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4722 Candler <B.Candler@pobox.com>.
4724 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4725 OK for addresses that are the result of redirections.
4727 PH/48 A number of further additions for the benefit of the new test suite,
4728 including a fake gethostbyname() that interfaces to the fake DNS resolver
4731 TF/06 The fix for widen_domains has also been applied to qualify_single and
4732 search_parents which are the other dnslookup options that can cause
4735 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4738 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4740 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4741 service" if it encountered an MX record with an empty target hostname.
4742 The message is now "an MX or SRV record indicated no SMTP service".
4744 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4745 if the DNS of the sending site is misconfigured. This is quite a
4746 common situation. This change restores the behaviour of treating a
4747 helo verification defer as a failure.
4749 PH/51 If self=fail was set on a router, the bounce message did not include the
4750 actual error message.
4756 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4758 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4759 command itself ended in error, the underlying message about the transport
4760 filter timeout was being overwritten with the pipe command error. Now the
4761 underlying error message should be appended to the second error message.
4763 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4765 PH/02 Reduce the amount of output that "make" produces by default. Full output
4766 can still be requested.
4768 PH/03 The warning log line about a condition test deferring for a "warn" verb
4769 was being output only once per connection, rather than after each
4770 occurrence (because it was using the same function as for successful
4771 "warn" verbs). This seems wrong, so I have changed it.
4773 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4774 it should not have, which might have caused a crash in the right
4775 circumstances, but probably never did.
4777 PH/04 Installed a modified version of Tony Finch's patch to make submission
4778 mode fix the return path as well as the Sender: header line, and to
4779 add a /name= option so that you can make the user's friendly name appear
4782 TF/03 Added the control = fakedefer ACL modifier.
4784 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4785 Mark Lowes for thorough testing.
4787 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4789 TK/03 Merged latest SRS patch from Miles Wilton.
4791 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4792 with the definition in sysexits.h (which is #included earlier).
4793 Fortunately, Exim does not actually use EX_OK. The code used to try to
4794 preserve the sysexits.h value, by assuming that macro definitions were
4795 scanned for macro replacements. I have been disabused of this notion,
4796 so now the code just undefines EX_OK before #including unistd.h.
4798 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4799 in the smtp transport. When a block could not be written in a single
4800 write() function, the timeout was being re-applied to each part-write.
4801 This seems wrong - if the receiver was accepting one byte at a time it
4802 would take for ever. The timeout is now adjusted when this happens. It
4803 doesn't have to be particularly precise.
4805 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4806 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4808 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4809 without a domain if the check to <postmaster@domain> fails.
4811 SC/01 Eximstats: added -xls and the ability to specify output files
4812 (patch written by Frank Heydlauf).
4814 SC/02 Eximstats: use FileHandles for outputting results.
4816 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4818 SC/04 Eximstats: fixed display of large numbers with -nvr option
4820 SC/05 Eximstats: fixed merging of reports with empty tables.
4822 SC/06 Eximstats: added the -include_original_destination flag
4824 SC/07 Eximstats: removed tabs and trailing whitespace.
4826 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4828 TK/06 MBOX spool code: Add real "From " MBOX separator line
4829 so the .eml file is really in mbox format (even though
4830 most programs do not really care). Patch from Alex Miller.
4832 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4833 The latter is generated from $received_to and is only set if the
4834 message has one envelope recipient. SA can use these headers,
4835 obviously out-of-the-box. Patch from Alex Miller.
4837 PH/08 The ${def test on a variable was returning false if the variable's
4838 value was "0", contrary to what the specification has always said!
4839 The result should be true unless the variable is empty.
4841 PH/09 The syntax error of a character other than { following "${if
4842 def:variable_name" (after optional whitespace) was not being diagnosed.
4843 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4844 accidental colon was present, for example, could give incorrect results.
4846 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4847 result is used (not including appendfile, where other changes are about
4850 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4851 This involved changing a lot of size variables from int to off_t. It
4852 should work with maildirs and everything.
4854 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4855 spamd dying while we are connected to it.
4857 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4860 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4861 function for BDB 4.3.
4863 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4865 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4866 the sources for PCRE have been re-arranged and more files are now
4869 PH/15 The code I had for printing potentially long long variables in PH/11
4870 above was not the best (it lost precision). The length of off_t variables
4871 is now inspected at build time, and an appropriate printing format (%ld
4872 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4873 to be "long long int" or "long int". This is needed for the internal
4874 formatting function string_vformat().
4876 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4877 the configuration file to be ":syslog", then the script "guesses" where
4878 the logs files are, rather than using the compiled in default. In our
4879 case the guess is not the same as the compiled default, so the script
4880 suddenly stopped working when I started to use syslog. The patch checks
4881 to see if log_file_path is "". If so, it attempts to read it from exim
4882 with no configuration file to get the compiled in version, before it
4883 falls back to the previous guessing code."
4885 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4886 implementing BATV in an Exim configuration. See NewStuff for the gory
4889 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4890 Makefile that are specific to HP-UX.
4892 PH/18 If the "use_postmaster" option was set for a recipient callout together
4893 with the "random" option, the postmaster address was used as the MAIL
4894 FROM address for the random test, but not for the subsequent recipient
4895 test. It is now used for both.
4897 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4898 patch removes a few documentation additions to RFC 3028, because the
4899 latest draft now contains them. It adds the new en;ascii-case comparator
4900 and a new error check for 8bit text in MIME parts. Comparator and
4901 require names are now matched exactly. I enabled the subaddress
4902 extension, but it is not well tested yet (read: it works for me)."
4904 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4905 rework some of the code of TK/09 above to avoid the hardwired use of
4906 "%lld" and "long long". Replaced the call to snprintf() with a call to
4909 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4910 records point to non-existent hosts", "retry timeout exceeded", and
4911 "retry time not reached for any host after a long failure period".
4913 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4914 experimental DomainKeys support:
4916 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4917 (2) On an error such as an illegally used "control", the wrong name for
4918 the control was given.
4920 These problems did NOT occur unless DomainKeys support was compiled.
4922 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4924 PH/24 Added ${if match_ip condition.
4926 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4927 databases so that it will be absolutely obvious if a crash occurs in the
4928 DB library. This is a regular occurrence (often caused by mis-matched
4931 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4932 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4933 picked up on a user's system that detects such things. There doesn't seem
4934 to be a gcc warning option for this - only an attribute that has to be
4935 put on the function's prototype. It seems that in Fedora Core 4 they have
4936 set this on a number of new functions. No doubt there will be more in due
4939 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4940 specify a transport. However, if an address that was verified by such a
4941 router was the subject of a callout, Exim crashed because it tried to
4942 read the rcpt_include_affixes from the non-existent transport. Now it
4943 just assumes that the setting of that option is false. This bug was
4944 introduced by 4.51/PH/31.
4946 PH/28 Changed -d+all to exclude +memory, because that information is very
4947 rarely of interest, but it makes the output a lot bigger. People tend to
4948 do -d+all out of habit.
4950 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4951 code in os-type was giving problems when libc.so lives in lib64, like on
4954 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4955 aren't the modern standard, and it seems that some systems' include files
4956 don't always have them. Exim was already checking for some of the newer
4957 ones like T_AAAA, and defining it itself. I've added checks for all the
4958 record types that Exim uses.
4960 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4961 not automatically generating a new one, as it is supposed to. This
4962 prevented TLS from working. If the file did exist, but contained invalid
4963 data, a new version was generated, as expected. It was only the case of a
4964 non-existent file that was broken.
4966 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4967 with a change in libdomainkeys > 0.64.
4969 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4970 from DNS. If the selector record carries the flag, it now has
4971 precedence over the domain-wide flag.
4973 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4975 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4976 the use of a port name that isn't defined in /etc/services) occurred, the
4977 message was deferred as in a normal delivery, and thus remained on the
4978 spool, instead of being failed because of the mua_wrapper setting. This
4979 is now fixed, and I tidied up some of the mua_wrapper messages at the
4982 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4983 the messages in a hash of arrays rather than using individual hashes.
4984 This is a bit cleaner and results in dramatic memory savings, albeit
4985 at a slight CPU cost.
4987 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4988 as requested by Marc Sherman.
4990 SC/10 Eximstats: added histograms for user specified patterns as requested
4993 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4995 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4996 fopen() in the content-scanning modules that did not already have it.
5002 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5003 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5005 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5007 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5009 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5010 to test Sieve filters that use "vacation".
5012 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5013 that changes the way the GnuTLS parameters are stored in the cache file.
5014 The new format can be generated externally. For backward compatibility,
5015 if the data in the cache doesn't make sense, Exim assumes it has read an
5016 old-format file, and it generates new data and writes a new file. This
5017 means that you can't go back to an older release without removing the
5020 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5021 work if there are any delivery delays because "one_time" forces the
5022 parent to be marked "delivered", so its unseen clone is never tried
5023 again. For this reason, Exim now forbids the simultaneous setting of
5026 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5027 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5028 entry if you want to know the details. The fix, however, neglected to
5029 consider the case when local delivery batching is involved. The test for
5030 "previously delivered" was not happening when checking to see if an
5031 address could be batched with a previous (undelivered) one; under
5032 certain circumstances this could lead to multiple deliveries to the same
5035 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5036 in its include files, and this causes problems building Exim.
5038 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5039 header_syntax) but Exim was just ignoring anything given after a slash.
5040 In particular, this caused confusion with an attempt to use "verify =
5041 reverse_host_lookup/defer_ok". An error is now given when options are
5042 supplied for verify items that do not have them. (Maybe reverse_host_
5043 lookup should have a defer_ok option, but that's a different point.)
5045 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5046 defined by RFC 821) to 2048, because there were problems with some AUTH
5047 commands, and RFC 1869 says the size should be increased for extended
5048 SMTP commands that take arguments.
5050 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5053 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5054 "unknown" error; now it says that the functionality isn't in the binary.
5056 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5057 an address' error message when a string expansion fails (syntax or
5058 whatever). Otherwise the password may appear in the log. Following change
5059 PH/42 below, there is no longer a chance of it appearing in a bounce
5062 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5064 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5065 of its addresses was ever tried. (Bugzilla bug #2.)
5067 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5068 the result incorrectly in the debug output. (It correctly added a newline
5069 to what was transported.)
5071 TF/01 Added $received_time.
5073 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5074 commented out examples of how to interface to a virus scanner and to
5075 SpamAssassin. Also added commented examples of av_scanner and
5076 spamd_address settings.
5078 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5079 and controls are allowed in which ACLs. There were a couple of minor
5080 errors. Some of the entries in the conditions table (which is a table of
5081 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5082 negation of where the condition IS allowed.
5084 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5086 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5087 header file does not have a version number, so I've had to invent a new
5088 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5089 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5090 radiusclient), but it was contributed by a Radius user.
5092 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5093 files or directories, for OpenSSL.
5095 PH/19 When an Exim process that is running as root has to create an Exim log
5096 file, it does so in a subprocess that runs as exim:exim so as to get the
5097 ownership right at creation (otherwise, other Exim processes might see
5098 the file with the wrong ownership). There was no test for failure of this
5099 fork() call, which would lead to the process getting stuck as it waited
5100 for a non-existent subprocess. Forks do occasionally fail when resources
5101 run out. I reviewed all the other calls to fork(); they all seem to check
5104 PH/20 When checking for unexpected SMTP input at connect time (before writing
5105 the banner), Exim was not dealing correctly with a non-positive return
5106 from the read() function. If the client had disconnected by this time,
5107 the result was a log entry for a synchronization error with an empty
5108 string after "input=" when read() returned zero. If read() returned -1
5109 (an event I could not check), uninitialized data bytes were printed.
5110 There were reports of junk text (parts of files, etc) appearing after
5113 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5115 PH/22 Added support for macro redefinition, and (re)definition in between
5116 driver and ACL definitions.
5118 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5119 forgetting to use the resulting value; it was using the unexpanded value.
5121 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5122 hadn't been configured. The fix is from Juergen Kreileder, who
5123 understands it better than I do:
5125 "Here's what I see happening with three configured cyrus_sasl
5126 authenticators configured (plain, login, cram-md5):
5128 On startup auth_cyrus_sasl_init() gets called for each of these.
5129 This means three calls to sasl_listmech() without a specified mech_list.
5130 => SASL tests which mechs of all available mechs actually work
5131 => three warnings about OTP not working
5132 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5134 With the patch, sasl_listmech() also gets called three times. But now
5135 SASL's mech_list option is set to the server_mech specified in the the
5136 authenticator. Or in other words, the answer from sasl_listmech()
5137 gets limited to just the mech you're testing for (which is different
5139 => the return list contains just 'plain' or 'login', 'cram-md5' or
5140 nothing depending on the value of ob->server_mech.
5142 I've just tested the patch: Authentication still works fine,
5143 unavailable mechs specified in the exim configuration are still
5144 caught, and the auth.log warnings about OTP are gone."
5146 PH/25 When debugging is enabled, the contents of the command line are added
5147 to the debugging output, even when log_selector=+arguments is not
5150 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5151 answer is "GNU", and only if the return is "GNU/something" is the answer
5154 PH/27 $acl_verify_message is now set immediately after the failure of a
5155 verification in an ACL, and so is available in subsequent modifiers. In
5156 particular, the message can be preserved by coding like this:
5158 warn !verify = sender
5159 set acl_m0 = $acl_verify_message
5161 Previously, $acl_verify_message was set only while expanding "message"
5162 and "log_message" when a very denied access.
5164 PH/28 Modified OS/os.c-Linux with
5166 -#ifndef OS_LOAD_AVERAGE
5167 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5169 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5170 nomenclature these days.)
5172 PH/29 Installed patch from the Sieve maintainer that adds the options
5173 sieve_useraddress and sieve_subaddress to the redirect router.
5175 PH/30 In these circumstances:
5176 . Two addresses routed to the same list of hosts;
5177 . First host does not offer TLS;
5178 . First host accepts first address;
5179 . First host gives temporary error to second address;
5180 . Second host offers TLS and a TLS session is established;
5181 . Second host accepts second address.
5182 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5183 and peerdn, if requested) that were in fact used only for the second
5186 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5187 attention to any local part prefix or suffix that was matched by the
5188 router that accepted the address. It now behaves in the same way as it
5189 does for delivery: the affixes are removed from the local part unless
5190 rcpt_include_affixes is set on the transport.
5192 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5193 timeout during the DATA phase of an incoming message.
5195 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5196 applied a patch sanctioned by the Sieve maintainer.
5198 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5199 the uid or gid is negative. A case of a negative gid caused this to be
5200 noticed. The fix allows for either to be negative.
5202 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5203 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5206 PH/36 The change PH/12 above was broken. Fixed it.
5208 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5209 the grounds that routing the same address twice would always produce the
5210 same answer. This might have been true once, but it is certainly no
5211 longer true now. Routing a child address may depend on the previous
5212 routing that produced that child. Some complicated redirection strategies
5213 went wrong when messages had multiple recipients, and made Exim's
5214 behaviour dependent on the order in which the addresses were given.
5216 I have moved the duplicate checking until after the routing is complete.
5217 Exim scans the addresses that are assigned to local and remote
5218 transports, and removes any duplicates. This means that more work will be
5219 done, as duplicates will always all be routed, but duplicates are
5220 presumably rare, so I don't expect this is of any significance.
5222 For deliveries to pipes, files, and autoreplies, the duplicate checking
5223 still happens during the routing process, since they are not going to be
5226 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5227 It corrects a timeout issue with spamd. This is Ian's comment: "The
5228 background is that sometimes spamd either never reads data from a
5229 connection it has accepted, or it never writes response data. The exiscan
5230 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5231 blindly assumes that writes won't block so it may never time out."
5233 PH/39 Allow G after quota size as well as K and M.
5235 PH/40 The value set for $authenticated_id in an authenticator may not contain
5236 binary zeroes or newlines because the value is written to log lines and
5237 to spool files. There was no check on this. Now the value is run through
5238 the string_printing() function so that such characters are converted to
5239 printable escape sequences.
5241 PH/41 $message_linecount is a new variable that contains the total number of
5242 lines in the message. Compare $body_linecount, which is the count for the
5245 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5246 bounce and delay warning messages, except in certain special cases, which
5249 (a) An SMTP error message from a remote host;
5250 (b) A message specified in a :fail: redirection;
5251 (c) A message specified in a "fail" command in a system filter;
5252 (d) A message specified in a FAIL return from the queryprogram router;
5253 (e) A message specified by the cannot_route_message router option.
5255 In these cases only, Exim does include the error details in bounce and
5256 warning messages. There are also a few cases where bland messages such
5257 as "unrouteable address" or "local delivery error" are given.
5259 PH/43 $value is now also set for the "else" part of a ${run expansion.
5261 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5262 being worked on, but at least Exim now implements the latest version to
5265 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5266 process to complete was treated as a delivery failure, a timeout while
5267 writing the message to the pipe was logged, but erroneously treated as a
5268 successful delivery. Such timeouts include transport filter timeouts. For
5269 consistency with the overall process timeout, these timeouts are now
5270 treated as errors, giving rise to delivery failures by default. However,
5271 there is now a new Boolean option for the pipe transport called
5272 timeout_defer, which, if set TRUE, converts the failures into defers for
5273 both kinds of timeout. A transport filter timeout is now identified in
5276 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5277 systems where "make" and "gmake" are different, calling "gmake" at top
5278 level broke things. I've arranged for the value of $(MAKE) to be passed
5279 from the Makefile to this script so that it can call the same version of
5283 A note about Exim versions 4.44 and 4.50
5284 ----------------------------------------
5286 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5287 changes of various kinds. As a consequence, a big documentation update was
5288 needed. This delayed the release for rather longer than seemed good, especially
5289 in the light of a couple of (minor) security issues. Therefore, the changes
5290 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5291 release. So 4.44 and 4.50 are in effect two different branches that both start
5294 I have left the 4.50 change log unchanged; it contains all the changes since
5295 4.43. The change log for 4.44 is below; many of its items are identical to
5296 those for 4.50. This seems to be the most sensible way to preserve the
5297 historical information.
5303 1. Minor wording change to the doc/README.SIEVE file.
5305 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5306 computation of the current number of files was incorrect.
5308 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5309 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5312 4. Give more explanation in the error message when the command for a transport
5313 filter fails to execute.
5315 5. There are several places where Exim runs a non-Exim command in a
5316 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5317 was being done only for the command run by the queryprogram router. It is
5318 now done for all such subprocesses. The other cases are: ${run, transport
5319 filters, and the commands run by the lmtp and pipe transports.
5321 6. Added CONFIGURE_GROUP build-time option.
5323 7. Some older OS have a limit of 256 on the maximum number of file
5324 descriptors. Exim was using setrlimit() to set 1000 as a large value
5325 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5326 systems. I've change it so that if it can't get 1000, it tries for 256.
5328 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5329 was an oversight, and furthermore, ever since the addition of extra
5330 controls (e.g. 4.43/32), the checks on when to allow different forms of
5331 "control" were broken. There should now be diagnostics for all cases when a
5332 control that does not make sense is encountered.
5334 9. Added the /retain_sender option to "control=submission".
5336 10. $recipients is now available in the predata ACL (oversight).
5338 11. Tidy the search cache before the fork to do a delivery from a message
5339 received from the command line. Otherwise the child will trigger a lookup
5340 failure and thereby defer the delivery if it tries to use (for example) a
5341 cached ldap connection that the parent has called unbind on.
5343 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5344 of $address_data from the recipient verification was clobbered by the
5345 sender verification.
5347 13. The value of address_data from a sender verification is now available in
5348 $sender_address_data in subsequent conditions in the ACL statement.
5350 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5352 15. Added a new option "connect=<time>" to callout options, to set a different
5355 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5356 was its contents. (It was OK if the option was not defined at all.)
5358 17. A "Completed" log line is now written for messages that are removed from
5359 the spool by the -Mrm option.
5361 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5362 information about exactly what failed.
5364 19. Added -dd to debug only the daemon process.
5366 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5367 handles timeouts, both on the server side and network timeouts. Renamed the
5368 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5370 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5371 It is now set to "smtps".
5373 22. $host_address is now set to the target address during the checking of
5374 ignore_target_hosts.
5376 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5377 being passed; this would have caused $sender_host_name to have been used if
5378 matching the list had actually called for a host name (not very likely,
5379 since this list is usually IP addresses). A host name is now passed as
5382 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5383 code that specifies a non-restarting handler (typically sigaction() in
5384 modern systems) in an attempt to fix a rare and obscure crash bug.
5386 25. Narrowed the window for a race in the daemon that could cause it to ignore
5387 SIGCHLD signals. This is not a major problem, because they are used only to
5388 wake it up if nothing else does.
5390 26. A malformed maildirsize file could cause Exim to calculate negative values
5391 for the mailbox size or file count. Odd effects could occur as a result.
5392 The maildirsize information is now recalculated if the size or filecount
5395 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5396 support for a long time. Removed HAVE_SYS_VFS_H.
5398 28. Installed the latest version of exipick from John Jetmore.
5400 29. In an address list, if the pattern was not a regular expression, an empty
5401 subject address (from a bounce message) matched only if the pattern was an
5402 empty string. Non-empty patterns were not even tested. This was the wrong
5403 because it is perfectly reasonable to use an empty address as part of a
5404 database query. An empty address is now tested by patterns that are
5405 lookups. However, all the other forms of pattern expect the subject to
5406 contain a local part and a domain, and therefore, for them, an empty
5407 address still always fails if the pattern is not itself empty.
5409 30. Exim went into a mad DNS loop when attempting to do a callout where the
5410 host was specified on an smtp transport, and looking it up yielded more
5411 than one IP address.
5413 31. Re-factored the code for checking spool and log partition space into a
5414 function that finds that data and another that does the check. The former
5415 is then used to implement four new variables: $spool_space, $log_space,
5416 $spool_inodes, and $log_inodes.
5418 32. The RFC2047 encoding function was originally intended for short strings
5419 such as real names; it was not keeping to the 75-character limit for
5420 encoded words that the RFC imposes. It now respects the limit, and
5421 generates multiple encoded words if necessary. To be on the safe side, I
5422 have increased the buffer size for the ${rfc2047: expansion operator from
5425 33. It is now permitted to omit both strings after an "if" condition; if the
5426 condition is true, the result is "true". As before, when the second string
5427 is omitted, a false condition yields an empty string. This makes it less
5428 cumbersome to write custom ACL and router conditions.
5430 34. Failure to deliver a bounce message always caused it to be frozen, even if
5431 there was an errors_to setting on the router. The errors_to setting is now
5434 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5435 canonical form (fully expanded) before being placed in
5436 $sender_host_address.
5438 36. The table in the code that translates DNS record types into text (T_A to
5439 "A" for instance) was missing entries for NS and CNAME. It is just possible
5440 that this could have caused confusion if both these types were looked up
5441 for the same domain, because the text type is used as part of Exim's
5442 per-process caching. But the chance of anyone hitting this buglet seems
5445 37. The dnsdb lookup has been extended in a number of ways.
5447 (1) There is a new type, "zns", which walks up the domain tree until it
5448 finds some nameserver records. It should be used with care.
5450 (2) There is a new type, "mxh", which is like "mx" except that it returns
5451 just the host names, not the priorities.
5453 (3) It is now possible to give a list of domains (or IP addresses) to be
5454 looked up. The behaviour when one of the lookups defers can be
5455 controlled by a keyword.
5457 (4) It is now possible to specify the separator character for use when
5458 multiple records are returned.
5460 38. The dnslists ACL condition has been extended: it is now possible to supply
5461 a list of IP addresses and/or domains to be looked up in a particular DNS
5464 39. Added log_selector=+queue_time_overall.
5466 40. When running the queue in the test harness, wait just a tad after forking a
5467 delivery process, to get repeatability of debugging output.
5469 41. Include certificate and key file names in error message when GnuTLS fails
5470 to set them up, because the GnuTLS error message doesn't include the name
5471 of the failing file when there is a problem reading it.
5473 42. Allow both -bf and -bF in the same test run.
5475 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5477 44. The "Exiscan patch" is now merged into the mainline Exim source.
5479 45. Sometimes the final signoff response after QUIT could fail to get
5480 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5481 < 0 before doing a fflush(). This bug looks as though it goes back to the
5482 introduction of TLS in release 3.20, but "sometimes" must have been rare
5483 because the tests only now provoked it.
5485 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5486 (this can affect the format of dates).
5488 47. exim_tidydb, when checking for the continued existence of a message for
5489 which it has found a message-specific retry record, was not finding
5490 messages that were in split spool directories. Consequently, it was
5491 deleting retry records that should have stayed in existence.
5493 48. Steve fixed some bugs in eximstats.
5495 49. The SPA authentication driver was not abandoning authentication and moving
5496 on to the next authenticator when an expansion was forced to fail,
5497 contradicting the general specification for all authenticators. Instead it
5498 was generating a temporary error. It now behaves as specified.
5500 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5501 (the order specifies the preference for clients). The order is now AES256,
5502 AES128, 3DES, ARCFOUR128.
5504 51. Small patch to Sieve code - explicitly set From: when generating an
5507 52. Exim crashed if a remote delivery caused a very long error message to be
5508 recorded - for instance if somebody sent an entire SpamAssassin report back
5509 as a large number of 550 error lines. This bug was coincidentally fixed by
5510 increasing the size of one of Exim's internal buffers (big_buffer) that
5511 happened as part of the Exiscan merge. However, to be on the safe side, I
5512 have made the code more robust (and fixed the comments that describe what
5515 53. Now that there can be additional text after "Completed" in log lines (if
5516 the queue_time_overall log selector is set), a one-byte patch to exigrep
5517 was needed to allow it to recognize "Completed" as not the last thing in
5520 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5521 patch that reportedly fixes this has been added. I am not expert enough to
5522 create a test for it. This is what the patch creator wrote:
5524 "I found a little strange behaviour of ldap code when working with
5525 Windows 2003 AD Domain, where users was placed in more than one
5526 Organization Units. When I tried to give exim partial DN, the exit code
5527 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5528 But simultaneously result of request was absolutely normal ldap result,
5529 so I produce this patch..."
5531 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5532 REFERENCE, so I have modified the code to exclude the patch when that macro
5535 55. Some experimental protocols are using DNS PTR records for new purposes. The
5536 keys for these records are domain names, not reversed IP addresses. The
5537 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5538 leaves it alone. Component reversal etc. now happens only for IP addresses.
5541 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5543 57. Double the size of the debug message buffer (to 2048) so that more of very
5544 long debug lines gets shown.
5546 58. The exicyclog utility now does better if the number of log files to keep
5547 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5549 59. Two changes related to the smtp_active_hostname option:
5551 (1) $smtp_active_hostname is now available as a variable.
5552 (2) The default for smtp_banner uses $smtp_active_hostname instead
5553 of $primary_hostname.
5555 60. The host_aton() function is supposed to be passed a string that is known
5556 to be a valid IP address. However, in the case of IPv6 addresses, it was
5557 not checking this. This is a hostage to fortune. Exim now panics and dies
5558 if the condition is not met. A case was found where this could be provoked
5559 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5560 components; fortuitously, this particular loophole had already been fixed
5561 by change 4.50/55 above.
5563 If there are any other similar loopholes, the new check in host_aton()
5564 itself should stop them being exploited. The report I received stated that
5565 data on the command line could provoke the exploit when Exim was running as
5566 exim, but did not say which command line option was involved. All I could
5567 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5568 running as the user.
5571 61. There was a buffer overflow vulnerability in the SPA authentication code
5572 (which came originally from the Samba project). I have added a test to the
5573 spa_base64_to_bits() function which I hope fixes it.
5576 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5577 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5579 63. The daemon start-up calls getloadavg() while still root for those OS that
5580 need the first call to be done as root, but it missed one case: when
5581 deliver_queue_load_max is set with deliver_drop_privilege. This is
5582 necessary for the benefit of the queue runner, because there is no re-exec
5583 when deliver_drop_privilege is set.
5585 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5586 This has been fixed.
5588 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5589 was in use, was not putting the data itself into the right store pool;
5590 consequently, it could be overwritten for a subsequent message in the same
5591 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5594 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5596 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5597 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5598 the function were treating the return as a boolean value, which happened to
5599 work because 0=false and not-0=true, but is not correct code.
5601 68. The host_aton() function was not handling scoped IPv6 addresses (those
5602 with, for example, "%eth0" on the end) correctly.
5604 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5605 negated items (that is, ~something) in unsigned ints. Some compilers
5606 apparently mutter when there is no cast.
5608 70. If an address verification called from an ACL failed, and did not produce a
5609 user-specific message (i.e. there was only a "system" message), nothing was
5610 put in $acl_verify_message. In this situation, it now puts the system
5613 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5614 session; change 4.31/43 added the unwanted input to the log line - except
5615 that it did not do this in the start of session case. It now does.
5617 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5618 This is wrong and can cause the other end to generate a synchronization
5619 error if it is another Exim or anything else that does the synchronization
5620 check. A QUIT command is no longer sent after a timeout.
5622 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5623 during host lookups.
5625 74. The defer_ok option of callout verification was not working if it was used
5626 when verifying addresses in header lines, that is, for this case:
5628 verify = header_sender/callout=defer_ok
5630 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5631 those file descriptors could be used for SMTP connections. If anything
5632 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5633 could be sent to the SMTP client, causing chaos. The daemon now opens
5634 stdin, stdout, and stderr to /dev/null when it puts itself into the
5637 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5638 log by default. The user can override this with suitable Perl magic.
5640 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5641 the log message when discard triggers, was not working for the DATA ACL or
5642 for the non-SMTP ACL.
5644 78. Error message wording change in sieve.c.
5646 79. If smtp_accept_max_per_host was set, the number of connections could be
5647 restricted to fewer than expected, because the daemon was trying to set up
5648 a new connection before checking whether the processes handling previous
5649 connections had finished. The check for completed processes is now done
5650 earlier. On busy systems, this bug wouldn't be noticed because something
5651 else would have woken the daemon, and it would have reaped the completed
5654 80. If a message was submitted locally by a user whose login name contained one
5655 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5656 It caused a spool format error. I have fixed the spool reading code. A
5657 related problem was that the "from" clause in the Received: line became
5658 illegal because of the space(s). It is now covered by ${quote_local_part.
5660 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5663 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5665 83. Patch from Sieve maintainer to fix unterminated string problem in
5666 "vacation" handling.
5668 84. Some minor changes to the Linux configuration files to help with other
5669 OS variants using glibc.
5671 85. One more patch for Sieve to update vacation handling to latest spec.
5674 ----------------------------------------------------
5675 See the note above about the 4.44 and 4.50 releases.
5676 ----------------------------------------------------
5682 1. Change 4.43/35 introduced a bug that caused file counts to be
5683 incorrectly computed when quota_filecount was set in an appendfile
5686 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5687 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5690 3. Give more explanation in the error message when the command for a transport
5691 filter fails to execute.
5693 4. There are several places where Exim runs a non-Exim command in a
5694 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5695 was being done only for the command run by the queryprogram router. It is
5696 now done for all such subprocesses. The other cases are: ${run, transport
5697 filters, and the commands run by the lmtp and pipe transports.
5699 5. Some older OS have a limit of 256 on the maximum number of file
5700 descriptors. Exim was using setrlimit() to set 1000 as a large value
5701 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5702 systems. I've change it so that if it can't get 1000, it tries for 256.
5704 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5705 was an oversight, and furthermore, ever since the addition of extra
5706 controls (e.g. 4.43/32), the checks on when to allow different forms of
5707 "control" were broken. There should now be diagnostics for all cases when a
5708 control that does not make sense is encountered.
5710 7. $recipients is now available in the predata ACL (oversight).
5712 8. Tidy the search cache before the fork to do a delivery from a message
5713 received from the command line. Otherwise the child will trigger a lookup
5714 failure and thereby defer the delivery if it tries to use (for example) a
5715 cached ldap connection that the parent has called unbind on.
5717 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5718 of $address_data from the recipient verification was clobbered by the
5719 sender verification.
5721 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5722 was its contents. (It was OK if the option was not defined at all.)
5724 11. A "Completed" log line is now written for messages that are removed from
5725 the spool by the -Mrm option.
5727 12. $host_address is now set to the target address during the checking of
5728 ignore_target_hosts.
5730 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5731 being passed; this would have caused $sender_host_name to have been used if
5732 matching the list had actually called for a host name (not very likely,
5733 since this list is usually IP addresses). A host name is now passed as
5736 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5737 code that specifies a non-restarting handler (typically sigaction() in
5738 modern systems) in an attempt to fix a rare and obscure crash bug.
5740 15. Narrowed the window for a race in the daemon that could cause it to ignore
5741 SIGCHLD signals. This is not a major problem, because they are used only to
5742 wake it up if nothing else does.
5744 16. A malformed maildirsize file could cause Exim to calculate negative values
5745 for the mailbox size or file count. Odd effects could occur as a result.
5746 The maildirsize information is now recalculated if the size or filecount
5749 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5750 support for a long time. Removed HAVE_SYS_VFS_H.
5752 18. Updated exipick to current release from John Jetmore.
5754 19. Allow an empty sender to be matched against a lookup in an address list.
5755 Previously the only cases considered were a regular expression, or an
5758 20. Exim went into a mad DNS lookup loop when doing a callout where the
5759 host was specified on the transport, if the DNS lookup yielded more than
5762 21. The RFC2047 encoding function was originally intended for short strings
5763 such as real names; it was not keeping to the 75-character limit for
5764 encoded words that the RFC imposes. It now respects the limit, and
5765 generates multiple encoded words if necessary. To be on the safe side, I
5766 have increased the buffer size for the ${rfc2047: expansion operator from
5769 22. Failure to deliver a bounce message always caused it to be frozen, even if
5770 there was an errors_to setting on the router. The errors_to setting is now
5773 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5774 canonical form (fully expanded) before being placed in
5775 $sender_host_address.
5777 24. Updated eximstats to version 1.33
5779 25. Include certificate and key file names in error message when GnuTLS fails
5780 to set them up, because the GnuTLS error message doesn't include the name
5781 of the failing file when there is a problem reading it.
5783 26. Expand error message when OpenSSL has problems setting up cert/key files.
5786 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5787 (this can affect the format of dates).
5789 28. exim_tidydb, when checking for the continued existence of a message for
5790 which it has found a message-specific retry record, was not finding
5791 messages that were in split spool directories. Consequently, it was
5792 deleting retry records that should have stayed in existence.
5794 29. eximstats updated to version 1.35
5795 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5796 1.35 - bugfix such that pie charts by volume are generated correctly
5798 30. The SPA authentication driver was not abandoning authentication and moving
5799 on to the next authenticator when an expansion was forced to fail,
5800 contradicting the general specification for all authenticators. Instead it
5801 was generating a temporary error. It now behaves as specified.
5803 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5804 (the order specifies the preference for clients). The order is now AES256,
5805 AES128, 3DES, ARCFOUR128.
5807 31. Small patch to Sieve code - explicitly set From: when generating an
5810 32. Exim crashed if a remote delivery caused a very long error message to be
5811 recorded - for instance if somebody sent an entire SpamAssassin report back
5812 as a large number of 550 error lines. This bug was coincidentally fixed by
5813 increasing the size of one of Exim's internal buffers (big_buffer) that
5814 happened as part of the Exiscan merge. However, to be on the safe side, I
5815 have made the code more robust (and fixed the comments that describe what
5818 33. Some experimental protocols are using DNS PTR records for new purposes. The
5819 keys for these records are domain names, not reversed IP addresses. The
5820 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5821 leaves it alone. Component reversal etc. now happens only for IP addresses.
5824 34. The host_aton() function is supposed to be passed a string that is known
5825 to be a valid IP address. However, in the case of IPv6 addresses, it was
5826 not checking this. This is a hostage to fortune. Exim now panics and dies
5827 if the condition is not met. A case was found where this could be provoked
5828 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5829 components; fortuitously, this particular loophole had already been fixed
5830 by change 4.50/55 or 4.44/33 above.
5832 If there are any other similar loopholes, the new check in host_aton()
5833 itself should stop them being exploited. The report I received stated that
5834 data on the command line could provoke the exploit when Exim was running as
5835 exim, but did not say which command line option was involved. All I could
5836 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5837 running as the user.
5840 35. There was a buffer overflow vulnerability in the SPA authentication code
5841 (which came originally from the Samba project). I have added a test to the
5842 spa_base64_to_bits() function which I hope fixes it.
5845 36. The daemon start-up calls getloadavg() while still root for those OS that
5846 need the first call to be done as root, but it missed one case: when
5847 deliver_queue_load_max is set with deliver_drop_privilege. This is
5848 necessary for the benefit of the queue runner, because there is no re-exec
5849 when deliver_drop_privilege is set.
5851 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5852 was in use, was not putting the data itself into the right store pool;
5853 consequently, it could be overwritten for a subsequent message in the same
5854 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5857 38. Sometimes the final signoff response after QUIT could fail to get
5858 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5859 < 0 before doing a fflush(). This bug looks as though it goes back to the
5860 introduction of TLS in release 3.20, but "sometimes" must have been rare
5861 because the tests only now provoked it.
5867 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5868 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5869 Now it returns an integer. A number of calls were still expecting a T/F
5870 return. Fortuitously, in all cases, the tests worked in OK situations,
5871 which is the norm. However, things would have gone wrong on any write
5872 failures on the smtp file descriptor. This function is used when sending
5873 messages over SMTP and also when doing verify callouts.
5875 2. When Exim is called to do synchronous delivery of a locally submitted
5876 message (the -odf or -odi options), it no longer closes stderr before doing
5879 3. Implemented the mua_wrapper option.
5881 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5883 5. Implemented the functions header_remove(), header_testname(),
5884 header_add_at_position(), and receive_remove_recipient(), and exported them
5887 6. If an ACL "warn" statement specified the addition of headers, Exim already
5888 inserted X-ACL-Warn: at the start if there was no header name. However, it
5889 was not making this test for the second and subsequent header lines if
5890 there were newlines in the string. This meant that an invalid header could
5891 be inserted if Exim was badly configured.
5893 7. Allow an ACL "warn" statement to add header lines at the start or after all
5894 the Received: headers, as well as at the end.
5896 8. Added the rcpt_4xx retry error code.
5898 9. Added postmaster_mailfrom=xxx to callout verification option.
5900 10. Added mailfrom=xxxx to the callout verification option, for verify=
5903 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5904 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5906 12. Inserted some casts to stop certain compilers warning when using pointer
5907 differences as field lengths or precisions in printf-type calls (mostly
5908 affecting debugging statements).
5910 13. Added optional readline() support for -be (dynamically loaded).
5912 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5913 same clock tick as a message's arrival, so that its received time was the
5914 same as the "first fail" time on the retry record, and that message
5915 remained on the queue past the ultimate address timeout, every queue runner
5916 would try a delivery (because it was past the ultimate address timeout) but
5917 after another failure, the ultimate address timeout, which should have then
5918 bounced the address, did not kick in. This was a "< instead of <=" error;
5919 in most cases the first failure would have been in the next clock tick
5920 after the received time, and all would be well.
5922 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5923 being recognized when the domain list was tested by the match_domain
5924 condition in an expansion string.
5926 16. Added the ${str2b64: operator.
5928 17. Exim was always calling setrlimit() to set a large limit for the number of
5929 processes, without checking whether the existing limit was already
5930 adequate. (It did check for the limit on file descriptors.) Furthermore,
5931 errors from getrlimit() and setrlimit() were being ignored. Now they are
5932 logged to the main and panic logs, but Exim does carry on, to try to do its
5933 job under whatever limits there are.
5935 18. Imported PCRE 5.0.
5937 19. Trivial typo in log message " temporarily refused connection" (the leading
5940 20. If the log selector return_path_on_delivery was set and an address was
5941 redirected to /dev/null, the delivery process crashed because it assumed
5942 that a return path would always be set for a "successful" delivery. In this
5943 case, the whole delivery is bypassed as an optimization, and therefore no
5946 21. Internal re-arrangement: the function for sending a challenge and reading
5947 a response while authentication was assuming a zero-terminated challenge
5948 string. It's now changed to take a pointer and a length, to allow for
5949 binary data in such strings.
5951 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5953 23. Exim was not respecting finduser_retries when seeking the login of the
5954 uid under which it was called; it was always trying 10 times. (The default
5955 setting of finduser_retries is zero.) Also, it was sleeping after the final
5956 failure, which is pointless.
5958 24. Implemented tls_on_connect_ports.
5960 25. Implemented acl_smtp_predata.
5962 26. If the domain in control=submission is set empty, Exim assumes that the
5963 authenticated id is a complete email address when it generates From: or
5964 Sender: header lines.
5966 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5967 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5968 chown and chgrp in /bin and hostname in /usr/bin.
5970 28. Exim was keeping the "process log" file open after each use, just as it
5971 does for the main log. This opens the possibility of it remaining open for
5972 long periods when the USR1 signal hits a daemon. Occasional processlog
5973 errors were reported, that could have been caused by this. Anyway, it seems
5974 much more sensible not to leave this file open at all, so that is what now
5977 29. The long-running daemon process does not normally write to the log once it
5978 has entered its main loop, and it closes the log before doing so. This is
5979 so that log files can straightforwardly be renamed and moved. However,
5980 there are a couple of unusual error situations where the daemon does write
5981 log entries, and I had neglected to close the log afterwards.
5983 30. The text of an SMTP error response that was received during a remote
5984 delivery was being truncated at 512 bytes. This is too short for some of
5985 the long messages that one sometimes sees. I've increased the limit to
5988 31. It is now possible to make retry rules that apply only when a message has a
5989 specific sender, in particular, an empty sender.
5991 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5992 it possible to be selective about when SMTP synchronization is enforced.
5994 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5996 32. Implemented hosts_connection_nolog.
5998 33. Added an ACL for QUIT.
6000 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6003 35. Added mailbox_size and mailbox_filecount to appendfile.
6005 36. Added control = no_multiline_responses to ACLs.
6007 37. There was a bug in the logic of the code that waits for the clock to tick
6008 in the case where the clock went backwards by a substantial amount such
6009 that the microsecond fraction of "now" was more than the microsecond
6010 fraction of "then" (but the whole seconds number was less).
6012 38. Added support for the libradius Radius client library this is found on
6013 FreeBSD (previously only the radiusclient library was supported).
6019 1. When certain lookups returned multiple values in the form name=value, the
6020 quoting of the values was not always being done properly. Specifically:
6021 (a) If the value started with a double quote, but contained no whitespace,
6023 (b) If the value contained whitespace other than a space character (i.e.
6024 tabs or newlines or carriage returns) it was not quoted.
6025 This fix has been applied to the mysql and pgsql lookups by writing a
6026 separate quoting function and calling it from the lookup code. The fix
6027 should probably also be applied to nisplus, ibase and oracle lookups, but
6028 since I cannot test any of those, I have not disturbed their existing code.
6030 2. A hit in the callout cache for a specific address caused a log line with no
6031 reason for rejecting RCPT. Now it says "Previous (cached) callout
6032 verification failure".
6034 3. There was an off-by-one bug in the queryprogram router. An over-long
6035 return line was truncated at 256 instead of 255 characters, thereby
6036 overflowing its buffer with the terminating zero. As well as fixing this, I
6037 have increased the buffer size to 1024 (and made a note to document this).
6039 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6040 when Exim is waiting for an SMTP response from a remote server, Exim
6041 restarts its select() call on the socket, thereby resetting its timeout.
6042 This is not a problem when such interrupts are rare. Somebody set up a cron
6043 job to run exiwhat every 2 minutes, which is less than the normal select()
6044 timeout (5 or 10 minutes). This meant that the select() timeout never
6045 kicked in because it was always reset. I have fixed this by comparing the
6046 time when an interrupt arrives with the time at the start of the first call
6047 to select(). If more time than the timeout has elapsed, the interrupt is
6048 treated as a timeout.
6050 5. Some internal re-factoring in preparation for the addition of Sieve
6051 extensions (by MH). In particular, the "personal" test is moved to a
6052 separate function, and given an option for scanning Cc: and Bcc: (which is
6053 not set for Exim filters).
6055 6. When Exim created an email address using the login of the caller as the
6056 local part (e.g. when creating a From: or Sender: header line), it was not
6057 quoting the local part when it contained special characters such as @.
6059 7. Installed new OpenBSD configuration files.
6061 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6062 try to make them clearer.
6064 9. Callout options, other than the timeout value, were being ignored when
6065 verifying sender addresses in header lines. For example, when using
6067 verify = header_sender/callout=no_cache
6069 the cache was (incorrectly) being used.
6071 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6072 only the Cygwin environment.
6074 11. When return_path_on_delivery was set as a log selector, if different remote
6075 addresses in the same message used different return paths and parallel
6076 remote delivery occurred, the wrong values would sometimes be logged.
6077 (Whenever a remote delivery process finished, the return path value from
6078 the most recently started remote delivery process was logged.)
6080 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6081 lines when AUTH and/or TLS are in use. This is the "received protocol"
6082 field. Exim used to use "asmtp" for authenticated SMTP, without any
6083 indication (in the protocol name) for TLS use. Now it follows the RFC and
6084 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6085 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6086 names appear in log lines as well as in Received: header lines.
6088 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6089 extensions, and comparison tests, and to fix some bugs.
6091 14. Changes to the "personal" filter test:
6093 (1) The test was buggy in that it was just doing the equivalent of
6094 "contains" tests on header lines. For example, if a user's address was
6095 anne@some.where, the "personal" test would incorrectly be true for
6097 To: susanne@some.where
6099 This test is now done by extracting each address from the header in turn,
6100 and checking the entire address. Other tests that are part of "personal"
6101 are now done using regular expressions (for example, to check local parts
6102 of addresses in From: header lines).
6104 (2) The list of non-personal local parts in From: addresses has been
6105 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6106 taken from the Sieve specification recommendations.
6108 (3) If the message contains any header line starting with "List-" it is
6109 treated as non-personal.
6111 (4) The test for "circular" in the Subject: header line has been removed
6112 because it now seems ill-conceived.
6114 15. Minor typos in src/EDITME comments corrected.
6116 16. Installed latest exipick from John Jetmore.
6118 17. If headers_add on a router specified a text string that was too long for
6119 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6120 of string_sprintf() is now avoided.
6122 18. $message_body_size was not set (it was always zero) when running the DATA
6123 ACL and the local_scan() function.
6125 19. For the "mail" command in an Exim filter, no default was being set for
6126 the once_repeat time, causing a random time value to be used if "once" was
6127 specified. (If the value happened to be <= 0, no repeat happened.) The
6128 default is now 0s, meaning "never repeat". The "vacation" command was OK
6129 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6130 (I found it when inspecting the code).
6132 20. There is now an overall timeout for performing a callout verification. It
6133 defaults to 4 times the callout timeout, which applies to individual SMTP
6134 commands during the callout. The overall timeout applies when there is more
6135 than one host that can be tried. The timeout is checked before trying the
6136 next host. This prevents very long delays if there are a large number of
6137 hosts and all are timing out (e.g. when the network connections are timing
6138 out). The value of the overall timeout can be changed by specifying an
6139 additional sub-option for "callout", called "maxwait". For example:
6141 verify = sender/callout=5s,maxwait=20s
6143 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6144 to the end before writing, but this should make it even safer).
6146 22. Exim was forgetting that it had advertised PIPELINING for the second and
6147 subsequent messages on an SMTP connection. It was also not resetting its
6148 memory on STARTTLS and an internal HELO.
6150 23. When Exim logs an SMTP synchronization error within a session, it now
6151 records whether PIPELINING has been advertised or not.
6153 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6154 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6155 rather than long int.
6157 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6159 26. Added the never_mail option to autoreply.
6165 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6166 crash if the getsockname() call failed; this can happen if a connection is
6167 closed very soon after it is established. The problem was simply in the
6168 order in which certain operations were done, causing Exim to try to write
6169 to the SMTP stream before it had set up the file descriptor. The bug has
6170 been fixed by making things happen in the correct order.
6176 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6177 before the connection was closed, thus losing the rejection response.
6179 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6180 some early Solaris releases, but causes trouble in current releases where
6181 socklen_t is defined.
6183 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6186 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6189 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6190 that releasing the top of it at the end releases what was used for sub-
6191 expansions (unless the block got too big). However, discard this block if
6192 the first thing is a variable or header, so that we can use its block when
6193 it is dynamic (useful for very large $message_headers, for example).
6195 6. Lookups now cache *every* query, not just the most recent. A new, separate
6196 store pool is used for this. It can be recovered when all lookup caches are
6197 flushed. Lookups now release memory at the end of their result strings.
6198 This has involved some general refactoring of the lookup sources.
6200 7. Some code has been added to the store_xxx() functions to reduce the amount
6201 of flapping under certain conditions.
6203 8. log_incoming_interface used to affect only the <= reception log lines. Now
6204 it causes the local interface and port to be added to several more SMTP log
6205 lines, for example "SMTP connection from", and rejection lines.
6207 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6209 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6211 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6212 could be overwritten at the end of the current message (or the start of a
6213 new message if it was set in a HELO ACL). The value is now preserved for
6214 the duration of the SMTP connection.
6216 12. If a transport had a headers_rewrite setting, and a matching header line
6217 contained an unqualified address, that address was qualified, even if it
6218 did not match any rewriting rules. The underlying bug was that the values
6219 of the flags that permit the existence of unqualified sender and recipient
6220 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6221 non-local messages, and by -bnq for local messages) were not being
6222 preserved with the message after it was received.
6224 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6225 "next input=" as part of the text comprising the host identity instead of
6226 the correct text. The code was using the same buffer for two different
6227 strings. However, depending on which order the printing function evaluated
6228 its arguments, the bug did not always show up. Under Linux, for example, my
6229 test suite worked just fine.
6231 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6232 doesn't work with some older versions of Perl. It has been changed to "my",
6233 which in any case is probably the better facility to use.
6235 15. A really picky compiler found some instances of statements for creating
6236 error messages that either had too many or two few arguments for the format
6239 16. The size of the buffer for calls to the DNS resolver has been increased
6240 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6241 for addresses that have a lot of PTR records. This alleviates a problem; it
6242 does not fully solve it.
6244 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6245 buffer now truncates the list and logs the incident, which is the same
6246 action as happens when Exim is looking up a host name and its aliases.
6247 Previously in this situation something unpredictable would happen;
6248 sometimes it was "internal error: store_reset failed".
6250 18. If a server dropped the connection unexpectedly when an Exim client was
6251 using GnuTLS and trying to read a response, the client delivery process
6252 crashed while trying to generate an error log message.
6254 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6255 string, for example:
6257 warn message = H1: something\nH2: something
6259 the text was added as a single header line from Exim's point of view
6260 though it ended up OK in the delivered message. However, searching for the
6261 second and subsequent header lines using $h_h2: did not work. This has been
6262 fixed. Similarly, if a system filter added multiple headers in this way,
6263 the routers could not see them.
6265 20. Expanded the error message when iplsearch is called with an invalid key to
6266 suggest using net-iplsearch in a host list.
6268 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6269 ACLs are no longer actually imposed (and a message to that effect is
6272 22. If a "gecos" field in a passwd entry contained escaped characters, in
6273 particular, if it contained a \" sequence, Exim got it wrong when building
6274 a From: or a Sender: header from that name. A second bug also caused
6275 incorrect handling when an unquoted " was present following a character
6276 that needed quoting.
6278 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6279 was not being matched caselessly.
6281 24. Arranged for all hyphens in the exim.8 source to be escaped with
6284 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6285 properly. Recipient callout cache records were still being keyed to include
6286 the sender, even when use_sender was set false. This led to far more
6287 callouts that were necessary. The sender is no longer included in the key
6288 when use_sender is false.
6290 26. Added "control = submission" modifier to ACLs.
6292 27. Added the ${base62d: operator to decode base 62 numbers.
6294 28. dnsdb lookups can now access SRV records.
6296 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6297 the configuration file.
6299 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6300 (-v) mode. This makes the output for a verbose queue run more intelligible.
6302 31. Added a use_postmaster feature to recipient callouts.
6304 32. Added the $body_zerocount variable, containing the number of binary zero
6305 bytes in the message body.
6307 33. The time of last modification of the "new" subdirectory is now used as the
6308 "mailbox time last read" when there is a quota error for a maildir
6311 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6313 35. Added +ignore_unknown as a special item in host lists.
6315 36. Code for decoding IPv6 addresses in host lists is now included, even if
6316 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6317 address was recognized as an IP address, but was then not correctly decoded
6318 into binary, causing unexpected and incorrect effects when compared with
6325 1. Very minor rewording of debugging text in manualroute to say "list of
6326 hosts" instead of "hostlist".
6328 2. If verify=header_syntax was set, and a header line with an unqualified
6329 address (no domain) and a large number of spaces between the end of the
6330 name and the colon was received, the reception process suffered a buffer
6331 overflow, and (when I tested it) crashed. This was caused by some obsolete
6332 code that should have been removed. The fix is to remove it!
6334 3. When running in the test harness, delay a bit after writing a bounce
6335 message to get a bit more predictability in the log output.
6337 4. Added a call to search_tidyup() just before forking a reception process. In
6338 theory, someone could use a lookup in the expansion of smtp_accept_max_
6339 per_host which, without the tidyup, could leave open a database connection.
6341 5. Added the variables $recipient_data and $sender_data which get set from a
6342 lookup success in an ACL "recipients" or "senders" condition, or a router
6343 "senders" option, similar to $domain_data and $local_part_data.
6345 6. Moved the writing of debug_print from before to after the "senders" test
6348 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6349 problems for message scanning, either using a data ACL, or using
6350 local_scan() because the Received: header was not generated till after they
6351 were called (in order to set the time as the time of reception completion).
6352 I have revised the way this works. The header is now generated after the
6353 body is received, but before the ACL or local_scan() are called. After they
6354 are run, the timestamp in the header is updated.
6360 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6361 before starting a queue runner without re-exec. This happened only when
6362 deliver_drop_privilege was set or when the Exim user was set to root. The
6363 effect of the bug was that timeouts during subsequent deliveries caused
6364 crashes instead of being properly handled. The handler is now left at its
6365 default (and expected) setting.
6367 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6368 message, again when deliver_drop_privilege is set or Exim is run as root.
6369 The bug described in (1) was not present in this case, but the tidying up
6370 of the other signals was missing. I have made the two cases consistent.
6372 3. The ignore_target_hosts setting on a manualroute router was being ignored
6373 for hosts that were looked up using the /MX notation.
6375 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6378 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6379 operated on the sender address. After changing the $sender_address to <>
6380 for the sender address verify, Exim was re-instated it as the original
6381 (before rewriting) address, but remembering that it had rewritten it, so it
6382 wasn't rewriting it again. This bug also had the effect of breaking the
6383 sender address verification caching when the sender address was rewritten.
6385 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6386 This has been changed so that if the ip literal address matches
6387 ignore_target_hosts, the router declines.
6389 7. Added expansion conditions match_domain, match_address, and match_local_
6390 part (NOT match_host).
6392 8. The placeholder for the Received: header didn't have a length field set.
6394 9. Added code to Exim itself and to exim_lock to test for a specific race
6395 condition that could lead to file corruption when using MBX delivery. The
6396 issue is with the lockfile that is created in /tmp. If this file is removed
6397 after a process has opened it but before that process has acquired a lock,
6398 there is the potential for a second process to recreate the file and also
6399 acquire a lock. This could lead to two Exim processes writing to the file
6400 at the same time. The added code performs the same test as UW imapd; it
6401 checks after acquiring the lock that its file descriptor still refers to
6402 the same named file.
6404 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6405 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6408 11. Added the smtp_active_hostname option. If used, this will typically be made
6409 to depend on the incoming interface address. Because $interface_address is
6410 not set up until the daemon has forked a reception process, error responses
6411 that can happen earlier (such as "too many connections") no longer contain
6414 12. If an expansion in a condition on a "warn" statement fails because a lookup
6415 defers, the "warn" statement is abandoned, and the next ACL statement is
6416 processed. Previously this caused the whole ACL to be aborted.
6418 13. Added the iplsearch lookup type.
6420 14. Added ident_timeout as a log selector.
6422 15. Added tls_certificate_verified as a log selector.
6424 16. Added a global option tls_require_ciphers (compare the smtp transport
6425 option of the same name). This controls incoming TLS connections.
6427 17. I finally figured out how to make tls_require_ciphers do a similar thing
6428 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6429 before starting the TLS session.
6431 18. Tabs are now shown as \t in -bP output.
6433 19. If the log selector return_path_on_delivery was set, Exim crashed when
6434 bouncing a message because it had too many Received: header lines.
6436 20. If two routers both had headers_remove settings, and the first one included
6437 a superfluous trailing colon, the final name in the first list and the
6438 first name in the second list were incorrectly joined into one item (with a
6439 colon in the middle).
6445 1. Added -C and -D options to the exinext utility, mainly to make it easier
6446 to include in the automated testing, but these could be helpful when
6447 multiple configurations are in use.
6449 2. The exinext utility was not formatting the output nicely when there was
6450 an alternate port involved in the retry record key, nor when there was a
6451 message id as well (for retries that were specific to a specific message
6452 and a specific host). It was also confused by IPv6 addresses, because of
6453 the additional colons they contain. I have fixed the IPv4 problem, and
6454 patched it up to do a reasonable job for IPv6.
6456 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6457 delivery, the log line now contains "pipelined" if PIPELINING was used.
6459 4. An SMTP transport process used to panic and die if the bind() call to set
6460 an explicit outgoing interface failed. This has been changed; it is now
6461 treated in the same way as a connect() failure.
6463 5. A reference to $sender_host_name in the part of a conditional expansion
6464 that was being skipped was still causing a DNS lookup. This no longer
6467 6. The def: expansion condition was not recognizing references to header lines
6468 that used bh_ and bheader_.
6470 7. Added the _cache feature to named lists.
6472 8. The code for checking quota_filecount in the appendfile transport was
6473 allowing one more file than it should have been.
6475 9. For compatibility with Sendmail, the command line option
6483 and sets the incoming protocol and host name (for trusted callers). The
6484 host name and its colon can be omitted when only the protocol is to be set.
6485 Note the Exim already has two private options, -pd and -ps, that refer to
6486 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6487 "s", but I don't think that's a major issue.
6489 10. A number of refactoring changes to the code, none of which should affect
6492 (a) The number of logging options was getting close to filling up the
6493 32-bit word that was used as a bit map. I have split them into two classes:
6494 those that are passed in the argument to log_write(), and those that are
6495 only ever tested independently outside of that function. These are now in
6496 separate 32-bit words, so there is plenty of room for expansion again.
6497 There is no change in the user interface or the logging behaviour.
6499 (b) When building, for example, log lines, the code previously used a
6500 macro that called string_cat() twice, in order to add two strings. This is
6501 not really sufficiently general. Furthermore, there was one instance where
6502 it was actually wrong because one of the argument was used twice, and in
6503 one call a function was used. (As it happened, calling the function twice
6504 did not affect the overall behaviour.) The macro has been replaced by a
6505 function that can join an arbitrary number of extra strings onto a growing
6508 (c) The code for expansion conditions now uses a table and a binary chop
6509 instead of a serial search (which was left over from when there were very
6510 few conditions). Also, it now recognizes conditions like "pam" even when
6511 the relevant support is not compiled in: a suitably worded error message is
6512 given if an attempt is made to use such a condition.
6514 11. Added ${time_interval:xxxxx}.
6516 12. A bug was causing one of the ddress fields not to be passed back correctly
6517 from remote delivery subprocesses. The field in question was not being
6518 subsequently used, so this caused to problems in practice.
6520 13. Added new log selectors queue_time and deliver_time.
6522 14. Might have fixed a bug in maildirsizefile handling that threw up
6523 "unexpected character" debug warnings, and recalculated the data
6524 unnecessarily. In any case, I expanded the warning message to give more
6527 15. Added the message "Restricted characters in address" to the statements in
6528 the default ACL that block characters like @ and % in local parts.
6530 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6531 Three changes have been made:
6533 (a) There was a serious bug; a negative response to MAIL caused the whole
6534 recipient domain to be cached as invalid, thereby blocking all messages
6535 to all local parts at the same domain, from all senders. This bug has
6536 been fixed. The domain is no longer cached after a negative response to
6537 MAIL if the sender used is not empty.
6539 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6542 (c) A new callout option, "use_sender" has been added for people who want
6543 the modified behaviour.
6549 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6552 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6553 indeed breaks things for older releases.
6555 3. Added additional logging to the case where there is a problem reading data
6556 from a filter that is running in a subprocess using a pipe, in order to
6557 try to track down a specific problem.
6559 4. Testing facility fudge: when running in the test harness and attempting
6560 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6561 getting "No route to host". Convert this to a timeout.
6563 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6566 6. Some OS don't have socklen_t but use size_t instead. This affects the
6567 fifth argument of getsockopt() amongst other things. This is now
6568 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6569 can be set for individual OS. I have set it for SunOS5, OSF1, and
6570 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6571 some earlier ones do not.
6573 7. Change 4.30/15 was not doing the test caselessly.
6575 8. The standard form for an IPv6 address literal was being rejected by address
6576 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6577 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6578 this, as well as the form without the "IPv6" on the front (but only when
6579 address literals are enabled, of course).
6581 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6583 10. Exim crashed if a message with an empty sender address specified by -f
6584 encountered a router with an errors_to setting. This could be provoked only
6585 by a command such as
6589 where an empty string was supplied; "<>" did not hit this bug.
6591 11. Installed PCRE release 4.5.
6593 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6594 remained set. It is now erased.
6596 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6597 times from message ids (which are base 36 rather than the normal 62).
6599 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6600 were being counted as actual protocol errors, and logged if the log
6601 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6602 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6603 and DATA following a set of rejected RCPTs do not count as protocol errors.
6604 In other words, Exim assumes they were pipelined, though this may not
6605 actually be the case. Of course, in all cases the client gets an
6606 appropriate error code.
6608 15. If a lookup fails in an ACL condition, a message about the failure may
6609 be available; it is used if testing the ACL cannot continue, because most
6610 such messages specify what the cause of the deferral is. However, some
6611 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6612 that caused an old message to be retained and used if a later statement
6613 caused a defer, replacing the real cause of the deferral.
6615 16. If an IP address had so many PTR records that the DNS lookup buffer
6616 was not large enough to hold them, Exim could crash while trying to process
6617 the truncated data. It now detects and logs this case.
6619 17. Further to 4.21/58, another change has been made: if (and only if) the
6620 first line of a message (the first header line) ends with CRLF, a bare LF
6621 in a subsequent header line has a space inserted after it, so as not to
6622 terminate the header.
6624 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6625 unnecessarily, used atoi() instead of strtol(), and didn't check the
6626 termination when getting file sizes from file names by regex.
6628 19. Completely re-implemented the support for maildirsize files, in the light
6629 of a number of problems with the previous contributed implementation
6630 (4.30/29). In particular:
6632 . If the quota is zero, the maildirsize file is maintained, but no quota is
6635 . If the maildir directory does not exist, it is created before any attempt
6636 to write a maildirsize file.
6638 . The quota value in the file is just a cache; if the quota is changed in
6639 the transport, the new value overrides.
6641 . A regular expression is available for excluding directories from the
6644 20. The autoreply transport checks the characters in options that define the
6645 message's headers; it allows continued headers, but it was checking with
6646 isspace() after an embedded newline instead of explicitly looking for a
6649 21. If all the "regular" hosts to which an address was routed had passed their
6650 expiry times, and had not reached their retry times, the address was
6651 bounced, even if fallback hosts were defined. Now Exim should go on to try
6654 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6655 equivalent code in the SMTP transport. Some hosts send humungous responses
6656 to HELO/EHLO, more than 1024 it seems.
6658 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6659 gives compiler warnings in some environments. I've now done it "properly",
6662 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6663 (because of problems with the built-in one) was declared to return uschar *
6664 instead of char *, causing compiler failure.
6666 25. Fixed a file descriptor leak when processing alias/forward files.
6668 26. Fixed a minor format string issue in dbfn.c.
6670 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6672 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6673 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6676 29. When a sender address is verified, it is cached, to save repeating the test
6677 when there is more than one recipient in a message. However, when the
6678 verification involves a callout, it is possible for different callout
6679 options to be set for different recipients. It is too complicated to keep
6680 track of this in the cache, so now Exim always runs a verification when a
6681 callout is required, relying on the callout cache for the optimization.
6682 The overhead is duplication of the address routing, but this should not be
6685 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6686 to be verified with callout=postmaster, and the main callout worked but the
6687 postmaster check failed, the verification correctly failed. However, if a
6688 subsequent RCPT command asked for sender verification *without* the
6689 postmaster check, incorrect caching caused this verification also to fail,
6692 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6693 it was not caching the DNS options (qualify_single, search_parents) that
6694 were used when the lookup failed. A subsequent lookup with different
6695 options therefore always gave the same answer, though there were cases
6696 where it should not have. (Example: a "domains = !$mx_any" option on a
6697 dnslookup router: the "domains" option is always processed without any
6698 widening, but the router might have qualify_single set.) Now Exim uses the
6699 cached value only when the same options are set.
6701 32. Added John Jetmore's "exipick" utility to the distribution.
6703 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6704 than a timeout (e.g. a certificate is required, and is not provided), an
6705 Exim server now closes the connection immediately. Previously it waited for
6706 the client to close - but if the client is SSL, it seems that they each
6707 wait for each other, leading to a delay before one of them times out.
6709 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6710 maintained 0.8.x compatibility because I don't think many are using it, and
6711 it is clearly obsolete.
6713 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6716 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6717 client certificate was expired. A simple patch fixes this, though I don't
6718 understand the full logic of why the verify callback is called multiple
6721 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6722 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6723 which causes problems with some clients (such as the Certicom SSL Plus
6724 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6725 disables the coutermeasure allowing Eudora to connect."
6727 38. Exim was not checking that a write() to a log file succeeded. This could
6728 lead to Bad Things if a log got too big, in particular if it hit a file
6729 size limit. Exim now panics and dies if it cannot write to a log file, just
6730 as it does if it cannot open a log file.
6732 39. Modified OS/Makefile-Linux so that it now contains
6734 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6736 The two -D definitions ensure that Exim is compiled with large file
6737 support, which makes it possible to handle log files that are bigger than
6740 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6741 instance) a domain was checked against a named list that involved a lookup,
6742 causing $domain_data to be set, then another domain was checked against the
6743 same list, then the first domain was re-checked, the value of $domain_data
6744 after the final check could be wrong. In particular, if the second check
6745 failed, it could be set empty. This bug probably also applied to
6748 41. The strip_trailing_dot option was not being applied to the address given
6749 with the -f command-line option.
6751 42. The code for reading a message's header from the spool was incrementing
6752 $received_count, but never initializing it. This meant that the value was
6753 incorrect (doubled) while delivering a message in the same process in which
6754 it was received. In the most common configuration of Exim, this never
6755 happens - a fresh exec is done - but it can happen when
6756 deliver_drop_privilege is set.
6758 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6759 it now includes up to 150 characters of the unexpected data in the log
6762 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6763 and building data strings. The size of both of these buffers was 10 000
6764 bytes - far larger than anybody would *ever* want, thought I. Needless to
6765 say, somebody hit the limit. I have increased the maximum line length to
6766 20 000 and the maximum data length of concatenated lines to 100 000. I have
6767 also fixed two bugs, because there was no checking on these buffers. Tsk,
6768 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6769 buffer is too small.
6771 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6772 lsearch lookups. Now it does.
6774 46. When parsing a route_list item in a manualroute router, a fixed-length
6775 buffer was used for the list of hosts. I made this 1024 bytes long,
6776 thinking that nobody would ever have a list of hosts that long. Wrong.
6777 Somebody had a whole pile of complicated expansion conditions, and the
6778 string was silently truncated, leading to an expansion error. It turns out
6779 that it is easier to change to an unlimited length (owing to other changes
6780 that have happened since this code was originally written) than to build
6781 structure for giving a limitation error. The length of the item that
6782 expands into the list of hosts is now unlimited.
6784 47. The lsearch lookup could not handle data where the length of text line was
6785 more than 4095 characters. Such lines were truncated, leading to shortened
6786 data being returned. It should now handle lines of any length.
6788 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6789 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6792 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6793 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6794 no longer happens in comment lines. A list of replacements is now placed
6795 at the head of all of the source files, except those whose only change is
6796 to replace PERL_COMMAND in the very first #! line.
6798 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6799 messages on the queue, with a bottom-up merge sort, using code contributed
6800 by Michael Haardt. This should make operations like -bp somewhat faster on
6801 large queues. It won't affect queue runners, except when queue_run_in_order
6804 51. Installed eximstats 1.31 in the distribution.
6806 52. Added support for SRV lookups to the dnslookup router.
6808 53. If an ACL referred to $message_body or $message_body_end, the value was not
6809 reset for any messages that followed in the same SMTP session.
6811 54. The store-handling optimization for building very long strings was not
6812 differentiating between the different store pools. I don't think this
6813 actually made any difference in practice, but I've tidied it.
6815 55. While running the routers to verify a sender address, $sender_address
6816 was still set to the sender address. This is wrong, because when routing to
6817 send a bounce to the sender, it would be empty. Therefore, I have changed
6818 it so that, while verifying a sender address, $sender_address is set to <>.
6819 (There is no change to what happens when verifying a recipient address.)
6821 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6822 target A or AAAA records (if not already returned) without resetting the
6823 qualify_single or search_parents options of the DNS resolver. These are
6824 inappropriate in this case because the targets of MX and SRV records must
6825 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6826 target that, when qualified, matched something in the local domain. These
6827 two options are now turned off when doing these lookups.
6829 57. It seems that at least some releases of Reiserfs (which does not have the
6830 concept of a fixed number of inodes) returns zero and not -1 for the
6831 number of available inodes. This interacted badly with check_spool_inodes,
6832 which assumed that -1 was the "no such thing" setting. What I have done is
6833 to check that the total number of inodes is greater than zero before doing
6834 the test of how many are available.
6836 58. When a "warn" ACL statement has a log_message modifier, the message is
6837 remembered, and not repeated. This is to avoid a lot of repetition when a
6838 message has many recipients that cause the same warning to be written.
6839 However, Exim was preserving the list of already written lines for an
6840 entire SMTP session, which doesn't seem right. The memory is now reset if a
6841 new message is started.
6843 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6844 debugging output unless log_rewrite was also set.
6846 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6847 of (int)(handle) when we know that handle contains (void *)(-1).
6849 61. The Exim daemon panic-logs an error return when it closes the incoming
6850 connection. However "connection reset by peer" seems to be common, and
6851 isn't really an error worthy of noting specially, so that particular error
6854 62. When Exim is trying to find all the local interfaces, it used to panic and
6855 die if the ioctl to get the interface flags failed. However, it seems that
6856 on at least one OS (Solaris 9) it is possible to have an interface that is
6857 included in the list of interfaces, but for which you get a failure error
6858 for this call. This happens when the interface is not "plumbed" into a
6859 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6860 failure of the "get flags" call assumes that the interface is down.
6862 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6863 makes life easier for people who are doing arithmetic on fields extracted
6864 from dates, where you often get leading zeros that should not be
6865 interpreted as octal.
6867 64. Added qualify_domain to the redirect router, to override the global
6870 65. If a pathologically long header line contained very many addresses (the
6871 report of this problem mentioned 10 000) and each of them was rewritten,
6872 Exim could use up a very large amount of memory. (It kept on making new
6873 copies of the header line as it rewrote, and never released the old ones.)
6874 At the expense of a bit more processing, the header rewriting function has
6875 been changed so that it no longer eats memory in this way.
6877 66. The generation of the Received: header has been moved from the time that a
6878 message starts to be received, to the time that it finishes. The timestamp
6879 in the Received: header should now be very close to that of the <= log
6880 line. There are two side-effects of this change:
6882 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6883 logged header lines no longer include the local Received: line, because
6884 it has not yet been created. The same applies to a copy of the message
6885 that is returned to a non-SMTP sender when a message is rejected.
6887 (b) When a filter file is tested using -bf, no additional Received: header
6888 is added to the test message. After some thought, I decided that this
6891 This change does not affect the value of $received_for. It is still set
6892 after address rewriting, but before local_scan() is called.
6894 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6896 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6897 gave an unhelpful panic error message, and a defer error. I have managed to
6898 change this behaviour so that it now rejects any supplied certificate,
6899 which seems right, as the list of acceptable certificates is empty.
6901 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6902 gave an unhelpful defer error. I have not managed to make this reject any
6903 supplied certificates, but the error message it gives is "no certificate
6904 supplied", which is not helpful.
6906 70. exigrep's output now also includes lines that are not associated with any
6907 message, but which match the given pattern. Implemented by a patch from
6908 Martin Sluka, which also tidied up the Perl a bit.
6910 71. Recipient callout verification, like sender verification, was using <> in
6911 the MAIL FROM command. This isn't really the right thing, since the actual
6912 sender may affect whether the remote host accepts the recipient or not. I
6913 have changed it to use the actual sender in the callout; this means that
6914 the cache record is now keyed on a recipient/sender pair, not just the
6915 recipient address. There doesn't seem to be a real danger of callout loops,
6916 since a callout by the remote host to check the sender would use <>.
6917 [SEE ABOVE: changed after hitting problems.]
6919 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6920 temporary errors. However, in the case of such a code being given after
6921 the end of a data transmission (i.e. after ".") Exim was failing to write
6922 a retry record for the message. (Yes, there was some broken host that was
6923 actually sending 8xx at this point.)
6925 73. An unknown lookup type in a host list could cause Exim to panic-die when
6926 the list was checked. (An example that provoked this was putting <; in the
6927 middle of a list instead of at the start.) If this happened during a DATA
6928 ACL check, a -D file could be left lying around. This kind of configuration
6929 error no longer causes Exim to die; instead it causes a defer error. The
6930 incident is still logged to the main and panic logs.
6932 74. Buglet left over from Exim 3 conversion. The message "too many messages
6933 in one connection" was written to the rejectlog but not the mainlog, except
6934 when address rewriting (yes!) was being logged.
6936 75. Added write_rejectlog option.
6938 76. When a system filter was run not as root (that is, when system_filter_user
6939 was set), the values of the $n variables were not being returned to the
6940 main process; thus, they were not subsequently available in the $sn
6943 77. Added +return_path_on_delivery log selector.
6945 78. A connection timeout was being treated differently from recipients deferred
6946 when testing hosts_max_try with a message that was older than the host's
6947 retry timeout. (The host should not be counted, thus allowing all hosts to
6948 be tried at least once before bouncing.) This may have been the cause of an
6949 occasionally reported bug whereby a message would remain on the queue
6950 longer than the retry timeout, but would be bounced if a delivery was
6951 forced. I say "may" because I never totally pinned down the problem;
6952 setting up timeout/retry tests is difficult. See also the next item.
6954 79. The ultimate address timeout was not being applied to errors that involved
6955 a combination of host plus message (for example, a timeout on a MAIL
6956 command). When an address resolved to a number of possible hosts, and they
6957 were not all tried for each delivery (e.g. because of hosts_max_try), a
6958 message could remain on the queue longer than the retry timeout.
6960 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6963 81. Fixed an obscure SMTP outgoing bug which required at least the following
6964 conditions: (a) there was another message waiting for the same server;
6965 (b) the server returned 5xx to all RCPT commands in the first message so
6966 that the message was not completed; (c) the server dropped the connection
6967 or gave a negative response to the RSET that Exim sends to abort the
6968 transaction. The observed case was a dropped connection after DATA that had
6969 been sent in pipelining mode. That is, the server had advertised PIPELINING
6970 but was not implementing it correctly. The effect of the bug was incorrect
6971 behaviour, such as trying another host, and this could lead to a crash.
6977 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6978 and daemon.c were passed as pointers to ints; they should have been
6979 pointers to socklen_t variables (which are typically unsigned ints).
6981 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6984 3. Fixed a really odd bug that affected only the testing scheme; patching a
6985 certain fixed string in the binary changed the value of another string that
6986 happened to be identical to the end of the original first string.
6988 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6989 name", it returns that address as the IP address. On some operating
6990 systems (e.g. Solaris), it also passes back the IP address string as the
6991 "host name". However, on others (e.g. Linux), it passes back an empty
6992 string. Exim wasn't checking for this, and was changing the host name to an
6993 empty string, assuming it had been canonicalized.
6995 5. Although rare, it is permitted to have more than one PTR record for a given
6996 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6997 all the names associated with an address, because they do in Solaris.
6998 However, it seems that they do not in Linux for data that comes from the
6999 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7000 I found this out when I moved to a new Linux workstation and tried to run
7001 the Exim test suite.
7003 To get round this problem I have changed the code so that it now does its
7004 own call to the DNS to look up PTR records when searching for a host name.
7005 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7006 addresses that are only in /etc/hosts are still found.
7008 This behaviour is, however, controlled by an option called host_lookup_
7009 order, which defaults to "bydns:byaddr". If people want to use the other
7010 order, or indeed, just use one or the other means of lookup, they can
7011 specify it in this variable.
7013 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7014 some operating systems, this comes back from gethostbyaddr() as an empty
7015 string, and this is what Exim used to test for. However, it seems that in
7016 other systems, "." is yielded. Exim now tests for this case too.
7018 7. The values of check_spool_space and check_log_space are now held internally
7019 as a number of kilobytes instead of an absolute number of bytes. If a
7020 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7021 kilobyte. This means that much larger values can be stored.
7023 8. Exim monitor: an attempt to get the action menu when not actually pointing
7024 at a message produces an empty menu entitled "No message selected". This
7025 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7026 no entries in it ("Shell widget menu has zero width and/or height"). So I
7027 have added a single, blank menu entry in this case.
7029 9. Added ${quote_local_part.
7031 10. MIME decoding is now applied to the contents of Subject: header lines when
7034 11. Now that a reference to $sender_host_address automatically causes a reverse
7035 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7036 host lookup before query-style lookups in lists that might use this
7037 variable. This has therefore been abolished, and the "net-" prefix is no
7038 longer necessary for query-style lookups.
7040 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7041 have been a typo for LFLAGS, so it has been changed.
7043 13. The install script calls Exim with "-C /dev/null" in order to find the
7044 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7045 to be output. However, since Exim outputs its version number before the
7046 error, it didn't break the script. It just looked ugly. I fixed this by
7047 always allowing "-C /dev/null" if the caller is root.
7049 14. Ignore overlarge ACL variable number when reading spool file - insurance
7050 against a later release with more variables having written the file.
7052 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7053 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7054 this, as well as the form without the "IPv6" on the front.
7056 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7057 OS/Makefile-Darwin file.
7059 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7060 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7062 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7063 variable. However, some OS use ints for inodes. I've added cast to long int
7064 to get rid of the compiler warning.
7066 19. I had forgotten to lock out "/../" in configuration file names when
7067 ALT_CONFIG_PREFIX was set.
7069 20. Routers used for verification do not need to specify transports. However,
7070 if such a router generated a host list, and callout was configured, Exim
7071 crashed, because it could not find a port number from the (non-existent)
7072 transport. It now assumes port 25 in this circumstance.
7074 21. Added the -t option to exigrep.
7076 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7077 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7078 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7079 them is set, LOOKUP_LSEARCH is forced.
7081 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7083 24. Added sender and host information to the "rejected by local_scan()" log
7084 line; previously there was no indication of these.
7086 25. Added .include_if_exists.
7088 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7089 Linux. It turns out that not all file systems support this. Apparently some
7090 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7091 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7092 not supported on the file descriptor, is now ignored when Exim is trying to
7093 sync a directory. This applies only to Linux.
7095 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7097 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7098 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7101 29. Added support for maildirsize files from supplied patch (modified a bit).
7103 30. The use of :fail: followed by an empty string could lead Exim to respond to
7104 sender verification failures with (e.g.):
7106 550 Verification failed for <xxx>
7107 550 Sender verify failed
7109 where the first response line was missing the '-' that indicates it is not
7110 the final line of the response.
7112 31. The loop for finding the name of the user that called Exim had a hardwired
7113 limit of 10; it now uses the value of finduser_retries, which is used for
7114 all other user lookups.
7116 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7119 33. Exim was neglecting to zero errno before one call of strtol() when
7120 expanding a string and expecting an integer value. On some systems this
7121 resulted in spurious "integer overflow" errors. Also, it was casting the
7122 result into an int without checking.
7124 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7125 did not work. The code looks as if it has *never* worked, though it appears
7126 to have been documented since at least release 1.62. I have made it work.
7128 35. The "timeout_DNS" error in retry rules, also documented since at least
7129 1.62, also never worked. As it isn't clear exactly what this means, and
7130 clearly it isn't a major issue, I have abolished the feature by treating it
7131 as "timeout", and writing a warning to the main and panic logs.
7133 36. The display of retry rules for -brt wasn't always showing the error code
7136 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7137 timeout_connect_A, timeout_connect_MX.
7139 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7140 to the empty sender.
7142 39. The daemon was not analysing the content of -oX till after it had closed
7143 stderr and disconnected from the controlling terminal. This meant that any
7144 syntax errors were only noted on the panic log, and the return code from
7145 the command was 0. By re-arranging the code a little, I've made the
7146 decoding happen first, so such errors now appear on stderr, and the return
7147 code is 1. However, the actual setting up of the sockets still happens in
7148 the disconnected process, so errors there are still only recorded on the
7151 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7152 connections (as happens on some IP stacks) was logged at start up time as
7153 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7154 it from "IPv6 and IPv4", which means that two separate sockets are being
7157 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7158 says whether AF_INET or AF_INET6 was passed as an argument.
7160 42. Exiwhat output was messed up when time zones were included in log
7163 43. Exiwhat now gives more information about the daemon's listening ports,
7164 and whether -tls-on-connect was used.
7166 44. The "port" option of the smtp transport is now expanded.
7168 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7169 silently ignored. Now an error message is written to the main and panic
7172 46. There's a new ACL modifier called "logwrite" which writes to a log file
7173 as soon as it is encountered.
7175 47. Added $local_user_uid and $local_user_gid at routing time.
7177 48. Exim crashed when trying to verify a sender address that was being
7180 49. Exim was recognizing only a space character after ".include". It now also
7181 recognizes a tab character.
7183 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7184 extracting the relevant information from the specification. The man page no
7185 longer contains scrambled data for the -d option, and I've added a section
7186 at the front about calling Exim under different names.
7188 51. Added "extra_headers" argument to the "mail" command in filter files.
7190 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7193 53. Installed eximstats 1.29.
7195 54. Added transport_filter_timeout as a generic transport option.
7197 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7198 Cc: header lines. This was required by RFC 822, but it not required by RFC
7201 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7202 incoming messages that did not have them. Now it does so only if the
7203 message originates locally, that is, if there is no associated remote host
7204 address. When Resent- header lines are present, this applies to the Resent-
7205 lines rather than the non-Resent- lines.
7207 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7208 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7210 58. Messages for configuration errors now include the name of the main
7211 configuration file - useful now that there may be more than one file in a
7212 list (.included file names were always shown).
7214 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7215 for those rare installations that do not start the daemon as root or run it
7216 setuid root. I've cut out the call to initgroups() if the daemon is not
7219 60. The Exim user and group can now be bound into the binary as text strings
7220 that are looked up at the start of Exim's processing.
7222 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7224 62. Added $mailstore_basename variable.
7226 63. Installed patch to sieve.c from Michael Haardt.
7228 64. When Exim failed to open the panic log after failing to open the main log,
7229 the original message it was trying to log was written to stderr and debug
7230 output, but if they were not available (the usual case in production), it
7231 was lost. Now it is written to syslog before the two lines that record the
7232 failures to open the logs.
7234 65. Users' Exim filters run in subprocesses under the user's uid. It is
7235 possible for a "deliver" command or an alias in a "personal" command to
7236 provoke an address rewrite. If logging of address rewriting is configured,
7237 this fails because the process is not running as root or exim. There may be
7238 a better way of dealing with this, but for the moment (because 4.30 needs
7239 to be released), I have disabled address rewrite logging when running a
7240 filter in a non-root, non-exim process.
7246 1. The buildconfig auxiliary program wasn't quoting the value set for
7247 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7248 not defined. This bug was masked in 4.22 by the effect that was fixed in
7251 2. Some messages that were rejected after a message id was allocated were
7252 shown as "incomplete" by exigrep. It no longer does this for messages that
7253 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7255 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7256 have allow_domain_literals set, the ID did not get logged in the <= line.
7257 Domain literals are now always recognized in Message-ID: header lines.
7259 4. The first argument for a ${extract expansion item is the key name or field
7260 number. Leading and trailing spaces in this item were not being ignored,
7261 causing some misleading effects.
7263 5. When deliver_drop_privilege was set, single queue runner processes started
7264 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7265 same command in the process it spins off) were not dropping privilege.
7267 6. When the daemon running as "exim" started a queue runner, it always
7268 re-executed Exim in the spun-off process. This is a waste of effort when
7269 deliver_drop_privilege is set. The new process now just calls the
7270 queue-runner function directly.
7276 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7279 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7280 ignored. Though the use of -oP was forcing the writing of a pid file, it
7281 was always written to the default place.
7283 3. If the message "no IP address found for host xxxx" is generated during
7284 incoming verification, it is now followed by identification of the incoming
7285 connection (so you can more easily find what provoked it).
7287 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7289 5. Added some features to "harden" Exim a bit more against certain attacks:
7291 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7292 be put in Local/Makefile. This is like the never_users runtime option,
7293 but it cannot be overridden. The default setting is "root".
7295 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7296 prefix string with which any file named in a -C command line option
7299 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7300 is retained for -C and -D only if the caller of Exim is root. Without
7301 it, the exim user may also use -C and -D and retain privilege.
7303 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7304 command line option is disabled.
7306 6. Macro names set by the -D option must start with an upper case letter, just
7307 like macro names defined in the configuration file.
7309 7. Added "dereference=" facility to LDAP.
7311 8. Two instances of the typo "uknown" in the source files are fixed.
7313 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7314 the Configure-Makefile script screwed up while processing it.
7316 10. Incorporated PCRE 4.4.
7318 11. The SMTP synchronization check was not operating right at the start of an
7319 SMTP session. For example, it could not catch a HELO sent before the client
7320 waited for the greeting. There is now a check for outstanding input at the
7321 point when the greeting is written. Because of the duplex, asynchronous
7322 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7323 way, but not yet received, when the check is performed.
7325 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7326 on TCP/IP sockets, because this apparently causes some broken clients to
7329 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7330 unchanged) from the Cygwin maintainer.
7332 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7333 was supported instead of testing for mbx. Effectively a typo.
7335 15. The spa authenticator server code was not checking that the input it
7336 received was valid base64.
7338 16. The debug output line for the "set" modifier in ACLs was not showing the
7339 name of the variable that was being set.
7341 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7343 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7344 lookup on its own. Something else was needed to trigger the lookup. For
7345 example, a match in host_lookup or the need for a host name in a host list.
7346 Now, if $sender_host_name is referenced and the host name has not yet been
7347 looked up, a lookup is performed. If the lookup fails, the variable remains
7348 empty, and $host_lookup_failed is set to "1".
7350 19. Added "eqi" as a case-independent comparison operator.
7352 20. The saslauthd authentication condition could segfault if neither service
7353 nor realm was specified.
7355 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7356 error message that was logged was misleading, and incoming SMTP
7357 connections were dropped. The message is now more accurate, and temporary
7358 errors are given to SMTP connections.
7360 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7361 (except RSET) with a 421 error, until QUIT is received. However, it was
7362 failing to send a response to QUIT.
7364 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7365 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7366 ACL is now run after the helo_try_verify_hosts code.
7368 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7369 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7370 case-independent, so other case variants are also recognized). Apparently
7371 some systems use these upper case variants.
7373 25. If more than two messages were waiting for the same host, and a transport
7374 filter was specified for the transport, Exim sent two messages over the
7375 same TCP/IP connection, and then failed with "socket operation on non-
7376 socket" when it tried to send the third.
7378 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7380 27. The extern definition of crypt16() in expand.c was not being excluded when
7381 the OS had its own crypt16() function.
7383 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7384 as a preferred synonym for return_size_limit, both as an option and as an
7387 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7389 30. Changed the default configuration ACL to relax the local part checking rule
7390 for addresses that are not in any local domains. For these addresses,
7391 slashes and pipe symbols are allowed within local parts, but the sequence
7392 /../ is explicitly forbidden.
7394 31. SPA server authentication was not clearing the challenge buffer before
7397 32. log_message in a "warn" ACL statement was writing to the reject log as
7398 well as to the main log, which contradicts the documentation and doesn't
7399 seem right (because no rejection is happening). So I have stopped it.
7401 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7402 However, I am unable to do any testing of this.
7404 34. Fixed an infelicity in the appendfile transport. When checking directories
7405 for a mailbox, to see if any needed to be created, it was accidentally
7406 using path names with one or more superfluous leading slashes; tracing
7407 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7409 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7410 contents are added to the log line that is written for every discarded
7411 recipient. (Previously a log_message setting was ignored.)
7413 36. The ${quote: operator now quotes the string if it is empty.
7415 37. The install script runs exim in order to find its version number. If for
7416 some reason other than non-existence or emptiness, which it checks, it
7417 could not run './exim', it was installing it with an empty version number,
7418 i.e. as "exim-". This error state is now caught, and the installation is
7421 38. An argument was missing from the function that creates an error message
7422 when Exim fails to connect to the socket for saslauthd authentication.
7423 This could cause Exim to crash, or give a corrupted message.
7425 39. Added isip, isip4, and isip6 to ${if conditions.
7427 40. The ACL variables $acl_xx are now saved with the message, and can be
7428 accessed later in routers, transports, and filters.
7430 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7431 strings in the file are not string-expanded.
7433 42. If a MAIL command specified a SIZE value that was too large to fit into an
7434 int variable, the check against message_size_limit failed. Such values are
7435 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7436 day this will have to be increased, but I don't think I want to be around
7437 when emails are that large.
7444 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7445 iconv() is not standard in FreeBSD.
7447 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7448 IPv6 enabled. The observed symptom was a segmentation fault on return from
7449 the function os_common_find_running_interfaces() in src/os.c.
7451 3. In the check_special_case() function in daemon.c I had used "errno" as an
7452 argument name, which causes warnings on some systems. This was basically a
7453 typo, since it was named "eno" in the comments!
7455 4. The code that waits for the clock to tick (at a resolution of some fraction
7456 of a second) so as to ensure message-id uniqueness was always waiting for
7457 at least one whole tick, when it could have waited for less. [This is
7458 almost certainly not relevant at current processor speeds, where it is
7459 unlikely to ever wait at all. But we try to future-proof.]
7461 5. The function that sleeps for a time interval that includes fractions of a
7462 second contained a race. It did not block SIGALRM between setting the
7463 timer, and suspending (a couple of lines later). If the interval was short
7464 and the sigsuspend() was delayed until after it had expired, the suspension
7465 never ended. On busy systems this could lead to processes getting stuck for
7468 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7469 process, before it forks any delivery processes. The open lookup caching
7470 mechanism meant that the open file or database connection was passed into
7471 the delivery process. The problem was that delivery processes always tidy
7472 up cached lookup data. This could cause a problem for the next delivery
7473 process started by the queue runner, because the external queue runner
7474 process does not know about the closure. So the next delivery process
7475 still has data in the lookup cache. In the case of a file lookup, there was
7476 no problem because closing a file descriptor in a subprocess doesn't affect
7477 the parent. However, if the lookup was caching a connection to a database,
7478 the connection was closed, and the second delivery process was likely to
7479 see errors such as "PGSQL: query failed: server closed the connection
7480 unexpectedly". The problem has been fixed by closing all cached lookups
7481 in a queue runner before running a delivery process.
7483 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7484 seem to have the "const" qualifier which it has on other OS. I've
7487 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7488 *of the same type* (client or server) with the same public name that an
7489 error should be diagnosed.
7491 9. When Exim looked up a host name for an IP address, but failed to find the
7492 original IP address when looking up the host name (a safety check), it
7493 output the message "<ip address> does not match any IP for NULL", which was
7494 confusing, to say the least. The bug was that the host name should have
7495 appeared instead of "NULL".
7497 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7498 user that is built into the binary, and the -C or -D options is used, root
7499 privilege is dropped before the configuration file is read. In addition,
7500 logging is switched to stderr instead of the normal log files. If the
7501 configuration then re-defines the Exim user, the unprivileged environment
7502 is probably not what is expected, so Exim logs a panic warning message (but
7505 However, if deliver_drop_privilege is set, the unprivileged state may well
7506 be exactly what is intended, so the warning has been cut out in that case,
7507 and Exim is allowed to try to write to its normal log files.
7513 1. smtp_return_error_details was not giving details for temporary sender
7514 or receiver verification errors.
7516 2. Diagnose a configuration error if two authenticators have the same public
7519 3. Exim used not to create the message log file for a message until the first
7520 delivery attempt. This could be confusing when incoming messages were held
7521 for policy or load reasons. The message log file is now created at the time
7522 the message is received, and an initial "Received" line is written to it.
7524 4. The automatically generated man page for command line options had a minor
7525 bug that caused no ill effects; however, a more serious problem was that
7526 the procedure for building the man page automatically didn't always
7527 operate. Consequently, release 4.20 contains an out-of-date version. This
7528 shouldn't happen again.
7530 5. When building Exim with embedded Perl support, the script that builds the
7531 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7532 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7534 6. The freeze_tell option was not being used for messages that were frozen on
7535 arrival, either by an ACL or by local_scan().
7537 7. Added the smtp_incomplete_transaction log selector.
7539 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7540 was accepting AUTH without a new EHLO.
7542 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7543 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7546 10. Logging of TCP/IP connections (when configured) now happens in the main
7547 daemon process instead of the child process, so that the TCP/IP connection
7548 count is more accurate (but it can never be perfect).
7550 11. The use of "drop" in a nested ACL was not being handled correctly in the
7551 outer ACL. Now, if condition failure induced by the nested "drop" causes
7552 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7553 or "require"), the connection is dropped.
7555 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7556 that yield "discard" can now be used with an "accept" or a "discard" verb,
7557 but an error is generated for any others (because I can't see a useful way
7558 to define what should happen).
7560 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7561 are now processed in the same way as lines in the Exim configuration file.
7562 In particular, continuation lines are supported.
7564 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7566 15. Added -ti meaning -t -i.
7568 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7569 domains, and warn by logging if others are found.
7571 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7572 returned from GIFCONF when Exim is trying to find the list of interfaces on
7573 a host. The code in os.c has been modified to copy each ifreq to an aligned
7574 structure in all cases.
7576 Also, in some cases, the returned ifreq's were being copied to a 'struct
7577 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7578 means the last couple of bytes of an IPv6 address could be chopped if the
7579 ifreq contained only a normal sockaddr (14 bytes storage).
7581 18. Named domain lists were not supported in the hosts_treat_as_local option.
7582 An entry such as +xxxx was not recognized, and was treated as a literal
7585 19. Ensure that header lines added by a DATA ACL are included in the reject log
7586 if the ACL subsequently rejects the message.
7588 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7589 MD5 (which is deprecated).
7591 21. When testing a filter file using -bf, Exim was writing a message when it
7592 took the sender from a "From " line in the message, but it was not doing so
7593 when it took $return_path from a Return-Path: header line. It now does.
7595 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7596 with a valid header line field name (a series of printing characters
7597 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7599 23. Changed "disc" in the source to "disk" to conform to the documentation and
7600 the book and for uniformity.
7602 24. Ignore Sendmail's -Ooption=value command line item.
7604 25. When execve() failed while trying to run a command in a pipe transport,
7605 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7606 could be confused with a return value of 69 from the command itself. This
7607 has been changed to 127, the value the shell returns if it is asked to run
7608 a non-existent command. The wording for the related log line suggests a
7609 non-existent command as the problem.
7611 26. If received_header_text expands to an empty string, do not add a Received:
7612 header line to the message. (Well, it adds a token one on the spool, but
7613 marks it "old" so that it doesn't get used or transmitted.)
7615 27. Installed eximstats 1.28 (addition of -nt option).
7617 28. There was no check for failure on the call to getsockname() in the daemon
7618 code. This can fail if there is a shortage of resources on the system, with
7619 ENOMEM, for example. A temporary error is now given on failure.
7621 29. Contrary to the C standard, it seems that in some environments, the
7622 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7623 program. Exim now does this explicitly; it affects the formatting of
7624 timestamps using strftime().
7626 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7627 complaints. I've now initialized all the variables, to avoid this.
7629 32. Header lines added by a system filter were not being "seen" during
7630 transport-time rewrites.
7632 33. The info_callback() function passed to OpenSSL is set up with type void
7633 (*)(SSL *, int, int), as described somewhere. However, when calling the
7634 function (actually a macro) that sets it up, the type void(*)() is
7635 expected. I've put in a cast to prevent warnings from picky compilers.
7637 34. If a DNS black list lookup found a CNAME record, but there were no A
7638 records associated with the domain it pointed at, Exim crashed.
7640 35. If a DNS black list lookup returned more than one A record, Exim ignored
7641 all but the first. It now scans all returned addresses if a particular IP
7642 value is being sought. In this situation, the contents of the
7643 $dnslist_value variable are a list of all the addresses, separated by a
7646 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7647 to accept a host name and all its aliases if the forward lookup for any of
7648 them yielded the IP address of the incoming connection. Now it accepts only
7649 those names whose forward lookup yields the correct IP address. Any other
7650 names are discarded. This closes a loophole whereby a rogue DNS
7651 administrator could create reverse DNS records to break through a
7652 wildcarded host restriction in an ACL.
7654 37. If a user filter or a system filter that ran in a subprocess used any of
7655 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7656 the wrong values were passed to the pipe command ($thisaddress had the
7657 value of $0, $0 had the value of $1, etc). This bug was introduced by
7658 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7660 38. Improved the line breaking for long SMTP error messages from ACLs.
7661 Previously, if there was no break point between 40 and 75 characters, Exim
7662 left the rest of the message alone. Two changes have been made: (a) I've
7663 reduced the minimum length to 35 characters; (b) if it can't find a break
7664 point between 35 and 75 characters, it looks ahead and uses the first one
7665 that it finds. This may give the occasional overlong line, but at least the
7666 remaining text gets split now.
7668 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7669 file descriptors might be low, and that setting 1000 would always raise it.
7670 It turns out that in some environments, the limit is already over 1000 and
7671 that lowering it causes trouble. So now Exim takes care not to decrease it.
7673 40. When delivering a message, the value of $return_path is set to $sender_
7674 address at the start of routing (routers may change the value). By an
7675 oversight, this default was not being set up when an address was tested by
7676 -bt or -bv, which affected the outcome if any router or filter referred to
7679 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7680 log only when "message" or "log_message" are set. However, if one of the
7681 conditions was an address verification, or a call to a nested ACL, the
7682 messages generated by the underlying test were being passed through. This
7683 no longer happens. The underlying message is available in $acl_verify_
7684 message for both "message" and "log_message" expansions, so it can be
7685 passed through if needed.
7687 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7688 new expansion $bh_ to give the encoded byte string without charset
7689 translation. Translation happens only if iconv() is available; HAVE_ICONV
7690 indicates this at build time. HEADERS_CHARSET gives the charset to
7691 translate to; headers_charset can change it in the configuration, and
7692 "headers charset" can change it in an individual filter file.
7694 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7695 that creates RFC 2047 encoded "words" labels them as that charset instead
7696 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7697 expansion operator; (ii) when Exim creates a From: line for a local
7698 message; (iii) when a header line is rewritten to include a "phrase" part.
7700 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7701 buggy, causing it to skip the first lines of messages whose message ID
7702 ended in 'D'. This would not have bitten before Exim release 4.14, because
7703 message IDs were unlikely to end in 'D' before then. The effect was to have
7704 incorrect size information for certain domains.
7706 45. #include "config.h" was missing at the start of the crypt16.c module. This
7707 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7710 46. If there was a timeout during a "random" callout check, Exim treated it as
7711 a failure of the random address, and carried on sending RSET and the real
7712 address. If the delay was just some slowness somewhere, the response to the
7713 original RCPT would be taken as a response to RSET and so on, causing
7714 mayhem of various kinds.
7716 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7717 when I implemented it. It didn't allow for the fact that some option values
7718 may legitimately be negative (e.g. size_addition), and it didn't even do
7719 the right test for positive values.
7721 48. Domain names in DNS records are case-independent. Exim always looks them up
7722 in lower case. Some resolvers return domain names in exactly the case they
7723 appear in the zone file, that is, they may contain uppercase letters. Not
7724 all resolvers do this - some return always lower case. Exim was treating a
7725 change of case by a resolver as a change of domain, similar to a widening
7726 of a domain abbreviation. This triggered its re-routing code and so it was
7727 trying to route what was effectively the same domain again. This normally
7728 caused routing to fail (because the router wouldn't handle the domain
7729 twice). Now Exim checks for this case specially, and just changes the
7730 casing of the domain that it ultimately uses when it transmits the message
7733 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7736 50. If a filter generated a file delivery with a non-absolute name (possible if
7737 no home directory exists for the router), the forbid_file option was not
7740 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7741 the existing equality matching.
7743 52. Exim was using ints instead of ino_t variables in some places where it was
7744 dealing with inode numbers.
7746 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7747 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7748 called TMPDIR, and if it finds it is different, it changes its value.
7750 54. The smtp_printf() function is now made available to local_scan() so
7751 additional output lines can be written before returning. There is also an
7752 smtp_fflush() function to enable the detection of a dropped connection.
7753 The variables smtp_input and smtp_batched_input are exported to
7756 55. Changed the default runtime configuration: the message "Unknown user"
7757 has been removed from the ACL, and instead placed on the localuser router,
7758 using the cannot_route_message feature. This means that any verification
7759 failures that generate their own messages won't get overridden. Similarly,
7760 the "Unrouteable address" message that was in the ACL for unverifiable
7761 relay addresses has also been removed.
7763 56. Added hosts_avoid_esmtp to the smtp transport.
7765 57. The exicyclog script was not checking for the esoteric option
7766 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7767 will work only if exicyclog is run under the appropriate euid.
7769 58. Following a discussion on the list, the rules by which Exim recognises line
7770 endings on incoming messages have been changed. The -dropcr and drop_cr
7771 options are now no-ops, retained only for backwards compatibility. The
7772 following line terminators are recognized: LF CRLF CR. However, special
7773 processing applies to CR:
7775 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7776 nor a local message in the state where . is a terminator.
7778 (ii) If a bare CR is encountered in a header line, an extra space is added
7779 after the line terminator so as not to end the header. The reasoning
7780 behind this is that bare CRs in header lines are most likely either
7781 to be mistakes, or people trying to play silly games.
7783 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7784 was being incorrectly given as 18 bytes larger than it should have been.
7785 This is a VOB (very old bug).
7787 60. This may never have affected anything current, but just in case it has:
7788 When the local host is found other than at the start of a list of hosts,
7789 the local host, those with the same MX, and any that follow, are discarded.
7790 When the list in question was part of a longer list of hosts, the following
7791 hosts (not currently being processed) were also being discarded. This no
7792 longer happens. I'm not sure if this situation could ever has previously
7795 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7798 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7799 header. This is something that is recommended in a new Internet Draft, and
7800 is something that is documented as being done by Sendmail. There are two
7801 possible values. For messages generated by the autoreply transport, Exim
7804 Auto-Submitted: auto-replied
7806 whereas for all other generated messages (e.g. bounces) it adds
7808 Auto-Submitted: auto-generated
7810 63. The "personal" condition in filters now includes a test for the
7811 Auto-Submitted: header. If it contains the string "auto-" the message it
7812 not considered personal.
7814 64. Added rcpt_include_affixes as a generic transport option.
7816 65. Added queue_only_override (default true).
7818 66. Added the syslog_duplication option.
7820 67. If what should have been the first header line of a message consisted of
7821 a space followed by a colon, Exim was mis-interpreting it as a header line.
7822 It isn't of course - it is syntactically invalid and should therefore be
7823 treated as the start of the message body. The misbehaviour could have
7824 caused a number of strange effects, including loss of data in subsequent
7825 header lines, and spool format errors.
7827 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7828 client host had authenticated. This control can now be exercised by an ACL
7829 for more flexibility.
7831 69. By default, callouts do not happen when testing with -bh. There is now a
7832 variant, -bhc, which does actually run the callout code, including
7833 consulting and updating the callout cache.
7835 70. Added support for saslauthd authentication, courtesy of Alexander
7838 71. If statvfs() failed on the spool or log directories while checking their
7839 size for availability, Exim confusingly gave the error "space shortage".
7840 Furthermore, in debugging mode it crashed with a floating point exception.
7841 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7842 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7843 serious problem, Exim now writes to the main and panic logs when this
7844 happens, with details of the failure. It then refuses to accept the
7845 incoming message, giving the message "spool directory problem" or "log
7846 directory problem" with a 421 code for SMTP messages.
7848 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7849 0, 1, and 2 exist, because some OS complain for execs without them (see
7850 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7851 descriptors. However, the code omitted to check that the open succeeded,
7852 causing mysterious errors if for some reason the permissions on /dev/null
7853 got screwed. Now Exim writes a message to the main and panic logs, and
7854 bombs out if it can't open /dev/null.
7856 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7857 interact so that it is all more flexible. It is supposed to remain
7858 backwards compatible. Also added extra_local_interfaces.
7860 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7861 to bomb out with an assertion failure - to the client this appears as a
7862 connection drop. This problem occurs in the part of the code that was taken
7863 from the Samba project. Fortunately, the assertion is in a very simple
7864 function, so I have fixed this by reproducing the function inline in the
7865 one place where it is called, and arranging for authentication to fail
7866 instead of killing the process with assert().
7868 75. The SPA client code was not working when the server requested OEM rather
7869 than Unicode encoding.
7871 76. Added code to make require_files with a specific uid setting more usable in
7872 the case where statting the file as root fails - usually a non-root-mounted
7873 NFS file system. When this happens and the failure is EACCES, Exim now
7874 forks a subprocess and does the per-uid checking as the relevant uid.
7876 77. Added process_log_path.
7878 78. If log_file_path was not explicitly set, a setting of check_log_space or
7879 check_log_inodes was ignored.
7881 79. If a space check for the spool or log partitions fails, the incident is now
7882 logged. Of course, in the latter case the data may get lost...
7884 80. Added the %p formatting code to string_format() so that it can be used to
7885 print addresses in debug_print(). Adjusted all the address printing in the
7886 debugging in store.c to use %p rather than %d.
7888 81. There was a concern that a line of code in smtp_in.c could overflow a
7889 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7890 initially expressed, the concern was not well-founded, because trailing
7891 spaces are removed early. However, if the trailing spaces were followed by
7892 a NULL, they did not get removed, so the overflow was possible. Two fixes
7895 (a) I re-wrote the offending code in a cleaner fashion.
7896 (b) If an incoming SMTP command contains a NULL character, it is rejected
7899 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7900 runs initgroups(), so that if the Exim user is in any additional groups,
7901 they will be used during message reception.
7907 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob