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.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
126 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
127 a remote which uses callout/hold. Previously the local was queued.
129 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
132 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
134 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
135 authenticator client_send option. Previously the next char, after a pair
136 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
137 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
138 documentation. There is still no way to get a leading ^ immediately
139 after a NUL (ie. for the password of a PLAIN method authenticator.
141 JH/39 Bug 2691: fix $local_part_data. When the matching list element
142 referred to a file, bad data was returned. This likely also affected
145 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
146 left undeleted; the attempt to re-create it then failed - resulting in
147 the usual "SIGHUP tp have daemon reload configuration" to not work.
148 This affected any platform not supporting "abstract" Unix-domain
149 sockets (i.e. not Linux).
151 JH/42 Bug 2692: Harden against a peer which reneges on a 452 "too many
152 recipients" response to RCPT in a later response, with a 250. The
153 previous coding assumed this would not happen, and under PIPELINING
154 would result in both lost and duplicate recipients for a message.
156 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
157 Previously the weighting was incorrectly applied. Similar fix for socks
158 proxies. Found and fixed by Heiko Schlichting.
160 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
161 not handle sub-lists included using the +namedlist syntax. While
162 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
163 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
166 HS/01 Enforce absolute PID file path name.
168 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
174 JH/01 Avoid costly startup code when not strictly needed. This reduces time
175 for some exim process initialisations. It does mean that the logging
176 of TLS configuration problems is only done for the daemon startup.
178 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
180 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
181 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
184 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
185 DKIM signing is being done. Previously a transport_filter would always
186 disable CHUNKING, falling back to traditional DATA.
188 JH/05 Regard command-line receipients as tainted.
190 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
192 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
193 PAM library frees one of the arguments given to it, despite the
194 documentation. Therefore a plain malloc must be used.
196 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
197 on-stack buffers were used, resulting in a taint trap when DSN information
198 copied from a received message was written into the buffer.
200 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
201 the ordering of its ARC headers. This caused a crash.
203 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
204 a new record was being constructed with information from the peer, a trap
207 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
208 installation would get error messages from DMARC verify, when it hit the
209 nonexistent file indicated by the default. Distros wanting DMARC enabled
210 should both provide the file and set the option.
211 Also enforce no DMARC verification for command-line sourced messages.
213 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
214 could, depending on the platform, hang at the STARTTLS response.
216 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
217 message on a connection. Previously if one message had ARC headers and
218 the following one did not, a crash could result when adding an
219 Authentication-Results: header.
221 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
222 Exim main code and Exim-related utities. The introduction of taint
223 tracking also did many adjustments to string handling. Since then, eximon
224 frequently terminated with an assert failure.
226 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
227 check for 452 responses. This slightly helps the inefficieny of doing
228 a large alias-expansion into a recipient-limited target. The max_rcpt
229 transport option still applies (and at the current default, will override
230 the new feature). The check is done for either cause of synch, and forces
231 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
232 connection. The new facility is not tunable at this time.
234 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
235 library live data was being used, so the results became garbage. Make
236 copies while it is still usable.
238 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
239 on delivery deferred (==) and failed (**) lines (if a delivery was
240 attemtped). Previously it was only on completion (=>) lines.
242 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
243 for the expansion of the server_scram_iter and server_scram_salt options.
245 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
246 are now specifically given a NO_DATA response without hitting the system
247 resolver. The library goes on to do the now-standard TXT lookup.
248 Use of dnsdb lookups is not affected.
250 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
251 only retrieve the errormessage once. Previously two calls to dlerror()
252 were used, and the second one (for mainlog/paniclog) retrieved null
255 JH/20 Taint checking: disallow use of tainted data for
256 - the appendfile transport file and directory options
257 - the pipe transport command
258 - the autoreply transport file, log and once options
259 - file names used by the redirect router (including filter files)
261 - paths used by single-key lookups
262 Previously this was permitted.
264 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
265 adjusted the size of a major service buffer; this failed because the
266 buffer was in use at the time. Change to a compile-time increase in the
267 buffer size, when this authenticator is compiled into exim.
269 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
270 previous fast-mode was untenable in the face of glibs using mmap to
271 support larger malloc requests.
273 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
274 New values supported, if defined on system where compiled:
275 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
276 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
278 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
279 running a limited number of proceses in parallel, a benefit is gained. The
280 amount varies with the platform hardware and load. The use of the option
281 queue_run_in_order means we cannot do this, as ordering becomes
284 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
285 had introduced a string-copy (for ensuring NUL-termination) which was not
286 appropriate for that case, which can include embedded NUL bytes in the
287 block of data. Investigation showed the copy to actually be needless, the
288 data being length-specified.
290 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
291 done during a receiving connection, and both used TLS, global info was
292 used rather than per-connection info for tracking the state of data
293 queued for transmission. This could result in a connection hang.
295 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
296 Previously, when delivering serveral messages down a single connection
297 only the first would provide a SIZE. This was due to the size information
298 not being properly tracked.
300 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
301 TAI (at 37 seconds currently), pretend to be in UTC for time-related
302 expansion and logging. Previously, spurious values such as a future
303 minute could be seen.
305 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
306 it could crash from a null-deref. This could also affect the
307 ${addresses: } operator and ${readsock } item.
309 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
310 message following a mime one, the variable was not reset.
312 JH/30 When an pipelined-connect fails at the first response, assume incorrect
313 cached capability (perhaps the peer reneged?) and immediately retry in
316 JH/31 Fix spurious detection of timeout while writing to transport filter.
318 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
319 an attempt to copy the string was made before checking it.
321 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
322 taint of the lookup key was maintained; we now regard the presence in the
323 filesystem as sufficient validation.
325 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
326 argument is supplied.
328 JH/35 The dsearch lookup now requires that the directory is an absolute path.
329 Previously this was not checked, and nonempty relative paths made an
330 access under Exim's current working directory.
332 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
333 Previously no event was raised.
335 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
336 parameter supplied by the sender MAIL FROM command. Previously it was
337 ignored, and only the check_spool_space option value for the required
340 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
341 the size of the signing public-key. Previously it was instead giving
342 the size of the signature hash.
344 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
345 the default. See the (new) dkim_verify_min_keysizes option.
347 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
348 and an ACL use a lookup for checking either the local_part or domain,
349 stale data could be accessed. Ensure that variable references are
350 dropped between messages.
352 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
353 by the client was not checked as pointing within response data before
354 being used. A malicious client could thus cause an out-of-bounds read and
355 possibly gain authentication. Fix by adding the check.
357 JH/42 Internationalisation: change the default for downconversion in the smtp
358 transport to be "if needed". Previously it was "as previously set" for
359 the message, which usually meant "if needed" for message-submission but
360 "no" for everything else. However, MTAs have been seen using SMTPUTF8
361 even when the envelope addresses did not need it, resulting in forwarding
362 failures to non-supporting MTAs. A downconvert in such cases will be
363 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
364 transport. The change does mean that addresses needing conversion will
365 be converted when previously a delivery failure would occur.
367 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
368 response was received it would be used unchecked in a fail-DSN, violating
369 standards on line-length limits. Truncate if needed.
371 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
372 doesn't work. (Additionally add a new main config option to configure the
379 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
380 side, to decode a TLS 1.3 packet capture.
382 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
383 Previously the default library behaviour applied, sending two, each in
386 JH/03 Debug output for ACL now gives the config file name and line number for
389 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
391 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
393 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
394 buffer overrun for (non-chunking) other transports.
396 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
397 TLS1.3, means that a server rejecting a client certificate is not visible
398 to the client until the first read of encrypted data (typically the
399 response to EHLO). Add detection for that case and treat it as a failed
400 TLS connection attempt, so that the normal retry-in-clear can work (if
401 suitably configured).
403 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
404 and/or domain. Found and fixed by Jason Betts.
406 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
407 configuration). If a CNAME target was not a wellformed name pattern, a
410 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
411 the OS reports them interleaved with other addresses.
413 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
414 used both for input and for a verify callout, both encrypted, SMTP
415 responses being sent by the server could be lost. This resulted in
416 dropped connections and sometimes bounces generated by a peer sending
419 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
420 string. Previously it was possible to cause undefined behaviour in a
421 library routine (usually a crash). Found by "zerons".
423 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
426 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
427 API was removed, so update to use the newer ones.
429 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
430 any timeout set, is taking a long time. Previously we would hang on to a
431 rotated logfile "forever" if the input was arriving with long gaps
432 (a previous attempt to fix addressed lack, for a long time, of initial
435 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
436 shared (NFS) environment. The length of the tempfile name is now
437 4 + 16 ("hdr.$message_exim_id") which might break on file
438 systems which restrict the file name length to lower values.
441 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
442 shared (NFS) environment.
444 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
445 did for all versions <4.90). Notably -M, -m, --invert, -I may be
448 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
449 on some platforms for bit 31.
451 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
452 to changes apparently associated with TLS1.3 handling some of the APIs
453 previously used were either nonfunctional or inappropriate. Strings
454 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
455 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
456 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
457 This affects log line X= elements, the $tls_{in,out}_cipher variables,
458 and the use of specific cipher names in the encrypted= ACL condition.
460 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
462 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
463 verification result was not updated unless hosts_require_ocsp applied.
465 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
466 queue_list_requires_admin set to false, non-admin users were denied the
469 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
470 directory-of-certs mode. Previously they were advertised despite the
473 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
474 A single TCP connection by a client will now hold a TLS connection open
475 for multiple message deliveries, by default. Previously the default was to
478 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
479 default. If built with the facility, DANE will be used. The facility
480 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
482 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
483 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
484 must be defined and you must still, unless you define DISABLE_TLS, manage
485 the the include-dir and library-file requirements that go with that
486 choice. Non-TLS builds are still supported.
488 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
491 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
492 default. If the platform supports and has the facility enabled, it will
493 be requested on all coneections.
495 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
496 controlled by the build-time option SUPPORT_PIPE_CONNECT.
498 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
500 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
501 success-DSN messages. Previously the From: header was always the default
502 one for these; the option was ignored.
504 JH/28 Fix the timeout on smtp response to apply to the whole response.
505 Previously it was reset for every read, so a teergrubing peer sending
506 single bytes within the time limit could extend the connection for a
507 long time. Credit to Qualsys Security Advisory Team for the discovery.
509 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
510 delivery address, which leaked information of the results of local
511 forwarding. Change to the original envelope recipient address, per
514 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
515 requested. Previously not bounce was generated and a log entry of
516 error ignored was made.
518 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
520 JH/32 Introduce a general tainting mechanism for values read from the input
521 channel, and values derived from them. Refuse to expand any tainted
522 values, to catch one form of exploit.
524 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
525 was unused and the unexpanded text used for the test. Found and
526 fixed by Ruben Jenster.
528 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
529 an attempt to use a TLS library read routine dereffed a nul pointer,
532 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
533 them in our smtp response.
535 JH/36 Have the general router option retry_use_local_part default to true when
536 any of the restrictive preconditions are set (to anything). Previously it
537 was only for check_local user. The change removes one item of manual
538 configuration which is required for proper retries when a remote router
539 handles a subset of addresses for a domain.
541 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
542 link count into consideration.
544 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
545 caused the extension of big_buffer, the following lines were ignored.
547 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
548 accordance with RFC 2308. Previously there was no expiry, so a longlived
549 receive process (eg. due to ACL delays) versus a short SOA value could
552 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
554 JH/39 Promote DMARC support to mainline.
556 JH/40 Bug 2452: Add a References: header to DSNs.
558 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
559 parameters. The relevant library call is documented as "Deprecated: This
560 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
561 3.6.0, DH parameters are negotiated following RFC7919."
563 HS/06 Change the default of dnssec_request_domains to "*"
565 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
566 carried on and emitted a BDAT command, even when PIPELINING was not
569 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
570 buffer was used for the filename, resulting in a trap when tainted
571 arguments (eg. $domain) were used.
573 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
574 recommended to avoid a possible server-load attack. The feature can be
575 re-enabled via the openssl_options main cofiguration option.
577 JH/45 local_scan API: documented the current smtp_printf() call. This changed
578 for version 4.90 - adding a "more data" boolean to the arguments.
579 Bumped the ABI version number also, this having been missed previously;
580 release versions 4.90 to 4.92.3 inclusive were effectively broken in
581 respect of usage of smtp_printf() by either local_scan code or libraries
582 accessed via the ${dlfunc } expansion item. Both will need coding
583 adjustment for any calls to smtp_printf() to match the new function
584 signature; a FALSE value for the new argument is always safe.
586 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
587 the file-offset (which the Linux syscall does, and exim expects); this
588 resulted in an indefinite loop.
590 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
591 to do ARC verification. The Authentication-Results: header line added
592 by the configuration then had no ARC item.
598 JH/01 Remove code calling the customisable local_scan function, unless a new
599 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
601 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
602 non-signal-safe functions being used.
604 JH/03 Bug 2269: When presented with a received message having a stupidly large
605 number of DKIM-Signature headers, disable DKIM verification to avoid
606 a resource-consumption attack. The limit is set at twenty.
608 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
609 report of oldest_pass in ${authres } in consequence, and separate out
610 some descriptions of reasons for verification fail.
612 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
613 files in the spool were present and unlocked. A queue-runner could spot
614 them, resulting in a duplicate delivery. Fix that by doing the unlock
615 after the unlink. Investigation by Tim Stewart. Take the opportunity to
616 add more error-checking on spoolfile handling while that code is being
619 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
620 No known attacks, no CVE, this is defensive hardening.
622 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
623 a queue-runner could start a delivery while other operations were ongoing.
624 Cutthrough delivery was a common victim, resulting in duplicate delivery.
625 Found and investigated by Tim Stewart. Fix by using the open message data
626 file handle rather than opening another, and not locally closing it (which
627 releases a lock) for that case, while creating the temporary .eml format
628 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
630 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
631 $sender_verify_failure/$recipient_verify_failure to "random".
633 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
636 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
637 Previously this would segfault.
639 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
642 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
643 like zero, since the resolver should be doing this for us, But we need one
644 as a CNAME but no MX presence gets the CNAME returned; we need to check
645 that doesn't point to an MX to declare it "no MX returned" rather than
646 "error, loop". A new main option is added so the older capability of
647 following some limited number of chain links is maintained.
649 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
651 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
652 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
653 now-deprecated earlier definitions used only specified the range up to TLS
654 1.2 (in the older-version library docs).
656 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
658 JH/15 Rework TLS client-side context management. Stop using a global, and
659 explicitly pass a context around. This enables future use of TLS for
660 connections to service-daemons (eg. malware scanning) while a client smtp
661 connection is using TLS; with cutthrough connections this is quite likely.
663 JH/16 Fix ARC verification to do AS checks in reverse order.
665 JH/17 Support a "tls" option on the ${readsocket } expansion item.
667 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
668 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
669 Previously the "utf8" would be re-prepended for every additional message.
671 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
672 Previously thery were accepted, resulting in issues when attempting to
673 forward messages to a non-supporting MTA.
675 PP/02 Let -n work with printing macros too, not just options.
677 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
678 one parent address was copied, and bogus data was used at delivery-logging
679 time. Either a crash (after delivery) or bogus log data could result.
680 Discovery and analysis by Tim Stewart.
682 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
683 Previously if the string ended mid-character, we did not insert the
684 promised '?' replacement.
686 PP/04 Documentation: current string operators work on bytes, not codepoints.
688 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
689 should pack well giving a smaller memory footprint so better caching and
690 therefore performance. Group the declarations where this can't be done so
691 that the byte-sized flag variables are not interspersed among pointer
692 variables, giving a better chance of good packing by the compiler.
694 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
695 non-null, to avoid issues with sites running BATV. Previously reports were
696 sent with an empty envelope sender so looked like bounces.
698 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
699 The ignore_error flag wasn't being returned from the filter subprocess so
700 was not set for later routers. Investigation and fix by Matthias Kurz.
702 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
703 and a msg:complete for the whole, when a message is manually removed using
704 -Mrm. Developement by Matthias Kurz, hacked on by JH.
706 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
707 a "Gnu special" function, asprintf() in the DB utility binary builds; I
708 hope that is portable enough.
710 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
711 requiring a known-CA anchor certificate; make it now rely entirely on the
712 TLSA as an anchor. Checking the name on the leaf cert against the name
713 on the A-record for the host is still done for TA (but not for EE mode).
715 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
716 included in delivery lines for non-proxied connections, when compiled with
717 SUPPORT_SOCKS and running with proxy logging enabled.
719 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
720 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
721 move the existing event to fire before the normal logging of message
722 failure so that custom logging is bracketed by normal logging.
724 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
725 msg:fail:internal event. Developement by Matthias Kurz.
727 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
728 far too small for todays use of crypto signatures stored there. Go all
729 the way to the max DNS message size of 64kB, even though this might be
730 overmuch for IOT constrained device use.
732 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
733 copy a string over itself. The library routine is documented as not
734 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
736 JH/32 For main options check_spool_space and check_inode_space, where the
737 platform supports 64b integers, support more than the previous 2^31 kB
738 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
739 the previous G, M, k.
741 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
742 $authenticated_fail_id variable on authentication failure. Previously
745 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
746 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
747 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
748 GNUTLS_SEC_PARAM_MEDIUM.
750 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
751 side. Previously we would continue as if no SNI had been received.
753 JH/36 Harden the handling of string-lists. When a list consisted of a sole
754 "<" character, which should be a list-separator specification, we walked
755 off past the nul-terimation.
757 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
758 causes) even when the retry time is not yet met. Previously they were
759 not, meaning that when (say) an account was over-quota and temp-rejecting,
760 and multiple senders' messages were queued, only one sender would get
761 notified on each configured delay_warning cycle.
763 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
765 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
766 cipher-suites, an error can be left on the stack even for a succeeding
767 accept; this results in impossible error messages when a later operation
770 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
771 return error codes indicating retry. Under TLS1.3 this becomes required.
773 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
774 it only wrote the new authenticators, resulting in a lack of tracking of
775 peer changes of ESMTP extensions until the next cache flush.
777 JH/41 Fix the loop reading a message header line to check for integer overflow,
778 and more-often against header_maxsize. Previously a crafted message could
779 induce a crash of the recive process; now the message is cleanly rejected.
781 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
782 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
788 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
789 When redis_servers is set to a list of > 1 element, and the Redis servers
790 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
791 case of MOVED into a DEFER case instead, thus moving the query onto the
792 next server in the list. For a cluster of N elements, all N servers must
793 be defined in redis_servers.
795 GF/02 Catch and remove uninitialized value warning in exiqsumm
796 Check for existence of @ARGV before looking at $ARGV[0]
798 JH/01 Replace the store_release() internal interface with store_newblock(),
799 which internalises the check required to safely use the old one, plus
800 the allocate and data copy operations duplicated in both (!) of the
801 extant use locations.
803 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
804 modifier. This matches the restriction on the commandline.
806 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
807 Previously only the last row was returned.
809 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
810 we assumed that tags in the header were well-formed, and parsed the
811 element content after inspecting only the first char of the tag.
812 Assumptions at that stage could crash the receive process on malformed
815 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
816 While running the DKIM ACL we operate on the Permanent memory pool so that
817 variables created with "set" persist to the DATA ACL. Also (at any time)
818 DNS lookups that fail create cache records using the Permanent pool. But
819 expansions release any allocations made on the current pool - so a dnsdb
820 lookup expansion done in the DKIM ACL releases the memory used for the
821 DNS negative-cache, and bad things result. Solution is to switch to the
822 Main pool for expansions.
823 While we're in that code, add checks on the DNS cache during store_reset,
824 active in the testsuite.
825 Problem spotted, and debugging aided, by Wolfgang Breyha.
827 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
828 When none of the hosts presented to a transport match an already-open
829 connection, close it and proceed with the list. Previously we would
830 queue the message. Spotted by Lena with Yahoo, probably involving
833 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
834 Previously a spurious "250 OK id=" response was appended to the proper
837 JH/08 The "support for" informational output now, which built with Content
838 Scanning support, has a line for the malware scanner interfaces compiled
839 in. Interface can be individually included or not at build time.
841 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
842 by the template makefile "src/EDITME". The "STREAM" support for an older
843 ClamAV interface method is removed.
845 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
846 rows affected is given instead).
848 JH/11 The runtime Berkeley DB library version is now additionally output by
849 "exim -d -bV". Previously only the compile-time version was shown.
851 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
852 SMTP connection. Previously, when one had more receipients than the
853 first, an abortive onward connection was made. Move to full support for
854 multiple onward connections in sequence, handling cutthrough connection
855 for all multi-message initiating connections.
857 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
858 routers. Previously, a multi-recipient message would fail to match the
859 onward-connection opened for the first recipient, and cause its closure.
861 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
862 a timeout on read on a GnuTLS initiating connection, resulting in the
863 initiating connection being dropped. This mattered most when the callout
864 was marked defer_ok. Fix to keep the two timeout-detection methods
867 JH/15 Relax results from ACL control request to enable cutthrough, in
868 unsupported situations, from error to silently (except under debug)
869 ignoring. This covers use with PRDR, frozen messages, queue-only and
872 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
874 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
875 metadata, resulting in a crash in free().
877 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
878 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
879 Broken also in d185889f4, with init system revamp.
881 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
882 Previously we abruptly closed the connection after reading a malware-
883 found indication; now we go on to read the "scan ok" response line,
886 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
887 ACL. Previously, a crash would result.
889 JH/19 Speed up macro lookups during configuration file read, by skipping non-
890 macro text after a replacement (previously it was only once per line) and
891 by skipping builtin macros when searching for an uppercase lead character.
893 JH/20 DANE support moved from Experimental to mainline. The Makefile control
894 for the build is renamed.
896 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
897 was allocated for every new TLS startup, meaning one per message. Fix
898 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
900 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
901 reported the original. Fix to report (as far as possible) the ACL
902 result replacing the original.
904 JH/23 Fix memory leak during multi-message connections using STARTTLS under
905 OpenSSL. Certificate information is loaded for every new TLS startup,
906 and the resources needed to be freed.
908 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
910 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
913 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
914 DATA response info to the (existing) per-recipient response info for
915 the "C=" log element. It can have useful tracking info from the
916 destination system. Patch from Simon Arlott.
918 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
919 length value. Previously this would segfault.
921 HS/02 Support Avast multiline protoocol, this allows passing flags to
922 newer versions of the scanner.
924 JH/28 Ensure that variables possibly set during message acceptance are marked
925 dead before release of memory in the daemon loop. This stops complaints
926 about them when the debug_store option is enabled. Discovered specifically
927 for sender_rate_period, but applies to a whole set of variables.
928 Do the same for the queue-runner and queue-list loops, for variables set
929 from spool message files. Do the same for the SMTP per-message loop, for
930 certain variables indirectly set in ACL operations.
932 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
933 as a multi-recipient message from a mailinglist manager). The coding had
934 an arbitrary cutoff number of characters while checking for more input;
935 enforced by writing a NUL into the buffer. This corrupted long / fast
936 input. The problem was exposed more widely when more pipelineing of SMTP
937 responses was introduced, and one Exim system was feeding another.
938 The symptom is log complaints of SMTP syntax error (NUL chars) on the
939 receiving system, and refused recipients seen by the sending system
940 (propating to people being dropped from mailing lists).
941 Discovered and pinpointed by David Carter.
943 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
944 replaced by the ${authresults } expansion.
946 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
948 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
949 allows proper process termination in container environments.
951 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
952 Previously the "final dot" had a newline after it; ensure it is CR,LF.
954 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
955 and "err_perm", deprecated since 4.83 when the RFC-defined words
956 "temperror" and "permerror" were introduced.
958 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
959 transport-filters or DKIM-signing. The restriction was lost in the
960 consolidation of verify-callout and delivery SMTP handling.
961 Extend the restriction to also cover ARC-signing.
963 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
964 in defer=pass mode supply a 450 to the initiator. Previously the message
967 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
968 tls_require_ciphers is used as before.
970 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
971 "pass_unscanned". Only tmpfails from the scanner are written to
972 the paniclog, as they may require admin intervention (permission
973 denied, license issues). Other scanner errors (like decompression
974 bombs) do not cause a paniclog entry.
976 JH/36 Fix reinitialisation of DKIM logging variable between messages.
977 Previously it was possible to log spurious information in receive log
980 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
981 triggered odd behaviour from Outlook Express clients.
983 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
986 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
987 since the IETF WG has not yet settled on that versus the original
988 "bare" representation.
990 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
991 Previously the millisecond value corrupted the output.
992 Fix also for syslog_pid=no and log_selector +pid, for which the pid
993 corrupted the output.
999 JH/01 Rework error string handling in TLS interface so that the caller in
1000 more cases is responsible for logging. This permits library-sourced
1001 string to be attached to addresses during delivery, and collapses
1002 pairs of long lines into single ones.
1004 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1005 during configuration. Wildcards are allowed and expanded.
1007 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1008 This permits better logging.
1010 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1011 do not close it down and have the child transport start it up again on
1012 the passed-on TCP connection. Instead, proxy the child (and any
1013 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1014 affected: the continued delivery log lines do not have any DNSSEC, TLS
1015 Certificate or OCSP information. TLS cipher information is still logged.
1017 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1018 identical IP addresses on different listening ports. Will also affect
1021 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1022 add noisy ifdef guards to special-case this sillyness.
1023 Patch from Bernd Kuhls.
1025 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1026 than 255 are no longer allowed.
1028 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1029 Disable the session-cache too, which might reduce our load. Since we
1030 currrectly use a new context for every connection, both as server and
1031 client, there is no benefit for these.
1032 GnuTLS appears to not support tickets server-side by default (we don't
1033 call gnutls_session_ticket_enable_server()) but client side is enabled
1034 by default on recent versions (3.1.3 +) unless the PFS priority string
1037 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1038 <https://reproducible-builds.org/specs/source-date-epoch/>.
1040 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1041 the check for any unsuccessful recipients did not notice the limit, and
1042 erroneously found still-pending ones.
1044 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1045 MSG_MORE. Only in-clear (not on TLS connections).
1047 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1048 transport-filter is involved we can buffer the headers in memory for
1049 creating the signature, and read the spool data file once for the
1050 signature and again for transmission.
1052 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1053 4.77 as the kernel support then wasn't solid, having issues in 64bit
1054 mode. Now, it's been long enough. Add support for FreeBSD also.
1056 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1057 case where the routing stage had gathered several addresses to send to
1058 a host before calling the transport for the first, we previously failed
1059 to close down TLS in the old transport process before passing the TCP
1060 connection to the new process. The new one sent a STARTTLS command
1061 which naturally failed, giving a failed delivery and bloating the retry
1062 database. Investigation and fix prototype from Wolfgang Breyha.
1064 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1065 false-negatives in the check that the sender had not preempted a response
1066 or prompt from Exim (running as a server), due to that code's lack of
1067 awareness of the SMTP input buffering.
1069 PP/04 Add commandline_checks_require_admin option.
1070 Exim drops privileges sanely, various checks such as -be aren't a
1071 security problem, as long as you trust local users with access to their
1072 own account. When invoked by services which pass untrusted data to
1073 Exim, this might be an issue. Set this option in main configuration
1074 AND make fixes to the calling application, such as using `--` to stop
1077 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1078 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1079 responses to those, into a single TLS record each way (this usually means
1080 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1083 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1084 affects you only if you're dancing at the edge of the param size limits.
1085 If you are, and this message makes sense to you, then: raise the
1086 configured limit or use OpenSSL 1.1. Nothing we can do for older
1089 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1090 cmdline element to get the documented default one. Previously it was
1093 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1094 are unsuable for non-message-specific reasons - eg. connection timeout,
1095 banner-time rejection.
1097 JH/16 Fix logging of delivery remote port, when specified by router, under
1100 PP/06 Repair manualroute's ability to take options in any order, even if one
1101 is the name of a transport.
1104 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1106 JH/17 Change the list-building routines interface to use the expanding-string
1107 triplet model, for better allocation and copying behaviour.
1109 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1110 Previously it was constructed the first time a possibly-matching string
1111 was met in the configuration file input during startup; now it is done
1114 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1115 compatible one, to avoid the (poorly documented) possibility of a config
1116 file in the working directory redirecting the DB files, possibly correpting
1117 some existing file. CVE-2017-10140 assigned for BDB.
1119 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1120 cache-hot. Previously, although the result was properly cached, the
1121 initial verify call returned a defer.
1123 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1124 the main verify for receipient in uncached-mode.
1126 JH/22 Retire historical build files to an "unsupported" subdir. These are
1127 defined as "ones for which we have no current evidence of testing".
1129 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1130 if present. Previously it was ignored.
1132 JH/24 Start using specified-initialisers in C structure init coding. This is
1133 a C99 feature (it's 2017, so now considered safe).
1135 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1136 if was a fixed-sized field and bitmask ops via macros; it is now more
1139 PP/07 GitHub PR 56: Apply MariaDB build fix.
1140 Patch provided by Jaroslav Škarvada.
1142 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1143 during Coverity cleanups [4.87 JH/47]
1144 Diagnosis and fix provided by Michael Fischer v. Mollard.
1146 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1147 the right size to place the terminating semicolon on its own folded
1148 line, the header hash was calculated to an incorrect value thanks to
1149 the (relaxed) space the fold became.
1151 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1152 and confused the parent.
1154 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1155 which could crash as a result. This could lead to undeliverable messages.
1157 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1160 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1161 matters on fast-turnover and PID-randomising systems, which were getting
1162 out-of-order delivery.
1164 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1165 a possibly-overlapping copy. The symptom was that "Remote host closed
1166 connection in response to HELO" was logged instead of the actual 4xx
1169 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1170 Previously only that bufferd was discarded, resulting in SYMTP command
1173 JH/32 DKIM: when a message has multiple signatures matching an identity given
1174 in dkim_verify_signers, run the dkim acl once for each. Previously only
1175 one run was done. Bug 2189.
1177 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1178 "panic the current process" to "deliberately defer". The panic log is
1179 still written with the problem list name; the mail and reject logs now
1180 get a temp-reject line for the message that was being handled, saying
1181 something like "domains check lookup or other defer". The SMTP 451
1182 message is still "Temporary local problem".
1184 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1185 A crafted sequence of BDAT commands could result in in-use memory beeing
1186 freed. CVE-2017-16943.
1188 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1189 from SMTP input. Previously it was always done; now only done for DATA
1190 and not BDAT commands. CVE-2017-16944.
1192 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1193 to the message (such as an overlong header line). Previously this was
1194 not done and we did not exit BDAT mode. Followon from the previous item
1195 though a different problem.
1201 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1202 than -2003 did; needs libidn2 in addition to libidn.
1204 JH/02 The path option on a pipe transport is now expanded before use.
1206 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1207 Patch provided by "Björn", documentation fix added too.
1209 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1210 missing a wire-to-host endian conversion.
1212 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1213 close after a BDAT command line could be taken as a following command,
1214 giving a synch failure. Fix by only checking for synch immediately
1215 before acknowledging the chunk.
1217 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1218 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1219 macro. Patches provided by Josh Soref.
1221 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1222 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1223 by IANA as a service extension. However, John Klensin suggests that we
1226 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1227 the dkim code may be unix-mode line endings rather than smtp wire-format
1228 CRLF, so prepend a CR to any bare LF.
1230 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1231 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1233 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1234 the first were themselves being wrongly included in the feed into dkim
1235 processing; with most chunk sizes in use this resulted in an incorrect
1236 body hash calculated value.
1238 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1239 DKIM signature block, for verification. Although advised against by
1240 standards it is specifically not ruled illegal.
1242 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1244 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1245 missing a body hash (the bh= tag).
1247 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1248 It seems that HAProxy sends the Proxy Protocol information in clear and
1249 only then does a TLS startup, so do the same.
1251 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1252 TCP connections (such as for Spamd) unless the daemon successfully set
1253 Fast Open mode on its listening sockets. This fixes breakage seen on
1254 too-old kernels or those not configured for Fast Open, at the cost of
1255 requiring both directions being enabled for TFO, and TFO never being used
1256 by non-daemon-related Exim processes.
1258 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1259 endings, at least on the first header line. Try to canonify any that get
1260 past that check, despite the cost.
1262 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1263 now limited to an arbitrary five deep, while parsing addresses with the
1264 strip_excess_angle_brackets option enabled.
1266 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1267 instead leave the unprompted TLS handshake in socket buffer for the
1268 TLS library to consume.
1270 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1272 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1274 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1275 region in one operation, for speed, and this leaves assigned pointers
1276 dangling. Add checks run only under the testsuite which checks all
1277 variables at a store-reset and panics on a dangling pointer; add code
1278 explicitly nulling out all the variables discovered. Fixes one known
1279 bug: a transport crash, where a dangling pointer for $sending_ip_address
1280 originally assigned in a verify callout, is re-used.
1282 PP/06 Drop '.' from @INC in various Perl scripts.
1284 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1286 PP/08 Reduce a number of compilation warnings under clang; building with
1287 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1288 should be warning-free.
1290 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1292 HS/01 Fix portability problems introduced by PP/08 for platforms where
1293 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1295 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1296 chunk. This allows us to accept broken chunked messages. We need a more
1297 general solution here.
1299 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1300 already-broken messages in the queue.
1302 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1304 JH/19 Fix reference counting bug in routing-generated-address tracking.
1310 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1311 supports it and a size is available (ie. the sending peer gave us one).
1313 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1314 years of being deprecated). The replacements are the ACLs
1315 acl_smtp_mime and acl_not_smtp_mime.
1317 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1318 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1319 or even an in-clear connection were permitted. Now, if the host lookup
1320 was dnssec and dane was requested then the host is only used if the
1321 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1322 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1323 if one fails this test.
1324 This means that a poorly-configured remote DNS will make it incommunicado;
1325 but it protects against a DNS-interception attack on it.
1327 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1328 non-noisy when a race steals the message being considered.
1330 JH/05 If main configuration option tls_certificate is unset, generate a
1331 self-signed certificate for inbound TLS connections.
1333 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1334 in rewrites and routers.
1336 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1337 and logged a warning sing 4.83; now they are a configuration file error.
1339 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1340 (lacking @domain). Apply the same qualification processing as RCPT.
1342 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1344 JH/10 Support ${sha256:} applied to a string (as well as the previous
1347 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1348 a cutthrough deliver is pending, as we always want to make a connection.
1349 This also avoids re-routing the message when later placing the cutthrough
1350 connection after a verify cache hit.
1351 Do not update it with the verify result either.
1353 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1354 when routing results in more than one destination address.
1356 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1357 signing (which inhibits the cutthrough capability). Previously only
1358 the presence of an option was tested; now an expansion evaluating as
1359 empty is permissible (obviously it should depend only on data available
1360 when the cutthrough connection is made).
1362 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1363 the relevant preceding SMTP command did not note the pipelining mode.
1365 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1366 Previously they were not counted.
1368 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1369 as one having no matching records. Previously we deferred the message
1370 that needed the lookup.
1372 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1373 distinguished as "(=".
1375 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1376 for missing MX records. Previously it only worked for missing A records.
1378 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1380 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1381 after the data-go-ahead and data-ack. Patch from Jason Betts.
1383 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1384 even for a "none" policy. Patch from Tony Meyer.
1386 JH/22 Fix continued use of a connection for further deliveries. If a port was
1387 specified by a router, it must also match for the delivery to be
1390 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1391 When one of the recipients of a message was unsuitable for the connection
1392 (has no matching addresses), we lost track of needing to mark it
1393 deferred. As a result mail would be lost.
1395 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1397 JH/25 Decoding ACL controls is now done using a binary search; the source code
1398 takes up less space and should be simpler to maintain. Merge the ACL
1399 condition decode tables also, with similar effect.
1401 JH/26 Fix problem with one_time used on a redirect router which returned the
1402 parent address unchanged. A retry would see the parent address marked as
1403 delivered, so not attempt the (identical) child. As a result mail would
1406 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1407 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1408 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1411 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1412 Main config options check_{log,spool}_{inodes,space} are now
1413 100 inodes, 10MB unless set otherwise in the configuration.
1415 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1416 Previously it only applied to the main-section connection policy
1419 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1421 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1422 by me. Added RFC7919 DH primes as an alternative.
1424 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1425 are not in the system include path.
1427 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1428 GnuTLS, when a session startup failed (eg because the client disconnected)
1429 Exim did stdio operations after fclose. This was exposed by a recent
1430 change which nulled out the file handle after the fclose.
1432 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1433 signed directly by the cert-signing cert, rather than an intermediate
1434 OCSP-signing cert. This is the model used by LetsEncrypt.
1436 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1438 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1439 an incoming connection.
1441 HS/02 Bug 1802: Do not half-close the connection after sending a request
1444 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1445 fallback to "prime256v1".
1447 JH/34 SECURITY: Use proper copy of DATA command in error message.
1448 Could leak key material. Remotely exploitable. CVE-2016-9963.
1454 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1455 and 3.4.4 - once the server is enabled to respond to an OCSP request
1456 it does even when not requested, resulting in a stapling non-aware
1457 client dropping the TLS connection.
1459 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1460 support variable-length bit vectors. No functional change.
1462 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1463 The I= interface field on outgoing lines is now after the H= remote
1464 host field, same as incoming lines. There is a separate
1465 outgoing_interface log selector which allows you to disable the
1468 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1469 If not running log_selector +smtp_connection the mainlog would be held
1470 open indefinitely after a "too many connections" event, including to a
1471 deleted file after a log rotate. Leave the per net connection logging
1472 leaving it open for efficiency as that will be quickly detected by the
1473 check on the next write.
1475 HS/01 Bug 1671: Fix post transport crash.
1476 Processing the wait-<transport> messages could crash the delivery
1477 process if the message IDs didn't exist for some reason. When
1478 using 'split_spool_directory=yes' the construction of the spool
1479 file name failed already, exposing the same netto behaviour.
1481 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1482 mime_regex ACL conditions.
1484 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1485 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1486 to HELO, local diagnostic string.
1488 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1489 log line to debug. Even when configured with a tls authenticator many
1490 client connections are expected to not authenticate in this way, so
1491 an authenticate fail is not an error.
1493 HS/02 Add the Exim version string to the process info. This way exiwhat
1494 gives some more detail about the running daemon.
1496 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1497 matter for fast-change records such as DNSBLs.
1499 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1500 retry record, even if constant. There may be multiple transports with
1501 different interface settings and the retry behaviour needs to be kept
1504 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1506 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1508 JH/10 Bug 840: fix log_defer_output option of pipe transport
1510 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1511 hosts_randomize. This matters a lot when combined with mua_wrapper.
1513 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1514 ${quote_pgsql:<string>} operator.
1516 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1518 JH/14 Change header file name for freeradius-client. Relevant if compiling
1519 with Radius support; from the Gentoo tree and checked under Fedora.
1521 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1523 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1524 option transport_filter, meaning no filtering.
1526 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1528 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1529 defaults to "*" (all hosts). The variable is now available when not built
1530 with TLS, default unset, mainly to enable keeping the testsuite sane.
1531 If a server certificate is not supplied (via tls_certificate) an error is
1532 logged, and clients will find TLS connections fail on startup. Presumably
1533 they will retry in-clear.
1534 Packagers of Exim are strongly encouraged to create a server certificate
1535 at installation time.
1537 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1538 with the $config_file variable.
1540 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1541 in transport context, after the attempt, and per-recipient. The latter type
1542 is per host attempted. The event data is the error message, and the errno
1543 information encodes the lookup type (A vs. MX) used for the (first) host,
1544 and the trailing two digits of the smtp 4xx response.
1546 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1547 to write to mainlog (or rejectlog, paniclog) in the window between file
1548 creation and permissions/ownership being changed. Particularly affects
1549 installations where exicyclog is run as root, rather than exim user;
1550 result is that the running daemon panics and dies.
1552 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1554 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1555 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1556 "pri" and "weight". Note that the previous implicit priority given by the
1557 list order is no longer honoured.
1559 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1560 for DKIM processing.
1562 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1563 by defining SUPPORT_SOCKS.
1565 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1566 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1567 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1568 variables are renamed to proxy_{local,external}_{address,port}.
1570 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1571 for a build by defining SUPPORT_I18N
1573 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1574 of the query string, and make ${quote_redis:} do that quoting.
1576 JH/29 Move Events support from Experimental to mainline, enabled by default
1577 and removable for a build by defining DISABLE_EVENT.
1579 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1581 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1582 cached by the daemon.
1584 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1585 by defining LOOKUP_REDIS. The libhiredis library is required.
1587 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1588 keys are given for lookup.
1590 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1591 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1592 only supported when built with TLS support. The PolarSSL SHA routines
1593 are still used when the TLS library is too old for convenient support.
1595 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1596 openssl_options), for security. OpenSSL forces this from version 1.1.0
1597 server-side so match that on older versions.
1599 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1600 allocation for $value could be released as the expansion processing
1601 concluded, but leaving the global pointer active for it.
1603 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1604 and to use the domains and local_parts ACL conditions.
1606 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1607 incorrectly not doubled on cutthrough transmission, hence seen as a
1608 body-termination at the receiving system - resulting in truncated mails.
1609 Commonly the sender saw a TCP-level error, and retransmitted the message
1610 via the normal store-and-forward channel. This could result in duplicates
1611 received - but deduplicating mailstores were liable to retain only the
1612 initial truncated version.
1614 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1616 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1618 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1619 we're in there, support oversigning also; bug 1309.
1621 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1623 HS/04 Add support for keep_environment and add_environment options.
1625 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1626 either intentional arithmetic overflow during PRNG, or testing config-
1629 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1630 delivery resulted in actual delivery. Cancel cutthrough before DATA
1633 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1634 rejects a recipient: pass the reject to the originator.
1636 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1637 Many were false-positives and ignorable, but it's worth fixing the
1640 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1641 for the new environment-manipulation done at startup. Move the routines
1642 from being local to tls.c to being global via the os.c file.
1644 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1645 an extract embedded as result-arg for a map, the first arg for extract
1646 is unavailable so we cannot tell if this is a numbered or keyed
1647 extraction. Accept either.
1653 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1656 JH/02 The smtp transport option "multi_domain" is now expanded.
1658 JH/03 The smtp transport now requests PRDR by default, if the server offers
1661 JH/04 Certificate name checking on server certificates, when exim is a client,
1662 is now done by default. The transport option tls_verify_cert_hostnames
1663 can be used to disable this per-host. The build option
1664 EXPERIMENTAL_CERTNAMES is withdrawn.
1666 JH/05 The value of the tls_verify_certificates smtp transport and main options
1667 default to the word "system" to access the system default CA bundle.
1668 For GnuTLS, only version 3.0.20 or later.
1670 JH/06 Verification of the server certificate for a TLS connection is now tried
1671 (but not required) by default. The verification status is now logged by
1672 default, for both outbound TLS and client-certificate supplying inbound
1675 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1678 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1679 Status Notification (bounce) messages are now MIME format per RFC 3464.
1680 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1681 under the control of the dsn_advertise_hosts option, and routers may
1682 have a dsn_lasthop option.
1684 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1685 default, modifiable by a malware= option. The list separator for
1686 the options can now be changed in the usual way. Bug 68.
1688 JH/10 The smtp_receive_timeout main option is now expanded before use.
1690 JH/11 The incoming_interface log option now also enables logging of the
1691 local interface on delivery outgoing connections.
1693 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1694 if the interface and destination host and port all match.
1696 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1699 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1700 Patch from Andrew Lewis.
1702 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1703 now supports optional time-restrictions, weighting, and priority
1704 modifiers per server. Patch originally by <rommer@active.by>.
1706 JH/16 The spamd_address main option now supports a mixed list of local
1707 and remote servers. Remote servers can be IPv6 addresses, and
1708 specify a port-range.
1710 JH/17 Bug 68: The spamd_address main option now supports an optional
1711 timeout value per server.
1713 JH/18 Bug 1581: Router and transport options headers_add/remove can
1714 now have the list separator specified.
1716 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1719 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1722 JH/21 Support for the A6 type of dns record is withdrawn.
1724 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1725 rather than the verbs used.
1727 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1728 from 255 to 1024 chars.
1730 JH/24 Verification callouts now attempt to use TLS by default.
1732 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1733 are generic router options now. The defaults didn't change.
1735 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1736 Original patch from Alexander Shikoff, worked over by JH.
1738 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1739 files if ZCAT_COMMAND is not executable.
1741 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1743 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1745 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1746 Normally benign, it bites when the pair was led to by a CNAME;
1747 modern usage is to not canonicalize the domain to a CNAME target
1748 (and we were inconsistent anyway for A-only vs AAAA+A).
1750 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1752 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1753 when evaluating $sender_host_dnssec.
1755 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1756 $sender_helo_dnssec variable.
1758 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1760 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1762 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1764 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1765 documented as working, but never had. Support all but $spam_report.
1767 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1768 added for tls authenticator.
1770 HS/03 Add perl_taintmode main config option
1776 TL/01 When running the test suite, the README says that variables such as
1777 no_msglog_check are global and can be placed anywhere in a specific
1778 test's script, however it was observed that placement needed to be near
1779 the beginning for it to behave that way. Changed the runtest perl
1780 script to read through the entire script once to detect and set these
1781 variables, reset to the beginning of the script, and then run through
1782 the script parsing/test process like normal.
1784 TL/02 The BSD's have an arc4random API. One of the functions to induce
1785 adding randomness was arc4random_stir(), but it has been removed in
1786 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1787 function when detected.
1789 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1790 cause callback expansion.
1792 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1793 syntax errors in an expansion can be treated as a string instead of
1794 logging or causing an error, due to the internal use of bool_lax
1795 instead of bool when processing it.
1797 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1798 server certificates when making smtp deliveries.
1800 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1802 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1804 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1806 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1807 Merged patch from Sebastian Wiedenroth.
1809 JH/05 Fix results-pipe from transport process. Several recipients, combined
1810 with certificate use, exposed issues where response data items split
1811 over buffer boundaries were not parsed properly. This eventually
1812 resulted in duplicates being sent. This issue only became common enough
1813 to notice due to the introduction of connection certificate information,
1814 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1816 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1817 size buffer was used, resulting in syntax errors when an expansion
1820 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1821 version 3.3.6 or later.
1823 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1824 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1825 both become "event_action", the variables become $event_name, $event_data
1826 and $event_defer_errno. There is a new variable $verify_mode, usable in
1827 routers, transports and related events. The tls:cert event is now also
1828 raised for inbound connections, if the main configuration event_action
1831 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1832 early OCSP support, but no stapling (appears to be less than 1.0.0).
1834 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1835 server certificate names available under the smtp transport option
1836 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1839 JH/10 Time-related extraction expansions from certificates now use the main
1840 option "timezone" setting for output formatting, and are consistent
1841 between OpenSSL and GnuTLS compilations. Bug 1541.
1843 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1844 encoded parameter in the incoming message. Bug 1558.
1846 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1847 include certificate info, eximon was claiming there were spoolfile
1850 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1852 JH/14 Log delivery-related information more consistently, using the sequence
1853 "H=<name> [<ip>]" wherever possible.
1855 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1856 are problematic for Debian distribution, omit them from the release
1859 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1861 JH/16 Fix string representation of time values on 64bit time_t architectures.
1864 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1865 output list separator was used.
1870 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1871 checkers that were complaining about end of non-void function with no
1874 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1875 This was a regression introduced in 4.83 by another bugfix.
1877 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1879 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1880 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1886 TF/01 Correctly close the server side of TLS when forking for delivery.
1888 When a message was received over SMTP with TLS, Exim failed to clear up
1889 the incoming connection properly after forking off the child process to
1890 deliver the message. In some situations the subsequent outgoing
1891 delivery connection happened to have the same fd number as the incoming
1892 connection previously had. Exim would try to use TLS and fail, logging
1893 a "Bad file descriptor" error.
1895 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1896 utilities have not been installed.
1898 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1899 temporary space as the ACL may create new global variables.
1901 TL/01 LDAP support uses per connection or global context settings, depending
1902 upon the detected version of the libraries at build time.
1904 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1905 to extract and use the src ip:port in logging and expansions as if it
1906 were a direct connection from the outside internet. PPv2 support was
1907 updated based on HAProxy spec change in May 2014.
1909 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1911 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1912 Properly escape header and check for NULL return.
1914 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1917 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1919 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1920 characters in header names, implemented as a verify condition.
1921 Contributed by Michael Fischer v. Mollard.
1923 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1924 results permerror and temperror. Previous values are deprecated but
1925 still accepted. In a future release, err_perm and err_temp will be
1926 completely removed, which will be a backward incompatibility if the
1927 ACL tests for either of these two old results. Patch contributed by
1928 user bes-internal on the mailing list.
1930 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1932 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1933 selectors, in both main and reject logs.
1935 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1938 JH/07 Add malware type "sock" for talking to simple daemon.
1940 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1942 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1943 routers/transports under cutthrough routing.
1945 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1946 numbers. Touch up "bool" conditional to keep the same definition.
1948 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1950 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1952 JH/12 Expand items in router/transport headers_add or headers_remove lists
1953 individually rather than the list as a whole. Bug 1452.
1955 Required for reasonable handling of multiple headers_ options when
1956 they may be empty; requires that headers_remove items with embedded
1957 colons must have them doubled (or the list-separator changed).
1959 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1960 view the policy declared in the DMARC record. Currently, $dmarc_status
1961 is a combined value of both the record presence and the result of the
1964 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1966 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1967 dnslookup router and the smtp transport (applying to the forward
1970 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1971 of ldap servers used for a specific lookup. Patch provided by Heiko
1974 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1975 New variable $lookup_dnssec_authenticated for observability.
1977 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1978 Patch submitted by Lars Timman.
1980 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1982 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1983 Requires trusted mode and valid format message id, aborts otherwise.
1984 Patch contributed by Heiko Schlichting.
1986 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1987 certextract with support for various fields. Bug 1358.
1989 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1990 is requested by default, modifiable by smtp transport option
1993 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1994 operate on certificate variables to give certificate fingerprints
1995 Also new ${sha256:cert_variable}.
1997 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1999 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2000 Christian Aistleitner.
2002 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2004 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2005 file. Patch from Wolfgang Breyha.
2007 JH/25 Expand the coverage of the delivery $host and $host_address to
2008 client authenticators run in verify callout. Bug 1476.
2010 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2011 align with daemon_smtp_ports. Bug 72.
2013 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2014 support and error reporting did not work properly.
2016 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2017 and is readable. Patch from Andrew Colin Kissa.
2019 TL/14 Enhance documentation of ${run expansion and how it parses the
2020 commandline after expansion, particularly in the case when an
2021 unquoted variable expansion results in an empty value.
2023 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2025 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2026 by renaming away from it. Addresses GH issue 15, reported by
2029 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2031 TL/15 SECURITY: prevent double expansion in math comparison functions
2032 (can expand unsanitized data). Not remotely exploitable.
2039 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2041 PP/02 Make -n do something, by making it not do something.
2042 When combined with -bP, the name of an option is not output.
2044 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2047 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2048 $sender_host_name and config options to manage this, and basic check
2051 PP/05 DSCP support for outbound connections and control modifier for inbound.
2053 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2054 (Only plugin which currently uses this is kerberos4, which nobody should
2055 be using, but we should make it available and other future plugins might
2056 conceivably use it, even though it would break NAT; stuff *should* be
2057 using channel bindings instead).
2059 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2060 name; added for Sendmail compatibility; requires admin caller.
2061 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2062 just ignore it); requires trusted caller.
2063 Also parse but ignore: -Ac -Am -X<logfile>
2066 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2068 TL/02 Add +smtp_confirmation as a default logging option.
2070 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2071 Patch by Magnus Holmgren from 2007-02-20.
2073 TL/04 Bugzilla 1281 - Spec typo.
2074 Bugzilla 1283 - Spec typo.
2075 Bugzilla 1290 - Spec grammar fixes.
2077 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2079 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2081 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2082 mailing list by by Dmitry Isaikin.
2084 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2086 JH/02 Support "G" suffix to numbers in ${if comparisons.
2088 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2090 NM/01 Bugzilla 1197 - Spec typo
2091 Bugzilla 1196 - Spec examples corrections
2093 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2095 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2096 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2099 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2100 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2102 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2103 "acl {{name}{arg}...}", and optional args on acl condition
2106 JH/05 Permit multiple router/transport headers_add/remove lines.
2108 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2110 JH/07 Avoid using a waiting database for a single-message-only transport.
2111 Performance patch from Paul Fisher. Bugzilla 1262.
2113 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2116 JH/09 Add $headers_added variable, with content from use of ACL modifier
2117 add_header (but not yet added to the message). Bugzilla 199.
2119 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2120 Pulled from Bugzilla 817 by Wolfgang Breyha.
2122 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2124 (nb: this is the same fix as in Exim 4.80.1)
2126 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2129 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2132 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2134 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2135 advertise SMTP AUTH mechanism to us, instead of a generic
2136 protocol violation error. Also, make Exim more robust to bad
2137 data from the Dovecot auth socket.
2139 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2141 When a queue runner is handling a message, Exim first routes the
2142 recipient addresses, during which it prunes them based on the retry
2143 hints database. After that it attempts to deliver the message to
2144 any remaining recipients. It then updates the hints database using
2147 So if a recipient address works intermittently, it can get repeatedly
2148 deferred at routing time. The retry hints record remains fresh so the
2149 address never reaches the final cutoff time.
2151 This is a fairly common occurrence when a user is bumping up against
2152 their storage quota. Exim had some logic in its local delivery code
2153 to deal with this. However it did not apply to per-recipient defers
2154 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2156 This change adds a proper retry rule check during routing so that the
2157 final cutoff time is checked against the message's age. We only do
2158 this check if there is an address retry record and there is not a
2159 domain retry record; this implies that previous attempts to handle
2160 the address had the retry_use_local_parts option turned on. We use
2161 this as an approximation for the destination being like a local
2162 delivery, as in LMTP.
2164 I suspect this new check makes the old local delivery cutoff check
2165 redundant, but I have not verified this so I left the code in place.
2167 TF/02 Correct gecos expansion when From: is a prefix of the username.
2169 Test 0254 submits a message to Exim with the header
2173 When I ran the test suite under the user fanf2, Exim expanded
2174 the header to contain my full name, whereas it should have added
2175 a Resent-Sender: header. It erroneously treats any prefix of the
2176 username as equal to the username.
2178 This change corrects that bug.
2180 GF/01 DCC debug and logging tidyup
2181 Error conditions log to paniclog rather than rejectlog.
2182 Debug lines prefixed by "DCC: " to remove any ambiguity.
2184 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2186 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2187 Bug spotted by Jeremy Harris; was flawed since initial commit.
2188 Would have resulted in OCSP responses post-SNI triggering an Exim
2189 NULL dereference and crash.
2191 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2193 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2194 Bug detection, analysis and fix by Samuel Thibault.
2195 Bugzilla 1331, Debian bug #698092.
2197 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2199 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2200 Server implementation by Todd Lyons, client by JH.
2201 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2202 config variable "prdr_enable" controls whether the server
2203 advertises the facility. If the client requests PRDR a new
2204 acl_data_smtp_prdr ACL is called once for each recipient, after
2205 the body content is received and before the acl_smtp_data ACL.
2206 The client is controlled by both of: a hosts_try_prdr option
2207 on the smtp transport, and the server advertisement.
2208 Default client logging of deliveries and rejections involving
2209 PRDR are flagged with the string "PRDR".
2211 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2212 fclose(). Diagnosis by Todd Lyons.
2214 PP/17 Update configure.default to handle IPv6 localhost better.
2215 Patch by Alain Williams (plus minor tweaks).
2218 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2219 This is now consistent with GnuTLS, and is now documented: the
2220 previous undocumented portable approach to treating the option as
2221 unset was to force an expansion failure. That still works, and
2222 an empty string is now equivalent.
2224 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2225 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2226 not performing validation itself.
2228 PP/20 Added force_command boolean option to pipe transport.
2229 Patch from Nick Koston, of cPanel Inc.
2231 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2234 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2236 PP/21 Fix eximon continuous updating with timestamped log-files.
2237 Broken in a format-string cleanup in 4.80, missed when I repaired the
2238 other false fix of the same issue.
2239 Report and fix from Heiko Schlichting.
2242 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2243 Report from Prashanth Katuri.
2245 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2246 It's SecureTransport, so affects any MacOS clients which use the
2247 system-integrated TLS libraries, including email clients.
2249 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2250 using a MIME ACL for non-SMTP local injection.
2251 Report and assistance in diagnosis by Warren Baker.
2253 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2255 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2257 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2258 last id that failed so it may be referenced in subsequent ACL's.
2260 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2263 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2264 ldap library initialization, allowing self-signed CA's to be
2265 used. Also properly sets require_cert option later in code by
2266 using NULL (global ldap config) instead of ldap handle (per
2267 session). Bug diagnosis and testing by alxgomz.
2269 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2270 the src/util/ subdirectory.
2272 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2273 renamed to Transport Post Delivery Action by Jeremy Harris, as
2276 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2277 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2278 redis_servers = needs to be configured which will be used by the redis
2279 lookup. Patch from Warren Baker, of The Packet Hub.
2281 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2283 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2284 hostname or reverse DNS when processing a host list. Used suggestions
2285 from multiple comments on this bug.
2287 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2289 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2290 Missed a few lines, added it to make the runtest require no keyboard
2293 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2294 contains upper case chars. Make router use caseful_local_part.
2296 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2297 support when GnuTLS has been built with p11-kit.
2303 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2305 This, or similar/improved, will also be change PP/11 of 4.82.
2311 PP/01 Handle short writes when writing local log-files.
2312 In practice, only affects FreeBSD (8 onwards).
2313 Bugzilla 1053, with thanks to Dmitry Isaikin.
2315 NM/01 Bugzilla 949 - Documentation tweak
2317 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2320 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2322 PP/02 Implemented gsasl authenticator.
2324 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2326 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2327 `pkg-config foo` for cflags/libs.
2329 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2330 with rest of GSASL and with heimdal_gssapi.
2332 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2333 `pkg-config foo` for cflags/libs for the TLS implementation.
2335 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2336 properties get this fed in as external SSF. A number of robustness
2337 and debugging improvements to the cyrus_sasl authenticator.
2339 PP/08 cyrus_sasl server now expands the server_realm option.
2341 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2342 Patch by Jeremy Harris.
2344 PP/10 Added dbmjz lookup type.
2346 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2348 PP/12 MAIL args handles TAB as well as SP, for better interop with
2349 non-compliant senders.
2350 Analysis and variant patch by Todd Lyons.
2352 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2353 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2354 Patch from Dirk Mueller <dmueller@suse.com>
2356 PP/13 tls_peerdn now print-escaped for spool files.
2357 Observed some $tls_peerdn in wild which contained \n, which resulted
2358 in spool file corruption.
2360 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2361 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2362 or write after TLS renegotiation, which otherwise led to messages
2365 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2366 as a tracking header (ie: a signed header comes before the signature).
2367 Patch from Wolfgang Breyha.
2369 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2370 comma-sep list; embedded commas doubled.
2372 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2374 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2376 Report and patch from Dmitry Banschikov.
2378 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2379 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2380 failures. We appear to now support TLS1.1+ with Exim.
2382 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2383 lets Exim select keys and certificates based upon TLS SNI from client.
2384 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2385 before an outbound SMTP session. New log_selector, +tls_sni.
2387 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2388 NULL dereference. Report and patch from Alun Jones.
2390 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2391 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2392 Not seeing resolver debug output on NetBSD, but suspect this is a
2393 resolver implementation change.
2395 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2396 Left warnings. Added "eximon gdb" invocation mode.
2398 PP/21 Defaulting "accept_8bitmime" to true, not false.
2400 PP/22 Added -bw for inetd wait mode support.
2402 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2403 locate the relevant includes and libraries. Made this the default.
2405 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2406 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2408 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2409 This may cause build issues on older platforms.
2411 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2412 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2413 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2414 Added SNI support via GnuTLS too.
2415 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2417 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2419 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2420 Applied second patch from Janne, implementing suggestion to default
2421 multiple-strings-in-record handling to match SPF spec.
2423 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2425 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2426 read-only, out of scope).
2427 Patch from Wolfgang Breyha, report from Stuart Northfield.
2429 PP/29 Fix three issues highlighted by clang analyser static analysis.
2430 Only crash-plausible issue would require the Cambridge-specific
2431 iplookup router and a misconfiguration.
2432 Report from Marcin Mirosław.
2434 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2436 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2437 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2438 As part of this, removing so much warning spew let me fix some minor
2439 real issues in debug logging.
2441 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2442 assignment on my part. Fixed.
2444 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2445 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2446 Janne Snabb (who went above and beyond: thank you).
2448 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2449 string otherwise requires a connection and a bunch more work and it's
2450 relatively easy to get wrong. Should also expose TLS library linkage
2453 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2454 64-bit ${eval} (JH/03).
2456 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2457 GNU libc to support some of the 64-bit stuff, should not lead to
2458 conflicts. Defined before os.h is pulled in, so if a given platform
2459 needs to override this, it can.
2461 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2462 protection layer was required, which is not implemented.
2463 Bugzilla 1254, patch from Wolfgang Breyha.
2465 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2466 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2467 tls_dhparam take prime identifiers. Also unbreak combination of
2468 OpenSSL+DH_params+TLSSNI.
2470 PP/39 Disable SSLv2 by default in OpenSSL support.
2476 PP/01 Solaris build fix for Oracle's LDAP libraries.
2477 Bugzilla 1109, patch from Stephen Usher.
2479 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2481 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2484 TF/02 Fix a couple more cases where we did not log the error message
2485 when unlink() failed. See also change 4.74-TF/03.
2487 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2488 lock up or crash if it happened to be inside a call to libc when it
2489 got a SIGUSR1 from exiwhat.
2491 The SIGUSR1 handler appends the current process status to the process
2492 log which is later printed by exiwhat. It used to use the general
2493 purpose logging code to do this, but several functions it calls are
2494 not safe for signals.
2496 The new output code in the SIGUSR1 handler is specific to the process
2497 log, and simple enough that it's easy to inspect for signal safety.
2498 Removing some special cases also simplifies the general logging code.
2499 Removing the spurious timestamps from the process log simplifies
2502 TF/04 Improved ratelimit ACL condition.
2504 The /noupdate option has been deprecated in favour of /readonly which
2505 has clearer semantics. The /leaky, /strict, and /readonly update modes
2506 are mutually exclusive. The update mode is no longer included in the
2507 database key; it just determines when the database is updated. (This
2508 means that when you upgrade Exim will forget old rate measurements.)
2510 Exim now checks that the per_* options are used with an update mode that
2511 makes sense for the current ACL. For example, when Exim is processing a
2512 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2513 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2514 must specify per_mail/readonly. If you omit the update mode it defaults to
2515 /leaky where that makes sense (as before) or /readonly where required.
2517 The /noupdate option is now undocumented but still supported for
2518 backwards compatibility. It is equivalent to /readonly except that in
2519 ACLs where /readonly is required you may specify /leaky/noupdate or
2520 /strict/noupdate which are treated the same as /readonly.
2522 A useful new feature is the /count= option. This is a generalization
2523 of the per_byte option, so that you can measure the throughput of other
2524 aggregate values. For example, the per_byte option is now equivalent
2525 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2527 The per_rcpt option has been generalized using the /count= mechanism
2528 (though it's more complicated than the per_byte equivalence). When it is
2529 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2530 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2531 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2532 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2533 Note that using per_rcpt with a non-readonly update mode in more than
2534 one ACL will cause the recipients to be double-counted. (The per_mail
2535 and per_byte options don't have this problem.)
2537 The handling of very low rates has changed slightly. If the computed rate
2538 is less than the event's count (usually one) then this event is the first
2539 after a long gap. In this case the rate is set to the same as this event's
2540 count, so that the first message of a spam run is counted properly.
2542 The major new feature is a mechanism for counting the rate of unique
2543 events. The new per_addr option counts the number of different
2544 recipients that someone has sent messages to in the last time period. It
2545 behaves like per_rcpt if all the recipient addresses are different, but
2546 duplicate recipient addresses do not increase the measured rate. Like
2547 the /count= option this is a general mechanism, so the per_addr option
2548 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2549 example, measure the rate that a client uses different sender addresses
2550 with the options per_mail/unique=$sender_address. There are further
2551 details in the main documentation.
2553 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2555 TF/06 Removed a few PCRE remnants.
2557 TF/07 Automatically extract Exim's version number from tags in the git
2558 repository when doing development or release builds.
2560 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2561 Bugzilla 879. Patch from Paul Fisher.
2563 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2564 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2567 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2569 PP/05 Variable $av_failed, true if the AV scanner deferred.
2570 Bugzilla 1078. Patch from John Horne.
2572 PP/06 Stop make process more reliably on build failure.
2573 Bugzilla 1087. Patch from Heiko Schlittermann.
2575 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2576 Bugzilla 1089. Patch from Heiko Schlittermann.
2578 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2579 Bugzilla 1131. Patch from Holger Weiß.
2581 PP/09 Handle IPv6 addresses with SPF.
2582 Bugzilla 860. Patch from Wolfgang Breyha.
2584 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2586 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2589 PP/11 match_* no longer expand right-hand-side by default.
2590 New compile-time build option, EXPAND_LISTMATCH_RHS.
2591 New expansion conditions, "inlist", "inlisti".
2593 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2595 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2597 PP/14 fix log_write() format string regression from TF/03.
2598 Bugzilla 1152. Patch from Dmitry Isaikin.
2604 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2606 PP/02 Harmonised TLS library version reporting; only show if debugging.
2607 Layout now matches that introduced for other libraries in 4.74 PP/03.
2609 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2611 PP/04 New "dns_use_edns0" global option.
2613 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2616 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2617 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2619 TK/01 Updated PolarSSL code to 0.14.2.
2620 Bugzilla 1097. Patch from Andreas Metzler.
2622 PP/07 Catch divide-by-zero in ${eval:...}.
2623 Fixes bugzilla 1102.
2625 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2628 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2629 format-string attack -- SECURITY: remote arbitrary code execution.
2631 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2632 time unintentionally subject to list matching rules, letting the header
2633 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2634 arbitrary string expansion). This allowed for information disclosure.
2636 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2637 INT_MIN/-1 -- value coerced to INT_MAX.
2643 NM/01 Workaround for PCRE version dependency in version reporting
2646 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2647 This fixes portability to compilers other than gcc, notably
2648 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2650 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2651 makefiles for portability to HP-UX and POSIX correctness.
2653 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2654 Also via indented variable definition in the Makefile.
2655 (Debugging by Oliver Heesakkers).
2657 PP/02 Restore caching of spamd results with expanded spamd_address.
2658 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2660 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2661 Improves build reliability. Fix from: Frank Elsner
2663 NM/02 Fix wide character breakage in the rfc2047 coding
2664 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2666 NM/03 Allow underscore in dnslist lookups
2667 Fixes bug 1026. Patch from Graeme Fowler
2669 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2670 Code patches from Adam Ciarcinski of NetBSD.
2672 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2675 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2676 is logged, to avoid truncation. Patch from John Horne.
2678 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2679 Patch from Jakob Hirsch.
2681 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2682 SQL string expansion failure details.
2683 Patch from Andrey Oktyabrski.
2685 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2686 Patch from Simon Arlott.
2688 PP/09 New lookups functionality failed to compile on old gcc which rejects
2689 extern declarations in function scope.
2690 Patch from Oliver Fleischmann
2692 PP/10 Use sig_atomic_t for flags set from signal handlers.
2693 Check getgroups() return and improve debugging.
2694 Fixed developed for diagnosis in bug 927 (which turned out to be
2697 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2698 Patch from Mark Zealey.
2700 PP/12 Bugzilla 1056: Improved spamd server selection.
2701 Patch from Mark Zealey.
2703 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2704 Based on patch from Heiko Schlittermann.
2706 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2707 Patch from Uwe Doering, sign-off by Michael Haardt.
2709 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2710 variable declaration deep within a block. Bug and patch from
2713 PP/15 lookups-Makefile IRIX compatibility coercion.
2715 PP/16 Make DISABLE_DKIM build knob functional.
2717 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2718 Patch by Simon Arlott
2720 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2721 variable argument macros. Our copy now differs from upstream.
2727 TF/01 Failure to get a lock on a hints database can have serious
2728 consequences so log it to the panic log.
2730 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2731 controlled using the smtp_confirmation log selector.
2733 TF/03 Include the error message when we fail to unlink a spool file.
2735 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2736 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2737 for maintaining out-of-tree patches for some time.
2739 PP/01 Bugzilla 139: Documentation and portability issues.
2740 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2741 Handle per-OS dynamic-module compilation flags.
2743 PP/02 Let /dev/null have normal permissions.
2744 The 4.73 fixes were a little too stringent and complained about the
2745 permissions on /dev/null. Exempt it from some checks.
2746 Reported by Andreas M. Kirchwitz.
2748 PP/03 Report version information for many libraries, including
2749 Exim version information for dynamically loaded libraries. Created
2750 version.h, now support a version extension string for distributors
2751 who patch heavily. Dynamic module ABI change.
2753 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2754 privilege escalation vulnerability whereby the Exim run-time user
2755 can cause root to append content of the attacker's choosing to
2758 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2761 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2762 If dropping privileges for untrusted macros, we disabled normal logging
2763 on the basis that it would fail; for the Exim run-time user, this is not
2764 the case, and it resulted in successful deliveries going unlogged.
2765 Fixed. Reported by Andreas Metzler.
2771 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2772 only prepend for the Resent-* case. Fixes regression introduced in
2773 Exim 4.70 by NM/22 for Bugzilla 607.
2775 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2776 increasing numbers of administrators be bitten by this.
2778 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2780 PP/03 Bugzilla 994: added openssl_options main configuration option.
2782 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2784 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2786 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2788 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2789 without a peer certificate, leading to a segfault because of an
2790 assumption that peers always have certificates. Be a little more
2791 paranoid. Problem reported by Martin Tscholak.
2793 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2794 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2795 NB: ClamAV planning to remove STREAM in "middle of 2010".
2796 CL also introduces -bmalware, various -d+acl logging additions and
2797 more caution in buffer sizes.
2799 PP/09 Implemented reverse_ip expansion operator.
2801 PP/10 Bugzilla 937: provide a "debug" ACL control.
2803 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2805 PP/12 Bugzilla 973: Implement --version.
2807 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2809 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2811 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2813 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2814 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2815 ignore trailing whitespace.
2817 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2819 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2822 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2823 Notification from Dr Andrew Aitchison.
2825 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2826 ExtendedDetectionInfo response format.
2827 Notification from John Horne.
2829 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2832 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2833 XSL and documented dependency on system catalogs, with examples of how
2836 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2839 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2840 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2841 configuration file which is writeable by the Exim user or group.
2843 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2844 of configuration files to cover files specified with the -C option if
2845 they are going to be used with root privileges, not just the default
2848 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2849 option (effectively making it always true).
2851 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2852 files to be used while preserving root privileges.
2854 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2855 that rogue child processes cannot use them.
2857 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2858 run-time user, instead of root.
2860 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2861 Exim run-time user without dropping privileges.
2863 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2864 result string, instead of calling string_vformat() twice with the same
2867 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2868 for other users. Others should always drop root privileges if they use
2869 -C on the command line, even for a whitelisted configure file.
2871 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2873 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2879 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2880 $data_path, and $header_path variables; fixed documentation bugs and
2883 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2884 exipick to access non-standard spools, including the "frozen" queue
2887 NM/01 Bugzilla 965: Support mysql stored procedures.
2888 Patch from Alain Williams
2890 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2892 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2893 Patch from Andreas Metzler
2895 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2896 Patch from Kirill Miazine
2898 NM/05 Bugzilla 671: Added umask to procmail example.
2900 JJ/03 installed exipick 20100323.0, fixing doc bug
2902 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2903 directory. Notification and patch from Dan Rosenberg.
2905 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2907 TK/02 Improve log output when DKIM signing operation fails.
2909 MH/01 Treat the transport option dkim_domain as a colon separated
2910 list, not as a single string, and sign the message with each element,
2911 omitting multiple occurences of the same signer.
2913 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2914 Bugzilla 985, 986. Patch by Simon Arlott
2916 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2917 Patch by Simon Arlott
2919 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2920 MBX locking. Notification from Dan Rosenberg.
2926 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2928 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2930 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2932 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2934 NM/04 Bugzilla 915: Flag broken perl installation during build.
2940 TK/01 Added patch by Johannes Berg that expands the main option
2941 "spamd_address" if it starts with a dollar sign.
2943 TK/02 Write list of recipients to X-Envelope-Sender header when building
2944 the mbox-format spool file for content scanning (suggested by Jakob
2947 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2948 (http://www.dcc-servers.net/) support via dccifd. Activated by
2949 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2951 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2952 by Mark Daniel Reidel <mr@df.eu>.
2954 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2955 When building exim an external PCRE library is now needed -
2956 PCRE is a system library on the majority of modern systems.
2957 See entry on PCRE_LIBS in EDITME file.
2959 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2960 conversation. Added nologin parameter to request.
2961 Patch contributed by Kirill Miazine.
2963 TF/01 Do not log submission mode rewrites if they do not change the address.
2965 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2967 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2968 log files in place. Contributed by Roberto Lima.
2970 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2972 TF/03 Bugzilla 615: When checking the local_parts router precondition
2973 after a local_part_suffix or local_part_prefix option, Exim now
2974 does not use the address's named list lookup cache, since this
2975 contains cached lookups for the whole local part.
2977 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2978 Robert Millan. Documentation is in experimental-spec.txt.
2980 TF/04 Bugzilla 668: Fix parallel build (make -j).
2982 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2984 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2985 Patch provided by Jan Srzednicki.
2987 TF/05 Leading white space used to be stripped from $spam_report which
2988 wrecked the formatting. Now it is preserved.
2990 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2991 that they are available at delivery time.
2993 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2995 TF/08 TLS error reporting now respects the incoming_interface and
2996 incoming_port log selectors.
2998 TF/09 Produce a more useful error message if an SMTP transport's hosts
2999 setting expands to an empty string.
3001 NM/06 Bugzilla 744: EXPN did not work under TLS.
3002 Patch provided by Phil Pennock.
3004 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3005 Patch provided by Richard Godbee.
3007 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3008 acl_smtp_notquit, added index entry.
3010 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3011 Patch provided by Eugene Bujak.
3013 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3014 accept(). Patch provided by Maxim Dounin.
3016 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3017 Patch provided by Phil Pennock.
3019 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3021 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3022 Patch provided by Brad "anomie" Jorsch.
3024 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3025 Patch provided by Dean Brooks.
3027 TK/05 Add native DKIM support (does not depend on external libraries).
3029 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3030 Patch provided by Graeme Fowler.
3032 NM/16 Bugzilla 851: Documentation example syntax fix.
3034 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3036 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3039 NM/19 Bugzilla 745: TLS version reporting.
3040 Patch provided by Phil Pennock.
3042 NM/20 Bugzilla 167: bool: condition support.
3043 Patch provided by Phil Pennock.
3045 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3046 clients. Patch provided by Phil Pennock.
3048 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3049 Patch provided by Brad "anomie" Jorsch.
3051 NM/23 Bugzilla 687: Fix misparses in eximstats.
3052 Patch provided by Heiko Schlittermann.
3054 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3055 Patch provided by Heiko Schlittermann.
3057 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3058 plus update to original patch.
3060 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3062 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3063 Patch provided by David Brownlee.
3065 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3067 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3069 NM/30 Bugzilla 888: TLS documentation bugfixes.
3071 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3073 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3074 Unlike the original bugzilla I have changed all shell scripts in src tree.
3076 NM/33 Bugzilla 898: Transport filter timeout fix.
3077 Patch by Todd Rinaldo.
3079 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3080 Patch by Serge Demonchaux.
3082 NM/35 Bugzilla 39: Base64 decode bug fixes.
3083 Patch by Jakob Hirsch.
3085 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3087 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3089 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3091 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3097 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3098 ALT-N's libdkim that I have put here:
3099 http://duncanthrax.net/exim-experimental/
3101 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3102 were called 'true' and it seems that C99 defines that as a reserved
3103 keyword to be used with 'bool' variable types. That means you could
3104 not include C99-style headers which use bools without triggering
3105 build errors in sieve.c.
3107 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3108 as mailq or other aliases. Changed the --help handling significantly
3109 to do whats expected. exim_usage() emits usage/help information.
3111 SC/01 Added the -bylocaldomain option to eximstats.
3113 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3115 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3117 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3123 PH/01 Another patch from the Sieve maintainer.
3125 PH/02 When an IPv6 address is converted to a string for single-key lookup
3126 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3127 dots are used instead of colons so that keys in lsearch files need not
3128 contain colons. This was done some time before quoting was made available
3129 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3130 (notated using the quote facility) so as to distinguish them from IPv4
3131 keys. This meant that lookups for IP addresses in host lists did not work
3132 for iplsearch lookups.
3134 This has been fixed by arranging for IPv6 addresses to be expressed with
3135 colons if the lookup type is iplsearch. This is not incompatible, because
3136 previously such lookups could never work.
3138 The situation is now rather anomalous, since one *can* have colons in
3139 ordinary lsearch keys. However, making the change in all cases is
3140 incompatible and would probably break a number of configurations.
3142 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3145 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3146 conversion specification without a maximum field width, thereby enabling
3147 a rogue spamd server to cause a buffer overflow. While nobody in their
3148 right mind would setup Exim to query an untrusted spamd server, an
3149 attacker that gains access to a server running spamd could potentially
3150 exploit this vulnerability to run arbitrary code as the Exim user.
3152 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3153 $primary_hostname instead of what libspf2 thinks the hosts name is.
3155 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3156 a directory entry by the name of the lookup key. Previously, if a
3157 symlink pointed to a non-existing file or a file in a directory that
3158 Exim lacked permissions to read, a lookup for a key matching that
3159 symlink would fail. Now it is enough that a matching directory entry
3160 exists, symlink or not. (Bugzilla 503.)
3162 PH/03 The body_linecount and body_zerocount variables are now exported in the
3165 PH/04 Added the $dnslist_matched variable.
3167 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3168 This means they are set thereafter only if the connection becomes
3171 PH/06 Added the client_condition to authenticators so that some can be skipped
3172 by clients under certain conditions.
3174 PH/07 The error message for a badly-placed control=no_multiline_responses left
3175 "_responses" off the end of the name.
3177 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3179 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3180 (without spaces) instead of just copying the configuration text.
3182 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3184 PH/11 Added $max_received_linelength.
3186 PH/12 Added +ignore_defer and +include_defer to host lists.
3188 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3189 way in which PCRE > 7.0 is built.
3191 PH/14 Implemented queue_only_load_latch.
3193 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3194 MAIL command. The effect was to mangle the value on 64-bit systems.
3196 PH/16 Another patch from the Sieve maintainer.
3198 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3200 PH/18 If a system quota error occurred while trying to create the file for
3201 a maildir delivery, the message "Mailbox is full" was not appended to the
3202 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3203 only to a quota excession during the actual writing of the file.
3205 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3206 characters?) which causes problems in log lines. The DN values are now
3207 passed through string_printing() before being added to log lines.
3209 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3210 and InterBase are left for another time.)
3212 PH/21 Added message_body_newlines option.
3214 PH/22 Guard against possible overflow in moan_check_errorcopy().
3216 PH/23 POSIX allows open() to be a macro; guard against that.
3218 PH/24 If the recipient of an error message contained an @ in the local part
3219 (suitably quoted, of course), incorrect values were put in $domain and
3220 $local_part during the evaluation of errors_copy.
3226 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3227 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3230 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3231 issue a MAIL command.
3233 PH/02 In an ACL statement such as
3235 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3237 if a client was not listed at all, or was listed with a value other than
3238 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3239 the condition was not true (as it should be), so access was not denied.
3240 The bug was that the ! inversion was incorrectly passed on to the second
3241 item. This has been fixed.
3243 PH/03 Added additional dnslists conditions == and =& which are different from
3244 = and & when the dns lookup returns more than one IP address.
3246 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3247 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3249 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3250 FSYNC, which compiles an option called disable_fsync that allows for
3251 bypassing fsync(). The documentation is heavily laced with warnings.
3253 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3255 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3256 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3257 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3258 including adding "make clean"; (3) Added -fPIC when compiling the test
3259 dynamically loaded module, to get rid of a warning.
3261 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3262 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3263 = 0s. The bug is otherwise harmless.
3265 PH/07 There was a bug in the dovecot authenticator such that the value of
3266 $auth1 could be overwritten, and so not correctly preserved, after a
3267 successful authentication. This usually meant that the value preserved by
3268 the server_setid option was incorrect.
3270 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3272 PH/09 Installed PCRE release 7.0.
3274 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3275 run for batched SMTP input. It is now run at the start of every message
3276 in the batch. While fixing this I discovered that the process information
3277 (output by running exiwhat) was not always getting set for -bs and -bS
3278 input. This is fixed, and it now also says "batched" for BSMTP.
3280 PH/11 Added control=no_pipelining.
3282 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3283 patch, slightly modified), and move the expansion of helo_data till after
3284 the connection is made in the smtp transport (so it can use these
3287 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3289 PH/14 Added log_selector = +pid.
3291 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3293 PH/16 Add ${if forany and ${if forall.
3295 PH/17 Added dsn_from option to vary the From: line in DSNs.
3297 PH/18 Flush SMTP output before performing a callout, unless control =
3298 no_callout_flush is set.
3300 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3301 was true (the default) a successful delivery failed to delete the retry
3302 item, thus causing premature timeout of the address. The bug is now
3305 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3307 PH/21 Long custom messages for fakedefer and fakereject are now split up
3308 into multiline responses in the same way that messages for "deny" and
3309 other ACL rejections are.
3311 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3312 with slight modification.
3314 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3315 draft, changing the syntax and factoring some duplicate code".
3317 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3318 for deliveries of the second and subsequent messages over the same SMTP
3321 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3322 ${reduce, with only minor "tidies".
3324 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3326 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3327 expansion side effects.
3329 PH/27 When a message times out after an over-quota error from an Exim-imposed
3330 quota, the bounce message says "mailbox is full". This message was not
3331 being given when it was a system quota that was exceeded. It now should
3334 MH/03 Made $recipients available in local_scan(). local_scan() already has
3335 better access to the recipient list through recipients_list[], but
3336 $recipients can be useful in postmaster-provided expansion strings.
3338 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3339 in the case of a MAIL command with additional options following the
3340 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3341 were accidentally chopped off.
3343 PH/29 SMTP synchronization checks are implemented when a command is read -
3344 there is a check that no more input is waiting when there shouldn't be
3345 any. However, for some commands, a delay in an ACL can mean that it is
3346 some time before the response is written. In this time, more input might
3347 arrive, invalidly. So now there are extra checks after an ACL has run for
3348 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3349 pipelining has not been advertised.
3351 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3353 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3354 returned in the SMTP response when smtp_return_error_details was false.
3355 This has been fixed.
3357 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3358 instead of the C I/O that was originally supplied, because problems were
3359 reported on Solaris.
3361 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3362 Exim which did not show up earlier: it was assuming that a call to
3363 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3364 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3365 was a macro that became an assignment, so it seemed to work. This has
3366 changed to a proper function call with a void return, hence the compile
3367 error. Exim's code has been fixed.
3369 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3372 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3374 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3376 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3377 because it turns out to be too big or there is a timeout) there is no
3378 "Completed" line in the log. When some messages of this type were
3379 selected by exigrep, they were listed as "not completed". Others were
3380 picked up by some special patterns. I have improved the selection
3381 criteria to be more general.
3383 PH/38 The host_find_failed option in the manualroute router can now be set
3384 to "ignore", to completely ignore a host whose IP address cannot be
3385 found. If all hosts are ignored, the behaviour is controlled by the new
3386 host_all_ignored option.
3388 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3389 homing or because of multiple MX records with /mx) generated more than
3390 one IP address, and the following item turned out to be the local host,
3391 all the secondary addresses of the first item were incorrectly removed
3392 from the list, along with the local host and any following hosts (which
3393 is what is supposed to happen).
3395 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3396 whoever called Exim into the -H file. In the case of the daemon it was
3397 behaving confusingly. When first started, it used values for whoever
3398 started the daemon, but after a SIGHUP it used the Exim user (because it
3399 calls itself on a restart). I have changed the code so that it now always
3402 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3403 message are rejected with the same error (e.g. no authentication or bad
3404 sender address), and a DATA command is nevertheless sent (as can happen
3405 with PIPELINING or a stupid MUA), the error message that was given to the
3406 RCPT commands is included in the rejection of the DATA command. This is
3407 intended to be helpful for MUAs that show only the final error to their
3410 PH/42 Another patch from the Sieve maintainer.
3412 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3413 Eximstats - Fixed some broken HTML links and added missing column headers
3415 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3416 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3418 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3420 PH/43 Yet another patch from the Sieve maintainer.
3422 PH/44 I found a way to check for a TCP/IP connection going away before sending
3423 the response to the final '.' that terminates a message, but only in the
3424 case where the client has not sent further data following the '.'
3425 (unfortunately, this is allowed). However, in many cases there won't be
3426 any further data because there won't be any more messages to send. A call
3427 to select() can be used: if it shows that the input is "ready", there is
3428 either input waiting, or the socket has been closed. An attempt to read
3429 the next input character can distinguish the two cases. Previously, Exim
3430 would have sent an OK response which the client would never have see.
3431 This could lead to message repetition. This fix should cure that, at
3432 least in a lot of common cases.
3434 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3435 advertised in response to EHLO.
3441 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3442 fixed by 4.65/MH/01 (is this a record?) are fixed:
3444 (i) An empty string was always treated as zero by the numeric comparison
3445 operators. This behaviour has been restored.
3447 (ii) It is documented that the numeric comparison operators always treat
3448 their arguments as decimal numbers. This was broken in that numbers
3449 starting with 0 were being interpreted as octal.
3451 While fixing these problems I realized that there was another issue that
3452 hadn't been noticed. Values of message_size_limit (both the global option
3453 and the transport option) were treated as octal if they started with 0.
3454 The documentation was vague. These values are now always treated as
3455 decimal, and I will make that clear in the documentation.
3461 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3462 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3465 MH/01 Don't check that the operands of numeric comparison operators are
3466 integers when their expansion is in "skipping" mode (fixes bug
3467 introduced by 4.64-PH/07).
3469 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3470 child addresses, Exim now panics and dies. Previously, because the count
3471 is held in a short int, deliveries were likely to be lost. As such a
3472 large number of recipients for a single message is ridiculous
3473 (performance will be very, very poor), I have chosen to impose a limit
3474 rather than extend the field.
3480 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3481 leftover -K file (the existence of which was triggered by #402).
3482 While we were at it, introduced process PID as part of the -K
3483 filename. This should rule out race conditions when creating
3486 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3487 processing considerably. Previous code took too long for large mails,
3488 triggering a timeout which in turn triggers #401.
3490 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3491 in the DK code in transports.c. sendfile() is not really portable,
3492 hence the _LINUX specificness.
3494 TF/01 In the add_headers option to the mail command in an Exim filter,
3495 there was a bug that Exim would claim a syntax error in any
3496 header after the first one which had an odd number of characters
3499 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3500 callout verification, Exim cached a "reject" for the entire domain. This
3501 is correct for most verifications, but it is not correct for a recipient
3502 verification with use_sender or use_postmaster set, because in that case
3503 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3504 case of MAIL FROM:<> rejection from other early rejections (e.g.
3505 rejection of HELO). When verifying a recipient using a non-null MAIL
3506 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3507 Whatever the result of the callout, the value of the domain cache is
3508 left unchanged (for any other kind of callout, getting as far as trying
3509 RCPT means that the domain itself is ok).
3511 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3514 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3515 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3516 session, the daemon ignores ECONNECT errors and logs others; it now
3517 ignores EPIPE as well.
3519 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3520 (quoted-printable decoding).
3522 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3523 later the small subsequent patch to fix an introduced bug.
3525 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3527 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3529 PH/08 An error is now given if message_size_limit is specified negative.
3531 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3532 to be given (somewhat) arbitrary names.
3534 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3537 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3538 miscellaneous code fixes
3540 PH/10 Added the log_reject_target ACL modifier to specify where to log
3543 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3544 hostname. This is wrong, because it relates to the incoming message (and
3545 probably the interface on which it is arriving) and not to the outgoing
3546 callout (which could be using a different interface). This has been
3547 changed to use the value of the helo_data option from the smtp transport
3548 instead - this is what is used when a message is actually being sent. If
3549 there is no remote transport (possible with a router that sets up host
3550 addresses), $smtp_active_hostname is used.
3552 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3553 tweaks were necessary in order to get it to work (see also 21 below):
3554 (a) The code assumed that strncpy() returns a negative number on buffer
3555 overflow, which isn't the case. Replaced with Exim's string_format()
3557 (b) There were several signed/unsigned issues. I just did the minimum
3558 hacking in of casts. There is scope for a larger refactoring.
3559 (c) The code used strcasecmp() which is not a standard C function.
3560 Replaced with Exim's strcmpic() function.
3561 (d) The code set only $1; it now sets $auth1 as well.
3562 (e) A simple test gave the error "authentication client didn't specify
3563 service in request". It would seem that Dovecot has changed its
3564 interface. Fortunately there's a specification; I followed it and
3565 changed what the client sends and it appears to be working now.
3567 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3570 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3571 address A is aliased to B and C, where B exists and C does not. Without
3572 -v the output is "A verified" because verification stops after a
3573 successful redirection if more than one address is generated. However,
3574 with -v the child addresses are also verified. Exim was outputting "A
3575 failed to verify" and then showing the successful verification for C,
3576 with its parentage. It now outputs "B failed to verify", showing B's
3577 parentage before showing the successful verification of C.
3579 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3580 look up a TXT record in a specific list after matching in a combined
3583 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3584 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3585 they consult the DNS. I had assumed they would set it the way they
3586 wanted; and indeed my experiments on Linux seem to show that in some
3587 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3588 To be on the safe side, however, I have now made the interface to
3589 host_find_byname() similar to host_find_bydns(), with an argument
3590 containing the DNS resolver options. The host_find_byname() function now
3591 sets these options at its start, just as host_find_bydns() does. The smtp
3592 transport options dns_qualify_single and dns_search_parents are passed to
3593 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3594 of host_find_byname() use the default settings of RES_DEFNAMES
3595 (qualify_single) but not RES_DNSRCH (search_parents).
3597 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3598 spool_read_header() do less string testing, by means of a preliminary
3599 switch on the second character of optional "-foo" lines. (This is
3600 overdue, caused by the large number of possibilities that now exist.
3601 Originally there were few.) While I was there, I also converted the
3602 str(n)cmp tests so they don't re-test the leading "-" and the first
3603 character, in the hope this might squeeze out yet more improvement.
3605 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3606 flag allowing group syntax was set by the header_syntax check but not
3607 turned off, possible causing trouble later; (2) The flag was not being
3608 set at all for the header_verify test, causing "group"-style headers to
3609 be rejected. I have now set it in this case, and also caused header_
3610 verify to ignore an empty address taken from a group. While doing this, I
3611 came across some other cases where the code for allowing group syntax
3612 while scanning a header line wasn't quite right (mostly, not resetting
3613 the flag correctly in the right place). These bugs could have caused
3614 trouble for malformed header lines. I hope it is now all correct.
3616 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3617 with the "reply" argument non-NULL. The code, however (which originally
3618 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3619 but it didn't always do it. This confused somebody who was copying the
3620 code for some other use. I have removed all the tests.
3622 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3623 feature that was used to support insecure browsers during the U.S. crypto
3624 embargo. It requires special client support, and Exim is probably the
3625 only MTA that supported it -- and would never use it because real RSA is
3626 always available. This code has been removed, because it had the bad
3627 effect of slowing Exim down by computing (never used) parameters for the
3628 RSA_EXPORT functionality.
3630 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3631 authenticator to fail if there's a tab character in the incoming data
3632 (there should never be unless someone is messing about, as it's supposed
3633 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3634 if the connection is using TLS or if the remote IP is the same as the
3635 local IP, and the "valid-client-cert option" if a client certificate has
3638 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3639 authenticators. This can be used for authorization after authentication
3640 succeeds. (In the case of plaintext, it servers for both authentication
3643 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3644 if any retry times were supplied.
3646 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3647 connection, where there is no client IP address to check. In this
3648 situation, the verify now always succeeds.
3650 PH/25 Applied John Jetmore's -Mset patch.
3652 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3654 PH/27 In a string expansion for a processed (not raw) header when multiple
3655 headers of the same name were present, leading whitespace was being
3656 removed from all of them, but trailing whitespace was being removed only
3657 from the last one. Now trailing whitespace is removed from each header
3658 before concatenation. Completely empty headers in a concatenation (as
3659 before) are ignored.
3661 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3662 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3664 PH/29 [Removed. This was a change that I later backed out, and forgot to
3665 correct the ChangeLog entry (that I had efficiently created) before
3666 committing the later change.]
3668 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3669 address errors (4xx response to RCPT) over the same connection as other
3670 messages routed to the same hosts. Such deliveries are always "forced",
3671 so retry times are not inspected. This resulted in far too many retries
3672 for the affected addresses. The effect occurred only when there were more
3673 hosts than the hosts_max_try setting in the smtp transport when it had
3674 the 4xx errors. Those hosts that it had tried were not added to the list
3675 of hosts for which the message was waiting, so if all were tried, there
3676 was no problem. Two fixes have been applied:
3678 (i) If there are any address or message errors in an SMTP delivery, none
3679 of the hosts (tried or untried) are now added to the list of hosts
3680 for which the message is waiting, so the message should not be a
3681 candidate for sending over the same connection that was used for a
3682 successful delivery of some other message. This seems entirely
3683 reasonable: after all the message is NOT "waiting for some host".
3684 This is so "obvious" that I'm not sure why it wasn't done
3685 previously. Hope I haven't missed anything, but it can't do any
3686 harm, as the worst effect is to miss an optimization.
3688 (ii) If, despite (i), such a delivery is accidentally attempted, the
3689 routing retry time is respected, so at least it doesn't keep
3690 hammering the server.
3692 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3693 in ${readsocket because some servers need this prod.
3695 PH/32 Added some extra debug output when updating a wait-xxx database.
3697 PH/33 The hint "could be header name not terminated by colon", which has been
3698 given for certain expansion errors for a long time, was not being given
3699 for the ${if def:h_colon_omitted{... case.
3701 PH/34 The spec says: "With one important exception, whenever a domain list is
3702 being scanned, $domain contains the subject domain." There was at least
3703 one case where this was not true.
3705 PH/35 The error "getsockname() failed: connection reset by peer" was being
3706 written to the panic log as well as the main log, but it isn't really
3707 panic-worthy as it just means the connection died rather early on. I have
3708 removed the panic log writing for the ECONNRESET error when getsockname()
3711 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3712 runs only) independently of the message's sender address. This meant
3713 that, if the 4xx error was in fact related to the sender, a different
3714 message to the same recipient with a different sender could confuse
3715 things. In particular, this can happen when sending to a greylisting
3716 server, but other circumstances could also provoke similar problems.
3717 I have changed the default so that the retry time for these errors is now
3718 based a combination of the sender and recipient addresses. This change
3719 can be overridden by setting address_retry_include_sender=false in the
3722 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3723 remote server are returned as part of bounce messages. This was not
3724 happening for LMTP over a pipe (the lmtp transport), but now it is the
3725 same for both kinds of LMTP.
3727 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3728 in header lines that were in fact CNAMEs. This is no longer the case.
3730 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3731 and queue runs started by the daemon processed all messages. This has
3732 been fixed so that -R and -S can now usefully be given with -q<time>.
3734 PH/40 Import PCRE release 6.7 (fixes some bugs).
3736 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3738 PH/42 Give an error if -q is specified more than once.
3740 PH/43 Renamed the variables $interface_address and $interface_port as
3741 $received_ip_address and $received_port, to make it clear that these
3742 values apply to message reception, and not to the outgoing interface when
3743 a message is delivered. (The old names remain recognized, of course.)
3745 PH/44 There was no timeout on the connect() call when using a Unix domain
3746 socket in the ${readsocket expansion. There now is.
3748 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3749 be meaningful with "accept".
3751 SC/01 Eximstats V1.43
3752 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3754 SC/02 Eximstats V1.44
3755 Use a glob alias rather than an array ref in the generated
3756 parser. This improves both readability and performance.
3758 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3759 Collect SpamAssassin and rejection statistics.
3760 Don't display local sender or destination tables unless
3761 there is data to show.
3762 Added average volumes into the top table text output.
3764 SC/04 Eximstats V1.46
3765 Collect data on the number of addresses (recipients)
3766 as well as the number of messages.
3768 SC/05 Eximstats V1.47
3769 Added 'Message too big' to the list of mail rejection
3770 reasons (thanks to Marco Gaiarin).
3772 SC/06 Eximstats V1.48
3773 Mainlog lines which have GMT offsets and are too short to
3774 have a flag are now skipped.
3776 SC/07 Eximstats V1.49 (Alain Williams)
3777 Added the -emptyok flag.
3779 SC/08 Eximstats V1.50
3780 Fixes for obtaining the IP address from reject messages.
3782 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3783 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3784 whitespace changes from 4.64-PH/27
3786 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3789 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3790 are found, allow negative numbers in numeric criteria)
3792 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3794 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3795 to match changes made in 4.64-PH/43
3797 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3799 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3800 "domain" is actually an IP address, and force a failure. This locks out
3801 those revolvers/nameservers that support "A-for-A" lookups, in
3802 contravention of the specifications.
3804 PH/48 When a host name was looked up from an IP address, and the subsequent
3805 forward lookup of the name timed out, the host name was left in
3806 $sender_host_name, contrary to the specification.
3808 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3809 restricted to single-key lookups, Exim was not diagnosing an error if
3810 * or *@ was used with a query-style lookup.
3812 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3814 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3815 long ago, but noone interested enough thought of it. Let's just say that
3816 the "1.1" means that there are some new functions that weren't there at
3817 some point in the past.
3819 PH/51 Error processing for expansion failure of helo_data from an smtp
3820 transport during callout processing was broken.
3822 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3823 tested/used via the -bh/-bhc/-bs options.
3825 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3826 bug, fixed in subsequent PCRE releases).
3828 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3829 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3831 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3837 SC/01 Use a glob alias rather than an array ref in eximstats generated
3838 parser. This improves both readability and performance.
3840 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3841 Don't display local sender or destination tables in eximstats unless
3842 there is data to show.
3843 Added average volumes into the eximstats top table text output.
3845 SC/03 Collect data on the number of addresses (recipients) as well
3846 as the number of messages in eximstats.
3848 TF/01 Correct an error in the documentation for the redirect router. Exim
3849 does (usually) call initgroups() when daemonizing.
3851 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3852 with consistent privilege compared to when running as a daemon.
3854 TF/03 Note in the spec that $authenticated_id is not set for local
3855 submissions from trusted users.
3857 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3858 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3860 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3861 by adding some example configuration directives to the default
3862 configuration file. A little bit of work is required to uncomment the
3863 directives and define how usernames and passwords are checked, but
3864 there is now a framework to start from.
3866 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3867 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3868 without this. I don't know how relevant this is to other LDAP libraries.
3870 PH/02 Add the verb name to the "unknown ACL verb" error.
3872 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3874 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3876 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3877 directory not expanded when it should be if an expanded home directory
3878 was set for the address (which is overridden by the transport).
3880 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3883 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3884 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3885 because it is too late at that time, and has no effect.
3887 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3888 security issue with \' (bugzilla #107). I could not use the
3889 PQescapeStringConn() function, because it needs a PGconn value as one of
3892 PH/08 When testing addresses using -bt, indicate those final addresses that
3893 are duplicates that would not cause an additional delivery. At least one
3894 person was confused, thinking that -bt output corresponded to deliveries.
3895 (Suppressing duplicates isn't a good idea as you lose the information
3896 about possibly different redirections that led to the duplicates.)
3898 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3899 systems where poll() doesn't work, in particular OS X.
3901 PH/10 Added more information to debugging output for retry time not reached.
3903 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3904 operations in malware.c.
3906 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3909 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3910 syslog_duplication set false, log lines that would normally be written
3911 both the the main log and to the reject log were not written to syslog at
3914 PH/14 In the default configuration, change the use of "message" in ACL warn
3915 statements to "add_header".
3917 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3918 not followed by a command (e.g. "seen endif").
3920 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3921 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3924 PH/17 Added extra conditions to the default value of delay_warning_condition
3928 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3929 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3930 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3933 The Auto-Submitted: and various List- headers are standardised, whereas I
3934 don't think Precedence: ever was.
3936 PH/18 Refactored debugging code in route_finduser() to show more information,
3937 in particular, the error code if getpwnam() issues one.
3939 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3940 This is apparently needed in addition to the PH/07 change above to avoid
3941 any possible encoding problems.
3943 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3944 but not after initializing Perl.
3946 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3947 output them only if debugging. By default they are written stderr,
3948 apparently, which is not desirable.
3950 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3953 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3956 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3958 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3959 authenticated or an ident call has been made. Suppress the default
3960 values for $authenticated_id and $authenticated_sender (but permit -oMai
3961 and -oMas) when testing with -bh.
3963 PH/24 Re-jigged the order of the tests in the default configuration so that the
3964 tests for valid domains and recipients precede the DNS black list and CSA
3965 tests, on the grounds that those ones are more expensive.
3967 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3968 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3969 This bug exists in every version of Exim that I still have, right back to
3972 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3973 However, an attempt to turn on case-sensitivity in a regex key by
3974 including (?-i) didn't work because the subject string was already
3975 lowercased, and the effects were non-intuitive. It turns out that a
3976 one-line patch can be used to allow (?-i) to work as expected.
3982 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3983 other effects) broke the use of negated acl sub-conditions.
3985 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3988 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3989 "Deny" causes Exim to reject the incoming connection with a 554 error.
3990 Unfortunately, if there is a major crisis, such as a disk failure,
3991 tcp-wrappers gives "deny", whereas what one would like would be some
3992 kind of temporary error. A kludge has been added to help with this.
3993 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3994 554 error is used if errno is still zero or contains ENOENT (which occurs
3995 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3998 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4000 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4001 errors. Otherwise a message that provokes a temporary error (when other
4002 messages do not) can cause a whole host to time out.
4004 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4005 addresses were routed directly to files or pipes from a redirect router.
4006 File deliveries just didn't batch; pipe deliveries might have suffered
4009 PH/06 A failure to get a lock for a hints database would erroneously always say
4010 "Failed to get write lock", even when it was really a read lock.
4012 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4013 of 0600. This has been changed to use the value of the lockfile_mode
4014 option (which defaults to 0600).
4016 PH/08 Applied small patch from the Sieve maintainer.
4018 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4019 folder from quota calculations, a direct delivery into this folder messed
4020 up the contents of the maildirsize file. This was because the regex was
4021 used only to exclude .Trash (or whatever) when the size of the mailbox
4022 was calculated. There was no check that a delivery was happening into an
4023 excluded directory. This bug has been fixed by ignoring all quota
4024 processing for deliveries into excluded directories.
4026 PH/10 Added the maildirfolder_create_regex option to appendfile.
4032 PH/01 The code for finding all the local interface addresses on a FreeBSD
4033 system running IPv6 was broken. This may well have applied to all BSD
4034 systems, as well as to others that have similar system calls. The broken
4035 code found IPv4 interfaces correctly, but gave incorrect values for the
4036 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4037 that it would not match correctly against @[] and not recognize the IPv6
4040 PH/02 The ipliteral router was not recognizing addresses of the form user@
4041 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4043 PH/03 Added disable_ipv6.
4045 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4046 decoded form, because it is most often used to construct To: headers
4047 lines in autoreplies, and the decoded form may well be syntactically
4048 invalid. However, $reply_address has leading white space removed, and all
4049 newlines turned into spaces so that the autoreply transport does not
4052 PH/05 If group was specified without a user on a router, and no group or user
4053 was specified on a transport, the group from the router was ignored.
4055 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4056 for visible compile-time settings that can be used to change these
4057 numbers, for those that want even more. Backwards compatibility with old
4058 spool files has been maintained. However, going back to a previous Exim
4059 release will lost any variables that are in spool files.
4061 PH/07 Two small changes when running in the test harness: increase delay when
4062 passing a TCP/IP connection to a new process, in case the original
4063 process has to generate a bounce, and remove special handling of
4064 127.0.0.2 (sic), which is no longer necessary.
4066 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4067 be the same on different OS.
4069 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4072 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4073 whether --show-vars was specified or not
4075 JJ/02 exipick: Added support for new ACL variable spool format introduced
4078 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4079 syntactically invalid From: or Reply-to: line, and a filter used this to
4080 generate an autoreply, and therefore failed to obtain an address for the
4081 autoreply, Exim could try to deliver to a non-existent relative file
4082 name, causing unrelated and misleading errors. What now happens is that
4083 it logs this as a hard delivery error, but does not attempt to create a
4086 PH/11 The exinext utility has a -C option for testing purposes, but although
4087 the given file was scanned by exinext itself; it wasn't being passed on
4088 when Exim was called.
4090 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4091 an end-of-file indication when reading a command response.
4093 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4094 compiled. In many other places in Exim, IPv6 addresses are always
4095 recognized, so I have changed this. It also means that IPv4 domain
4096 literals of the form [IPV4:n.n.n.n] are now always recognized.
4098 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4099 used if the router is not running as root, for example, when verifying at
4100 ACL time, or when using -bh. The debugging output from this situation was
4101 non-existent - all you got was a failure to exec. I have made two
4104 (a) Failures to set uid/gid, the current directory, or a process leader
4105 in a subprocess such as that created by queryprogram now generate
4106 suitable debugging output when -d is set.
4108 (b) The queryprogram router detects when it is not running as root,
4109 outputs suitable debugging information if -d is set, and then runs
4110 the subprocess without attempting to change uid/gid.
4112 PH/15 Minor change to Makefile for building test_host (undocumented testing
4115 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4116 additional section of a DNS packet that returns MX or SRV records.
4117 Instead, it always explicitly searches for A/AAAA records. This avoids
4118 major problems that occur when a DNS server includes only records of one
4119 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4120 fixed another bug: if SRV records were looked up and the corresponding
4121 address records were *not* found in the additional section, the port
4122 values from the SRV records were lost.
4124 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4125 using the correct key (the original address) when searching the retry
4126 rules in order to find which one to use for generating the retry hint.
4128 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4129 adding the default one. Similarly, if it contains a Reply-To: header, the
4130 errors_reply_to option, if set, is not used.
4132 PH/19 When calculating a retry time, Exim used to measure the "time since
4133 failure" by looking at the "first failed" field in the retry record. Now
4134 it does not use this if it is later than than the arrival time of the
4135 message. Instead it uses the arrival time. This makes for better
4136 behaviour in cases where some deliveries succeed, thus re-setting the
4137 "first failed" field. An example is a quota failure for a huge message
4138 when small messages continue to be delivered. Without this change, the
4139 "time since failure" will always be short, possible causing more frequent
4140 delivery attempts for the huge message than are intended.
4141 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4143 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4144 $1, $2, $3) because the numerical variables can be reset during some
4145 expansion items (e.g. "match"), thereby losing the authentication data.
4147 PH/21 Make -bV show the size of off_t variables so that the test suite can
4148 decide whether to run tests for quotas > 2G.
4150 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4151 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4152 filecount value is greater than 2G or if a quota value is greater than 2G
4153 on a system where the size of off_t is not greater than 4, a panic error
4156 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4157 never match. The debug and -bh output now contains an explicit error
4158 message indicating a malformed IPv4 address or mask.
4160 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4161 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4162 PH/23 above applies.
4164 PH/25 Do not write to syslog when running in the test harness. The only
4165 occasion when this arises is a failure to open the main or panic logs
4166 (for which there is an explicit test).
4168 PH/26 Added the /no_tell option to "control=freeze".
4170 PH/27 If a host name lookup failed very early in a connection, for example, if
4171 the IP address matched host_lookup and the reverse lookup yielded a name
4172 that did not have a forward lookup, an error message of the form "no IP
4173 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4174 could be logged. Now it outputs the IP address instead of "NULL".
4176 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4177 allows the sender and the authenticated sender to be set when
4178 submitting a message from within Exim. Since child_open_exim() is
4179 documented for local_scan(), the new function should be too.
4181 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4182 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4183 results in an empty string is now treated as unset.
4185 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4187 PH/31 Added sender_verify_fail logging option.
4189 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4190 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4191 tidied the source and removed it altogether.
4193 PH/33 When a queue run was abandoned because the load average was too high, a
4194 log line was always written; now it is written only if the queue_run log
4195 selector is set. In addition, the log line for abandonment now contains
4196 information about the queue run such as the pid. This is always present
4197 in "start" and "stop" lines but was omitted from the "abandon" line.
4199 PH/34 Omit spaces between a header name and the colon in the error message that
4200 is given when verify = headers_syntax fails (if there are lots of them,
4201 the message gets confusing).
4203 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4204 names, as there are now some PTR records that contain slashes. This check
4205 is only to protect against broken name servers that fall over on strange
4206 characters, so the fact that it applies to all lookups doesn't matter.
4208 PH/36 Now that the new test suite is complete, we can remove some of the
4209 special code in Exim that was needed for the old test suite. For example,
4210 sorting DNS records because real resolvers return them in an arbitrary
4211 order. The new test suite's fake resolver always returns records in the
4214 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4215 bounces) except when queue_only is set, to avoid logging races between
4216 the different processes.
4218 PH/38 Panic-die if .include specifies a non-absolute path.
4220 PH/39 A tweak to the "H" retry rule from its user.
4222 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4223 a label. They prevented compilation on older perls.
4225 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4226 a warning to be raised on newish perls.
4228 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4229 on queue. Changes to match documented behaviour of showing count of
4230 messages matching specified criteria.
4232 PH/40 Changed the default ident timeout from 30s to 5s.
4234 PH/41 Added support for the use of login_cap features, on those BSD systems
4235 that have them, for controlling the resources used by pipe deliveries.
4237 PH/42 The content-scanning code uses fopen() to create files in which to put
4238 message data. Previously it was not paying any attention to the mode of
4239 the files. Exim runs with umask(0) because the rest of the code creates
4240 files with open(), and sets the required mode explicitly. Thus, these
4241 files were ending up world-writeable. This was not a big issue, because,
4242 being within the spool directory, they were not world-accessible. I have
4243 created a function called modefopen, which takes an additional mode
4244 argument. It sets umask(777), creates the file, chmods it to the required
4245 mode, then resets the umask. All the relevant calls to fopen() in the
4246 content scanning code have been changed to use this function.
4248 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4249 to 24 hours. This avoids potential overflow problems when processing G
4250 and H retry rules. I suspect nobody ever tinkers with this value.
4252 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4254 PH/45 When the plaintext authenticator is running as a client, the server's
4255 challenges are checked to ensure they are valid base64 strings. By
4256 default, the authentication attempt is cancelled if an invalid string is
4257 received. Setting client_ignore_invalid_base64 true ignores these errors.
4258 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4259 they are received. Thus, the responses can be made to depend on the
4260 challenges. If an invalid string is ignored, an empty string is placed in
4263 PH/46 Messages that are created by the autoreply transport now contains a
4264 References: header, in accordance with RFCs 2822 and 3834.
4266 PH/47 Added authenticated_sender_force to the smtp transport.
4268 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4270 PH/49 Installed latest patch from the Sieve maintainer.
4272 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4273 was also set, the appendfile transport was unnecessarily scanning a
4274 directory of message files (e.g. for maildir delivery) to find the count
4275 of files (along with the size), even though it did not need this
4276 information. It now does the scan only if it needs to find either the
4277 size of the count of files.
4279 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4281 PH/52 Two bugs concerned with error handling when the smtp transport is
4284 (i) Exim was not creating retry information for temporary errors given
4285 for individual recipients after the DATA command when the smtp transport
4286 was used in LMTP mode. This meant that they could be retried too
4287 frequently, and not timed out correctly.
4289 (ii) Exim was setting the flag that allows error details to be returned
4290 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4291 recipients that were returned after the DATA command.
4293 PH/53 This is related to PH/52, but is more general: for any failing address,
4294 when detailed error information was permitted to be returned to the
4295 sender, but the error was temporary, then after the final timeout, only
4296 "retry timeout exceeded" was returned. Now it returns the full error as
4297 well as "retry timeout exceeded".
4299 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4300 do this, and (what is worse) MTAs that accept it.
4302 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4303 will now be deprecated.
4305 PH/56 New os.c-cygwin from the Cygwin maintainer.
4307 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4308 formats (previously only available in exim formats via -bpr, -bpru,
4309 and -bpra. Now also available in native and exiqgrep formats)
4311 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4312 with very large, slow to parse queues
4314 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4316 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4318 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4319 responses to authentication challenges, though it was showing the
4320 challenges; (ii) I've removed the CR characters from the debug output for
4323 PH/58 Allow for the insertion of a newline as well as a space when a string
4324 is turned into more than one encoded-word during RFC 2047 encoding. The
4325 Sieve code now uses this.
4327 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4328 data_4xx, lost_connection, tls_required.
4330 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4331 message was being sent as an SMTP response.
4333 PH/61 Add -l and -k options to exicyclog.
4335 PH/62 When verifying, if an address was redirected to one new address, so that
4336 verification continued, and the new address failed or deferred after
4337 having set something in $address_data, the value of $address_data was not
4338 passed back to the ACL. This was different to the case when no
4339 redirection occurred. The value is now passed back in both cases.
4341 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4342 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4343 use login_cap.h, so on its own it isn't the distinguishing feature. The
4344 new name refers directly to the setclassresources() function.
4346 PH/65 Added configuration files for NetBSD3.
4348 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4350 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4351 is preferred over IPv4.
4353 PH/68 The bounce_return_message and bounce_return_body options were not being
4354 honoured for bounces generated during the reception of non-SMTP messages.
4355 In particular, this applied to messages rejected by the ACL. This bug has
4356 been fixed. However, if bounce_return_message is true and bounce_return_
4357 body is false, the headers that are returned for a non-SMTP message
4358 include only those that have been read before the error was detected.
4359 (In the case of an ACL rejection, they have all been read.)
4361 PH/69 The HTML version of the specification is now built in a directory called
4362 spec_html instead of spec.html, because the latter looks like a path with
4363 a MIME-type, and this confuses some software.
4365 PH/70 Catch two compiler warnings in sieve.c.
4367 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4368 function verify_get_ident() calls ip_connect() to connect a socket, but
4369 if the "connect()" function timed out, ip_connect() used to close the
4370 socket. However, verify_get_ident() also closes the socket later, and in
4371 between Exim writes to the log, which may get opened at this point. When
4372 the socket was closed in ip_connect(), the log could get the same file
4373 descriptor number as the socket. This naturally causes chaos. The fix is
4374 not to close the socket in ip_connect(); the socket should be closed by
4375 the function that creates it. There was only one place in the code where
4376 this was missing, in the iplookup router, which I don't think anybody now
4377 uses, but I've fixed it anyway.
4379 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4380 well as to direct DNS lookups. Otherwise the handling of names in host
4381 lists is inconsistent and therefore confusing.
4387 PH/01 Two changes to the default runtime configuration:
4389 (1) Move the checks for relay_from_hosts and authenticated clients from
4390 after to before the (commented out) DNS black list checks.
4392 (2) Add control=submission to the relay_from_hosts and authenticated
4393 clients checks, on the grounds that messages accepted by these
4394 statements are most likely to be submissions.
4396 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4398 (1) Generate an error if the third argument for the ${prvs expansion is
4401 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4404 (3) Reset the variables that are obtained from the first argument of
4405 ${prvscheck and used in the second argument before leaving the code,
4406 because their memory is reclaimed, so using them afterwards may do
4409 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4410 one (it's much easier than Tom thought :-).
4412 (5) Because of (4), we can now allow for the use of $prvscheck_result
4413 inside the third argument.
4415 PH/03 For some reason, the default setting of PATH when running a command from
4416 a pipe transport was just "/usr/bin". I have changed it to
4419 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4420 anything to be listed in the output from -bV.
4422 PH/05 When a filter generated an autoreply, the entire To: header line was
4423 quoted in the delivery log line, like this:
4425 => >A.N.Other <ano@some.domain> <original@ddress> ...
4427 This has been changed so that it extracts the operative address. There
4428 may be more than one such address. If so, they are comma-separated, like
4431 => >ano@some.domain,ona@other.domain <original@ddress> ...
4433 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4434 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4435 not being looked up in the rDNS to get a host name, Exim was showing the
4436 IP address twice in Received: lines, even though the IP addresses were
4437 identical. For example:
4439 Received: from [1.2.3.4] (helo=[1.2.3.4])
4441 However, if the real host name was known, it was omitting the HELO data
4442 if it matched the actual IP address. This has been tidied up so that it
4443 doesn't show the same IP address twice.
4445 PH/07 When both +timestamp and +memory debugging was on, the value given by
4446 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4447 called by the debug printing, thereby overwriting the timestamp buffer.
4448 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4450 PH/08 When the original message was included in an autoreply transport, it
4451 always said "this is a copy of the message, including all the headers",
4452 even if body_only or headers_only was set. It now gives an appropriate
4455 PH/09 Applied a patch from the Sieve maintainer which:
4457 o fixes some comments
4458 o adds the (disabled) notify extension core
4459 o adds some debug output for the result of if/elsif tests
4460 o points to the current vacation draft in the documentation
4461 and documents the missing references header update
4465 o fixes a bug in processing the envelope test (when testing
4466 multiple envelope elements, the last element determined the
4469 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4470 Electronic Mail") by including:
4472 Auto-submitted: auto-generated
4474 in the messages that it generates (bounce messages and others, such as
4475 warnings). In the case of bounce messages for non-SMTP messages, there was
4476 also a typo: it was using "Auto_submitted" (underscore instead of
4477 hyphen). Since every message generated by Exim is necessarily in response
4478 to another message, thes have all been changed to:
4480 Auto-Submitted: auto-replied
4482 in accordance with these statements in the RFC:
4484 The auto-replied keyword:
4486 - SHOULD be used on messages sent in direct response to another
4487 message by an automatic process,
4489 - MUST NOT be used on manually-generated messages,
4491 - MAY be used on Delivery Status Notifications (DSNs) and Message
4492 Disposition Notifications (MDNs),
4494 - MUST NOT be used on messages generated by automatic or periodic
4495 processes, except for messages which are automatic responses to
4498 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4499 to the default Received: header definition.
4501 PH/12 Added log selector acl_warn_skipped (default on).
4503 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4504 variables because (a) they are in the wrong storage pool and (b) even if
4505 they were copied, it wouldn't work properly because of the caching.
4507 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4508 checking when decoding. Apparently there are clients that generate
4509 overlong encoded strings. Why am I not surprised?
4511 PH/15 If the first argument of "${if match_address" was not empty, but did not
4512 contain an "@" character, Exim crashed. Now it writes a panic log message
4513 and treats the condition as false.
4515 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4517 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4518 extension "envelope-auth". The code is finished and in agreement with
4519 other implementations, but there is no documentation so far and in fact,
4520 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4521 not changing the active code.
4523 Print executed "if" and "elsif" statements when debugging is used. This
4524 helps a great deal to understand what a filter does.
4526 Document more things not specified clearly in RFC3028. I had all this
4527 sorted out, when out of a sudden new issues came to my mind. Oops."
4529 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4532 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4533 canonical form (as documented). However, after a host name lookup from
4534 the IP address, check_host() was doing a simple string comparison with
4535 addresses acquired from the DNS when checking that the found name did
4536 have the original IP as one of its addresses. Since any found IPv6
4537 addresses are likely to be in abbreviated form, the comparison could
4538 fail. Luckily, there already exists a function for doing the comparison
4539 by converting both addresses to binary, so now that is used instead of
4540 the text comparison.
4542 PH/20 There was another similar case to PH/19, when a complete host name was
4543 given in a host list; looking up its IP address could give an abbreviated
4544 form, whereas the current host's name might or might not be abbreviated.
4545 The same fix has been applied.
4551 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4552 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4555 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4556 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4558 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4560 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4561 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4562 Also, for "Auto-Submitted", treat anything other than "no" as
4563 non-personal, in accordance with RFC 3834. (Previously it treated
4564 anything starting "auto-" as non-personal.)
4566 TF/01 The control=submission/name=... option had a problem with syntax
4567 errors if the name included a slash character. The /name= option
4568 now slurps the rest of the string, so it can include any characters
4569 but it must come last in the list of options (after /sender_retain
4572 PH/05 Some modifications to the interface to the fake nameserver for the new
4580 TK/01 Added the "success_on_redirect" address verification option. See
4581 NewStuff for rationale and an example.
4583 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4585 PH/02 Patch to exigrep to allow it to work on syslog lines.
4587 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4588 fread() to skip over the body file's header line, because in Cygwin the
4589 header line is locked and is inaccessible.
4591 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4592 co-exist for some time) to make it clear that it is the Exim ID that is
4593 referenced, not the Message-ID: header line.
4595 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4596 string_format() function, because snprintf() does not exist on all
4599 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4600 ${stat: expansion item.
4602 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4603 protocol synchronization error", to keep the pedants happy.
4605 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4606 well as for IRIX systems, when gcc is being used. See the host.c source
4609 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4611 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4614 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4615 transport and to the smtp transport in LMTP mode.
4617 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4619 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4620 run a filter in a subprocess. This could lead to confusion in subsequent
4621 lookups in the parent process. There should also be a search_tidyup() at
4622 the end of the subprocess.
4624 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4625 only if the host matched helo_try_verify_hosts, which caused the
4626 verification to occur when the EHLO/HELO command was issued. The ACL just
4627 tested the remembered result. Now, if a previous verification attempt has
4628 not happened, "verify = helo" does it there and then.
4630 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4632 TK/03 Fix log output including CR from clamd.
4634 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4635 exist provoked a memory error which could cause a segfault.
4637 PH/15 Installed PCRE 6.2
4639 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4641 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4642 of the problem. Specifically, suggested +O2 rather than +O1 for the
4645 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4647 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4648 record, the port was not correctly logged when the outgoing_port log
4649 selector was set (it logged the transort's default port).
4651 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4652 fallback_hosts, and "hosts" in the smtp transport.
4654 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4655 host errors such as "Connection refused".
4657 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4658 authentication with radiusclient 0.4.9:
4660 - Error returned from rc_read_config was caught wrongly
4661 - Username/password not passed on to radius server due to wrong length.
4663 The presumption is that some radiusclient API changes for 4.51/PH/17
4664 were not taken care of correctly. The code is still untested by me (my
4665 Linux distribution still has 0.3.2 of radiusclient), but it was
4666 contributed by a Radius user.
4668 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4669 expanding the "port" option of the smtp transport.
4671 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4672 while reading a MIME header. Thanks to Tom Hughes for a patch.
4674 PH/24 Include config.h inside local_scan.h so that configuration settings are
4677 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4678 that in an ACL for RCPT (for example), you can examine exactly what was
4681 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4682 commands, but it was not correctly comparing the address with the actual
4683 client host address. Thus, it would show the EHLO address in Received:
4684 header lines when this was not necessary.
4686 PH/27 Added the % operator to ${eval:}.
4688 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4689 it should be ignoring failures (because with -C, for example, it has lost
4690 privilege). It wasn't ignoring creation failures other than "already
4693 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4696 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4697 because an input line was too long, either on its own, or by virtue of
4698 too many continuations, the temporary file was not being removed, and the
4699 return code was incorrect.
4701 PH/31 Missing "BOOL" in function definition in filtertest.c.
4703 PH/32 Applied Sieve patches from the maintainer.
4705 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4707 PH/33 Added "verify = not_blind".
4709 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4710 Local/Makefile (with some defaults set). These are used in built scripts
4711 such as exicyclog, but they have never been used in the exim_install
4712 script (though there are many overriding facilities there). I have
4713 arranged that the exim_install script now takes note of these two
4716 PH/35 Installed configuration files for Dragonfly.
4718 PH/36 When a locally submitted message by a trusted user did not contain a
4719 From: header, and the sender address was obtained from -f or from an SMTP
4720 MAIL command, and the trusted user did not use -F to supply a sender
4721 name, $originator_name was incorrectly used when constructing a From:
4722 header. Furthermore, $originator_name was used for submission mode
4723 messages from external hosts without From: headers in a similar way,
4724 which is clearly wrong.
4726 PH/37 Added control=suppress_local_fixups.
4728 PH/38 When log_selector = +received_sender was set, and the addition of the
4729 sender made the log line's construction buffer exactly full, or one byte
4730 less than full, an overflow happened when the terminating "\n" was
4733 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4734 when the result of a list match is failure because a DNS lookup failed.
4736 PH/40 RM_COMMAND is now used in the building process.
4738 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4739 the "build-* directories that it finds.
4741 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4742 address in a domain literal was a prefix of an interface address.
4744 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4745 when verifying a sender address, unless rewrite_headers is false.
4747 PH/44 Wrote a long comment about why errors_to addresses are verified as
4748 recipients, not senders.
4750 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4751 the ratelimit ACL was added.
4753 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4755 PH/46 Added extra information about PostgreSQL errors to the error string.
4757 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4758 avoiding the need to install special zones in a real server. This is
4759 backwards compatible; if it can't find the fake resolver, it drops back.
4760 Thus, both old and new test suites can be run.
4762 TF/02 Added util/ratelimit.pl
4764 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4765 clock is set back in time.
4767 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4768 Candler <B.Candler@pobox.com>.
4770 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4771 OK for addresses that are the result of redirections.
4773 PH/48 A number of further additions for the benefit of the new test suite,
4774 including a fake gethostbyname() that interfaces to the fake DNS resolver
4777 TF/06 The fix for widen_domains has also been applied to qualify_single and
4778 search_parents which are the other dnslookup options that can cause
4781 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4784 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4786 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4787 service" if it encountered an MX record with an empty target hostname.
4788 The message is now "an MX or SRV record indicated no SMTP service".
4790 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4791 if the DNS of the sending site is misconfigured. This is quite a
4792 common situation. This change restores the behaviour of treating a
4793 helo verification defer as a failure.
4795 PH/51 If self=fail was set on a router, the bounce message did not include the
4796 actual error message.
4802 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4804 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4805 command itself ended in error, the underlying message about the transport
4806 filter timeout was being overwritten with the pipe command error. Now the
4807 underlying error message should be appended to the second error message.
4809 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4811 PH/02 Reduce the amount of output that "make" produces by default. Full output
4812 can still be requested.
4814 PH/03 The warning log line about a condition test deferring for a "warn" verb
4815 was being output only once per connection, rather than after each
4816 occurrence (because it was using the same function as for successful
4817 "warn" verbs). This seems wrong, so I have changed it.
4819 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4820 it should not have, which might have caused a crash in the right
4821 circumstances, but probably never did.
4823 PH/04 Installed a modified version of Tony Finch's patch to make submission
4824 mode fix the return path as well as the Sender: header line, and to
4825 add a /name= option so that you can make the user's friendly name appear
4828 TF/03 Added the control = fakedefer ACL modifier.
4830 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4831 Mark Lowes for thorough testing.
4833 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4835 TK/03 Merged latest SRS patch from Miles Wilton.
4837 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4838 with the definition in sysexits.h (which is #included earlier).
4839 Fortunately, Exim does not actually use EX_OK. The code used to try to
4840 preserve the sysexits.h value, by assuming that macro definitions were
4841 scanned for macro replacements. I have been disabused of this notion,
4842 so now the code just undefines EX_OK before #including unistd.h.
4844 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4845 in the smtp transport. When a block could not be written in a single
4846 write() function, the timeout was being re-applied to each part-write.
4847 This seems wrong - if the receiver was accepting one byte at a time it
4848 would take for ever. The timeout is now adjusted when this happens. It
4849 doesn't have to be particularly precise.
4851 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4852 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4854 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4855 without a domain if the check to <postmaster@domain> fails.
4857 SC/01 Eximstats: added -xls and the ability to specify output files
4858 (patch written by Frank Heydlauf).
4860 SC/02 Eximstats: use FileHandles for outputting results.
4862 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4864 SC/04 Eximstats: fixed display of large numbers with -nvr option
4866 SC/05 Eximstats: fixed merging of reports with empty tables.
4868 SC/06 Eximstats: added the -include_original_destination flag
4870 SC/07 Eximstats: removed tabs and trailing whitespace.
4872 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4874 TK/06 MBOX spool code: Add real "From " MBOX separator line
4875 so the .eml file is really in mbox format (even though
4876 most programs do not really care). Patch from Alex Miller.
4878 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4879 The latter is generated from $received_to and is only set if the
4880 message has one envelope recipient. SA can use these headers,
4881 obviously out-of-the-box. Patch from Alex Miller.
4883 PH/08 The ${def test on a variable was returning false if the variable's
4884 value was "0", contrary to what the specification has always said!
4885 The result should be true unless the variable is empty.
4887 PH/09 The syntax error of a character other than { following "${if
4888 def:variable_name" (after optional whitespace) was not being diagnosed.
4889 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4890 accidental colon was present, for example, could give incorrect results.
4892 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4893 result is used (not including appendfile, where other changes are about
4896 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4897 This involved changing a lot of size variables from int to off_t. It
4898 should work with maildirs and everything.
4900 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4901 spamd dying while we are connected to it.
4903 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4906 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4907 function for BDB 4.3.
4909 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4911 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4912 the sources for PCRE have been re-arranged and more files are now
4915 PH/15 The code I had for printing potentially long long variables in PH/11
4916 above was not the best (it lost precision). The length of off_t variables
4917 is now inspected at build time, and an appropriate printing format (%ld
4918 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4919 to be "long long int" or "long int". This is needed for the internal
4920 formatting function string_vformat().
4922 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4923 the configuration file to be ":syslog", then the script "guesses" where
4924 the logs files are, rather than using the compiled in default. In our
4925 case the guess is not the same as the compiled default, so the script
4926 suddenly stopped working when I started to use syslog. The patch checks
4927 to see if log_file_path is "". If so, it attempts to read it from exim
4928 with no configuration file to get the compiled in version, before it
4929 falls back to the previous guessing code."
4931 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4932 implementing BATV in an Exim configuration. See NewStuff for the gory
4935 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4936 Makefile that are specific to HP-UX.
4938 PH/18 If the "use_postmaster" option was set for a recipient callout together
4939 with the "random" option, the postmaster address was used as the MAIL
4940 FROM address for the random test, but not for the subsequent recipient
4941 test. It is now used for both.
4943 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4944 patch removes a few documentation additions to RFC 3028, because the
4945 latest draft now contains them. It adds the new en;ascii-case comparator
4946 and a new error check for 8bit text in MIME parts. Comparator and
4947 require names are now matched exactly. I enabled the subaddress
4948 extension, but it is not well tested yet (read: it works for me)."
4950 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4951 rework some of the code of TK/09 above to avoid the hardwired use of
4952 "%lld" and "long long". Replaced the call to snprintf() with a call to
4955 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4956 records point to non-existent hosts", "retry timeout exceeded", and
4957 "retry time not reached for any host after a long failure period".
4959 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4960 experimental DomainKeys support:
4962 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4963 (2) On an error such as an illegally used "control", the wrong name for
4964 the control was given.
4966 These problems did NOT occur unless DomainKeys support was compiled.
4968 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4970 PH/24 Added ${if match_ip condition.
4972 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4973 databases so that it will be absolutely obvious if a crash occurs in the
4974 DB library. This is a regular occurrence (often caused by mis-matched
4977 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4978 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4979 picked up on a user's system that detects such things. There doesn't seem
4980 to be a gcc warning option for this - only an attribute that has to be
4981 put on the function's prototype. It seems that in Fedora Core 4 they have
4982 set this on a number of new functions. No doubt there will be more in due
4985 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4986 specify a transport. However, if an address that was verified by such a
4987 router was the subject of a callout, Exim crashed because it tried to
4988 read the rcpt_include_affixes from the non-existent transport. Now it
4989 just assumes that the setting of that option is false. This bug was
4990 introduced by 4.51/PH/31.
4992 PH/28 Changed -d+all to exclude +memory, because that information is very
4993 rarely of interest, but it makes the output a lot bigger. People tend to
4994 do -d+all out of habit.
4996 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4997 code in os-type was giving problems when libc.so lives in lib64, like on
5000 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5001 aren't the modern standard, and it seems that some systems' include files
5002 don't always have them. Exim was already checking for some of the newer
5003 ones like T_AAAA, and defining it itself. I've added checks for all the
5004 record types that Exim uses.
5006 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5007 not automatically generating a new one, as it is supposed to. This
5008 prevented TLS from working. If the file did exist, but contained invalid
5009 data, a new version was generated, as expected. It was only the case of a
5010 non-existent file that was broken.
5012 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5013 with a change in libdomainkeys > 0.64.
5015 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5016 from DNS. If the selector record carries the flag, it now has
5017 precedence over the domain-wide flag.
5019 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5021 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5022 the use of a port name that isn't defined in /etc/services) occurred, the
5023 message was deferred as in a normal delivery, and thus remained on the
5024 spool, instead of being failed because of the mua_wrapper setting. This
5025 is now fixed, and I tidied up some of the mua_wrapper messages at the
5028 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5029 the messages in a hash of arrays rather than using individual hashes.
5030 This is a bit cleaner and results in dramatic memory savings, albeit
5031 at a slight CPU cost.
5033 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5034 as requested by Marc Sherman.
5036 SC/10 Eximstats: added histograms for user specified patterns as requested
5039 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5041 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5042 fopen() in the content-scanning modules that did not already have it.
5048 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5049 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5051 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5053 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5055 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5056 to test Sieve filters that use "vacation".
5058 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5059 that changes the way the GnuTLS parameters are stored in the cache file.
5060 The new format can be generated externally. For backward compatibility,
5061 if the data in the cache doesn't make sense, Exim assumes it has read an
5062 old-format file, and it generates new data and writes a new file. This
5063 means that you can't go back to an older release without removing the
5066 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5067 work if there are any delivery delays because "one_time" forces the
5068 parent to be marked "delivered", so its unseen clone is never tried
5069 again. For this reason, Exim now forbids the simultaneous setting of
5072 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5073 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5074 entry if you want to know the details. The fix, however, neglected to
5075 consider the case when local delivery batching is involved. The test for
5076 "previously delivered" was not happening when checking to see if an
5077 address could be batched with a previous (undelivered) one; under
5078 certain circumstances this could lead to multiple deliveries to the same
5081 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5082 in its include files, and this causes problems building Exim.
5084 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5085 header_syntax) but Exim was just ignoring anything given after a slash.
5086 In particular, this caused confusion with an attempt to use "verify =
5087 reverse_host_lookup/defer_ok". An error is now given when options are
5088 supplied for verify items that do not have them. (Maybe reverse_host_
5089 lookup should have a defer_ok option, but that's a different point.)
5091 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5092 defined by RFC 821) to 2048, because there were problems with some AUTH
5093 commands, and RFC 1869 says the size should be increased for extended
5094 SMTP commands that take arguments.
5096 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5099 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5100 "unknown" error; now it says that the functionality isn't in the binary.
5102 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5103 an address' error message when a string expansion fails (syntax or
5104 whatever). Otherwise the password may appear in the log. Following change
5105 PH/42 below, there is no longer a chance of it appearing in a bounce
5108 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5110 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5111 of its addresses was ever tried. (Bugzilla bug #2.)
5113 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5114 the result incorrectly in the debug output. (It correctly added a newline
5115 to what was transported.)
5117 TF/01 Added $received_time.
5119 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5120 commented out examples of how to interface to a virus scanner and to
5121 SpamAssassin. Also added commented examples of av_scanner and
5122 spamd_address settings.
5124 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5125 and controls are allowed in which ACLs. There were a couple of minor
5126 errors. Some of the entries in the conditions table (which is a table of
5127 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5128 negation of where the condition IS allowed.
5130 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5132 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5133 header file does not have a version number, so I've had to invent a new
5134 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5135 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5136 radiusclient), but it was contributed by a Radius user.
5138 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5139 files or directories, for OpenSSL.
5141 PH/19 When an Exim process that is running as root has to create an Exim log
5142 file, it does so in a subprocess that runs as exim:exim so as to get the
5143 ownership right at creation (otherwise, other Exim processes might see
5144 the file with the wrong ownership). There was no test for failure of this
5145 fork() call, which would lead to the process getting stuck as it waited
5146 for a non-existent subprocess. Forks do occasionally fail when resources
5147 run out. I reviewed all the other calls to fork(); they all seem to check
5150 PH/20 When checking for unexpected SMTP input at connect time (before writing
5151 the banner), Exim was not dealing correctly with a non-positive return
5152 from the read() function. If the client had disconnected by this time,
5153 the result was a log entry for a synchronization error with an empty
5154 string after "input=" when read() returned zero. If read() returned -1
5155 (an event I could not check), uninitialized data bytes were printed.
5156 There were reports of junk text (parts of files, etc) appearing after
5159 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5161 PH/22 Added support for macro redefinition, and (re)definition in between
5162 driver and ACL definitions.
5164 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5165 forgetting to use the resulting value; it was using the unexpanded value.
5167 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5168 hadn't been configured. The fix is from Juergen Kreileder, who
5169 understands it better than I do:
5171 "Here's what I see happening with three configured cyrus_sasl
5172 authenticators configured (plain, login, cram-md5):
5174 On startup auth_cyrus_sasl_init() gets called for each of these.
5175 This means three calls to sasl_listmech() without a specified mech_list.
5176 => SASL tests which mechs of all available mechs actually work
5177 => three warnings about OTP not working
5178 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5180 With the patch, sasl_listmech() also gets called three times. But now
5181 SASL's mech_list option is set to the server_mech specified in the the
5182 authenticator. Or in other words, the answer from sasl_listmech()
5183 gets limited to just the mech you're testing for (which is different
5185 => the return list contains just 'plain' or 'login', 'cram-md5' or
5186 nothing depending on the value of ob->server_mech.
5188 I've just tested the patch: Authentication still works fine,
5189 unavailable mechs specified in the exim configuration are still
5190 caught, and the auth.log warnings about OTP are gone."
5192 PH/25 When debugging is enabled, the contents of the command line are added
5193 to the debugging output, even when log_selector=+arguments is not
5196 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5197 answer is "GNU", and only if the return is "GNU/something" is the answer
5200 PH/27 $acl_verify_message is now set immediately after the failure of a
5201 verification in an ACL, and so is available in subsequent modifiers. In
5202 particular, the message can be preserved by coding like this:
5204 warn !verify = sender
5205 set acl_m0 = $acl_verify_message
5207 Previously, $acl_verify_message was set only while expanding "message"
5208 and "log_message" when a very denied access.
5210 PH/28 Modified OS/os.c-Linux with
5212 -#ifndef OS_LOAD_AVERAGE
5213 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5215 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5216 nomenclature these days.)
5218 PH/29 Installed patch from the Sieve maintainer that adds the options
5219 sieve_useraddress and sieve_subaddress to the redirect router.
5221 PH/30 In these circumstances:
5222 . Two addresses routed to the same list of hosts;
5223 . First host does not offer TLS;
5224 . First host accepts first address;
5225 . First host gives temporary error to second address;
5226 . Second host offers TLS and a TLS session is established;
5227 . Second host accepts second address.
5228 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5229 and peerdn, if requested) that were in fact used only for the second
5232 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5233 attention to any local part prefix or suffix that was matched by the
5234 router that accepted the address. It now behaves in the same way as it
5235 does for delivery: the affixes are removed from the local part unless
5236 rcpt_include_affixes is set on the transport.
5238 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5239 timeout during the DATA phase of an incoming message.
5241 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5242 applied a patch sanctioned by the Sieve maintainer.
5244 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5245 the uid or gid is negative. A case of a negative gid caused this to be
5246 noticed. The fix allows for either to be negative.
5248 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5249 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5252 PH/36 The change PH/12 above was broken. Fixed it.
5254 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5255 the grounds that routing the same address twice would always produce the
5256 same answer. This might have been true once, but it is certainly no
5257 longer true now. Routing a child address may depend on the previous
5258 routing that produced that child. Some complicated redirection strategies
5259 went wrong when messages had multiple recipients, and made Exim's
5260 behaviour dependent on the order in which the addresses were given.
5262 I have moved the duplicate checking until after the routing is complete.
5263 Exim scans the addresses that are assigned to local and remote
5264 transports, and removes any duplicates. This means that more work will be
5265 done, as duplicates will always all be routed, but duplicates are
5266 presumably rare, so I don't expect this is of any significance.
5268 For deliveries to pipes, files, and autoreplies, the duplicate checking
5269 still happens during the routing process, since they are not going to be
5272 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5273 It corrects a timeout issue with spamd. This is Ian's comment: "The
5274 background is that sometimes spamd either never reads data from a
5275 connection it has accepted, or it never writes response data. The exiscan
5276 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5277 blindly assumes that writes won't block so it may never time out."
5279 PH/39 Allow G after quota size as well as K and M.
5281 PH/40 The value set for $authenticated_id in an authenticator may not contain
5282 binary zeroes or newlines because the value is written to log lines and
5283 to spool files. There was no check on this. Now the value is run through
5284 the string_printing() function so that such characters are converted to
5285 printable escape sequences.
5287 PH/41 $message_linecount is a new variable that contains the total number of
5288 lines in the message. Compare $body_linecount, which is the count for the
5291 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5292 bounce and delay warning messages, except in certain special cases, which
5295 (a) An SMTP error message from a remote host;
5296 (b) A message specified in a :fail: redirection;
5297 (c) A message specified in a "fail" command in a system filter;
5298 (d) A message specified in a FAIL return from the queryprogram router;
5299 (e) A message specified by the cannot_route_message router option.
5301 In these cases only, Exim does include the error details in bounce and
5302 warning messages. There are also a few cases where bland messages such
5303 as "unrouteable address" or "local delivery error" are given.
5305 PH/43 $value is now also set for the "else" part of a ${run expansion.
5307 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5308 being worked on, but at least Exim now implements the latest version to
5311 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5312 process to complete was treated as a delivery failure, a timeout while
5313 writing the message to the pipe was logged, but erroneously treated as a
5314 successful delivery. Such timeouts include transport filter timeouts. For
5315 consistency with the overall process timeout, these timeouts are now
5316 treated as errors, giving rise to delivery failures by default. However,
5317 there is now a new Boolean option for the pipe transport called
5318 timeout_defer, which, if set TRUE, converts the failures into defers for
5319 both kinds of timeout. A transport filter timeout is now identified in
5322 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5323 systems where "make" and "gmake" are different, calling "gmake" at top
5324 level broke things. I've arranged for the value of $(MAKE) to be passed
5325 from the Makefile to this script so that it can call the same version of
5329 A note about Exim versions 4.44 and 4.50
5330 ----------------------------------------
5332 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5333 changes of various kinds. As a consequence, a big documentation update was
5334 needed. This delayed the release for rather longer than seemed good, especially
5335 in the light of a couple of (minor) security issues. Therefore, the changes
5336 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5337 release. So 4.44 and 4.50 are in effect two different branches that both start
5340 I have left the 4.50 change log unchanged; it contains all the changes since
5341 4.43. The change log for 4.44 is below; many of its items are identical to
5342 those for 4.50. This seems to be the most sensible way to preserve the
5343 historical information.
5349 1. Minor wording change to the doc/README.SIEVE file.
5351 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5352 computation of the current number of files was incorrect.
5354 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5355 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5358 4. Give more explanation in the error message when the command for a transport
5359 filter fails to execute.
5361 5. There are several places where Exim runs a non-Exim command in a
5362 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5363 was being done only for the command run by the queryprogram router. It is
5364 now done for all such subprocesses. The other cases are: ${run, transport
5365 filters, and the commands run by the lmtp and pipe transports.
5367 6. Added CONFIGURE_GROUP build-time option.
5369 7. Some older OS have a limit of 256 on the maximum number of file
5370 descriptors. Exim was using setrlimit() to set 1000 as a large value
5371 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5372 systems. I've change it so that if it can't get 1000, it tries for 256.
5374 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5375 was an oversight, and furthermore, ever since the addition of extra
5376 controls (e.g. 4.43/32), the checks on when to allow different forms of
5377 "control" were broken. There should now be diagnostics for all cases when a
5378 control that does not make sense is encountered.
5380 9. Added the /retain_sender option to "control=submission".
5382 10. $recipients is now available in the predata ACL (oversight).
5384 11. Tidy the search cache before the fork to do a delivery from a message
5385 received from the command line. Otherwise the child will trigger a lookup
5386 failure and thereby defer the delivery if it tries to use (for example) a
5387 cached ldap connection that the parent has called unbind on.
5389 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5390 of $address_data from the recipient verification was clobbered by the
5391 sender verification.
5393 13. The value of address_data from a sender verification is now available in
5394 $sender_address_data in subsequent conditions in the ACL statement.
5396 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5398 15. Added a new option "connect=<time>" to callout options, to set a different
5401 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5402 was its contents. (It was OK if the option was not defined at all.)
5404 17. A "Completed" log line is now written for messages that are removed from
5405 the spool by the -Mrm option.
5407 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5408 information about exactly what failed.
5410 19. Added -dd to debug only the daemon process.
5412 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5413 handles timeouts, both on the server side and network timeouts. Renamed the
5414 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5416 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5417 It is now set to "smtps".
5419 22. $host_address is now set to the target address during the checking of
5420 ignore_target_hosts.
5422 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5423 being passed; this would have caused $sender_host_name to have been used if
5424 matching the list had actually called for a host name (not very likely,
5425 since this list is usually IP addresses). A host name is now passed as
5428 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5429 code that specifies a non-restarting handler (typically sigaction() in
5430 modern systems) in an attempt to fix a rare and obscure crash bug.
5432 25. Narrowed the window for a race in the daemon that could cause it to ignore
5433 SIGCHLD signals. This is not a major problem, because they are used only to
5434 wake it up if nothing else does.
5436 26. A malformed maildirsize file could cause Exim to calculate negative values
5437 for the mailbox size or file count. Odd effects could occur as a result.
5438 The maildirsize information is now recalculated if the size or filecount
5441 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5442 support for a long time. Removed HAVE_SYS_VFS_H.
5444 28. Installed the latest version of exipick from John Jetmore.
5446 29. In an address list, if the pattern was not a regular expression, an empty
5447 subject address (from a bounce message) matched only if the pattern was an
5448 empty string. Non-empty patterns were not even tested. This was the wrong
5449 because it is perfectly reasonable to use an empty address as part of a
5450 database query. An empty address is now tested by patterns that are
5451 lookups. However, all the other forms of pattern expect the subject to
5452 contain a local part and a domain, and therefore, for them, an empty
5453 address still always fails if the pattern is not itself empty.
5455 30. Exim went into a mad DNS loop when attempting to do a callout where the
5456 host was specified on an smtp transport, and looking it up yielded more
5457 than one IP address.
5459 31. Re-factored the code for checking spool and log partition space into a
5460 function that finds that data and another that does the check. The former
5461 is then used to implement four new variables: $spool_space, $log_space,
5462 $spool_inodes, and $log_inodes.
5464 32. The RFC2047 encoding function was originally intended for short strings
5465 such as real names; it was not keeping to the 75-character limit for
5466 encoded words that the RFC imposes. It now respects the limit, and
5467 generates multiple encoded words if necessary. To be on the safe side, I
5468 have increased the buffer size for the ${rfc2047: expansion operator from
5471 33. It is now permitted to omit both strings after an "if" condition; if the
5472 condition is true, the result is "true". As before, when the second string
5473 is omitted, a false condition yields an empty string. This makes it less
5474 cumbersome to write custom ACL and router conditions.
5476 34. Failure to deliver a bounce message always caused it to be frozen, even if
5477 there was an errors_to setting on the router. The errors_to setting is now
5480 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5481 canonical form (fully expanded) before being placed in
5482 $sender_host_address.
5484 36. The table in the code that translates DNS record types into text (T_A to
5485 "A" for instance) was missing entries for NS and CNAME. It is just possible
5486 that this could have caused confusion if both these types were looked up
5487 for the same domain, because the text type is used as part of Exim's
5488 per-process caching. But the chance of anyone hitting this buglet seems
5491 37. The dnsdb lookup has been extended in a number of ways.
5493 (1) There is a new type, "zns", which walks up the domain tree until it
5494 finds some nameserver records. It should be used with care.
5496 (2) There is a new type, "mxh", which is like "mx" except that it returns
5497 just the host names, not the priorities.
5499 (3) It is now possible to give a list of domains (or IP addresses) to be
5500 looked up. The behaviour when one of the lookups defers can be
5501 controlled by a keyword.
5503 (4) It is now possible to specify the separator character for use when
5504 multiple records are returned.
5506 38. The dnslists ACL condition has been extended: it is now possible to supply
5507 a list of IP addresses and/or domains to be looked up in a particular DNS
5510 39. Added log_selector=+queue_time_overall.
5512 40. When running the queue in the test harness, wait just a tad after forking a
5513 delivery process, to get repeatability of debugging output.
5515 41. Include certificate and key file names in error message when GnuTLS fails
5516 to set them up, because the GnuTLS error message doesn't include the name
5517 of the failing file when there is a problem reading it.
5519 42. Allow both -bf and -bF in the same test run.
5521 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5523 44. The "Exiscan patch" is now merged into the mainline Exim source.
5525 45. Sometimes the final signoff response after QUIT could fail to get
5526 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5527 < 0 before doing a fflush(). This bug looks as though it goes back to the
5528 introduction of TLS in release 3.20, but "sometimes" must have been rare
5529 because the tests only now provoked it.
5531 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5532 (this can affect the format of dates).
5534 47. exim_tidydb, when checking for the continued existence of a message for
5535 which it has found a message-specific retry record, was not finding
5536 messages that were in split spool directories. Consequently, it was
5537 deleting retry records that should have stayed in existence.
5539 48. Steve fixed some bugs in eximstats.
5541 49. The SPA authentication driver was not abandoning authentication and moving
5542 on to the next authenticator when an expansion was forced to fail,
5543 contradicting the general specification for all authenticators. Instead it
5544 was generating a temporary error. It now behaves as specified.
5546 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5547 (the order specifies the preference for clients). The order is now AES256,
5548 AES128, 3DES, ARCFOUR128.
5550 51. Small patch to Sieve code - explicitly set From: when generating an
5553 52. Exim crashed if a remote delivery caused a very long error message to be
5554 recorded - for instance if somebody sent an entire SpamAssassin report back
5555 as a large number of 550 error lines. This bug was coincidentally fixed by
5556 increasing the size of one of Exim's internal buffers (big_buffer) that
5557 happened as part of the Exiscan merge. However, to be on the safe side, I
5558 have made the code more robust (and fixed the comments that describe what
5561 53. Now that there can be additional text after "Completed" in log lines (if
5562 the queue_time_overall log selector is set), a one-byte patch to exigrep
5563 was needed to allow it to recognize "Completed" as not the last thing in
5566 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5567 patch that reportedly fixes this has been added. I am not expert enough to
5568 create a test for it. This is what the patch creator wrote:
5570 "I found a little strange behaviour of ldap code when working with
5571 Windows 2003 AD Domain, where users was placed in more than one
5572 Organization Units. When I tried to give exim partial DN, the exit code
5573 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5574 But simultaneously result of request was absolutely normal ldap result,
5575 so I produce this patch..."
5577 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5578 REFERENCE, so I have modified the code to exclude the patch when that macro
5581 55. Some experimental protocols are using DNS PTR records for new purposes. The
5582 keys for these records are domain names, not reversed IP addresses. The
5583 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5584 leaves it alone. Component reversal etc. now happens only for IP addresses.
5587 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5589 57. Double the size of the debug message buffer (to 2048) so that more of very
5590 long debug lines gets shown.
5592 58. The exicyclog utility now does better if the number of log files to keep
5593 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5595 59. Two changes related to the smtp_active_hostname option:
5597 (1) $smtp_active_hostname is now available as a variable.
5598 (2) The default for smtp_banner uses $smtp_active_hostname instead
5599 of $primary_hostname.
5601 60. The host_aton() function is supposed to be passed a string that is known
5602 to be a valid IP address. However, in the case of IPv6 addresses, it was
5603 not checking this. This is a hostage to fortune. Exim now panics and dies
5604 if the condition is not met. A case was found where this could be provoked
5605 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5606 components; fortuitously, this particular loophole had already been fixed
5607 by change 4.50/55 above.
5609 If there are any other similar loopholes, the new check in host_aton()
5610 itself should stop them being exploited. The report I received stated that
5611 data on the command line could provoke the exploit when Exim was running as
5612 exim, but did not say which command line option was involved. All I could
5613 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5614 running as the user.
5617 61. There was a buffer overflow vulnerability in the SPA authentication code
5618 (which came originally from the Samba project). I have added a test to the
5619 spa_base64_to_bits() function which I hope fixes it.
5622 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5623 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5625 63. The daemon start-up calls getloadavg() while still root for those OS that
5626 need the first call to be done as root, but it missed one case: when
5627 deliver_queue_load_max is set with deliver_drop_privilege. This is
5628 necessary for the benefit of the queue runner, because there is no re-exec
5629 when deliver_drop_privilege is set.
5631 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5632 This has been fixed.
5634 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5635 was in use, was not putting the data itself into the right store pool;
5636 consequently, it could be overwritten for a subsequent message in the same
5637 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5640 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5642 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5643 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5644 the function were treating the return as a boolean value, which happened to
5645 work because 0=false and not-0=true, but is not correct code.
5647 68. The host_aton() function was not handling scoped IPv6 addresses (those
5648 with, for example, "%eth0" on the end) correctly.
5650 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5651 negated items (that is, ~something) in unsigned ints. Some compilers
5652 apparently mutter when there is no cast.
5654 70. If an address verification called from an ACL failed, and did not produce a
5655 user-specific message (i.e. there was only a "system" message), nothing was
5656 put in $acl_verify_message. In this situation, it now puts the system
5659 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5660 session; change 4.31/43 added the unwanted input to the log line - except
5661 that it did not do this in the start of session case. It now does.
5663 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5664 This is wrong and can cause the other end to generate a synchronization
5665 error if it is another Exim or anything else that does the synchronization
5666 check. A QUIT command is no longer sent after a timeout.
5668 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5669 during host lookups.
5671 74. The defer_ok option of callout verification was not working if it was used
5672 when verifying addresses in header lines, that is, for this case:
5674 verify = header_sender/callout=defer_ok
5676 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5677 those file descriptors could be used for SMTP connections. If anything
5678 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5679 could be sent to the SMTP client, causing chaos. The daemon now opens
5680 stdin, stdout, and stderr to /dev/null when it puts itself into the
5683 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5684 log by default. The user can override this with suitable Perl magic.
5686 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5687 the log message when discard triggers, was not working for the DATA ACL or
5688 for the non-SMTP ACL.
5690 78. Error message wording change in sieve.c.
5692 79. If smtp_accept_max_per_host was set, the number of connections could be
5693 restricted to fewer than expected, because the daemon was trying to set up
5694 a new connection before checking whether the processes handling previous
5695 connections had finished. The check for completed processes is now done
5696 earlier. On busy systems, this bug wouldn't be noticed because something
5697 else would have woken the daemon, and it would have reaped the completed
5700 80. If a message was submitted locally by a user whose login name contained one
5701 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5702 It caused a spool format error. I have fixed the spool reading code. A
5703 related problem was that the "from" clause in the Received: line became
5704 illegal because of the space(s). It is now covered by ${quote_local_part.
5706 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5709 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5711 83. Patch from Sieve maintainer to fix unterminated string problem in
5712 "vacation" handling.
5714 84. Some minor changes to the Linux configuration files to help with other
5715 OS variants using glibc.
5717 85. One more patch for Sieve to update vacation handling to latest spec.
5720 ----------------------------------------------------
5721 See the note above about the 4.44 and 4.50 releases.
5722 ----------------------------------------------------
5728 1. Change 4.43/35 introduced a bug that caused file counts to be
5729 incorrectly computed when quota_filecount was set in an appendfile
5732 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5733 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5736 3. Give more explanation in the error message when the command for a transport
5737 filter fails to execute.
5739 4. There are several places where Exim runs a non-Exim command in a
5740 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5741 was being done only for the command run by the queryprogram router. It is
5742 now done for all such subprocesses. The other cases are: ${run, transport
5743 filters, and the commands run by the lmtp and pipe transports.
5745 5. Some older OS have a limit of 256 on the maximum number of file
5746 descriptors. Exim was using setrlimit() to set 1000 as a large value
5747 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5748 systems. I've change it so that if it can't get 1000, it tries for 256.
5750 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5751 was an oversight, and furthermore, ever since the addition of extra
5752 controls (e.g. 4.43/32), the checks on when to allow different forms of
5753 "control" were broken. There should now be diagnostics for all cases when a
5754 control that does not make sense is encountered.
5756 7. $recipients is now available in the predata ACL (oversight).
5758 8. Tidy the search cache before the fork to do a delivery from a message
5759 received from the command line. Otherwise the child will trigger a lookup
5760 failure and thereby defer the delivery if it tries to use (for example) a
5761 cached ldap connection that the parent has called unbind on.
5763 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5764 of $address_data from the recipient verification was clobbered by the
5765 sender verification.
5767 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5768 was its contents. (It was OK if the option was not defined at all.)
5770 11. A "Completed" log line is now written for messages that are removed from
5771 the spool by the -Mrm option.
5773 12. $host_address is now set to the target address during the checking of
5774 ignore_target_hosts.
5776 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5777 being passed; this would have caused $sender_host_name to have been used if
5778 matching the list had actually called for a host name (not very likely,
5779 since this list is usually IP addresses). A host name is now passed as
5782 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5783 code that specifies a non-restarting handler (typically sigaction() in
5784 modern systems) in an attempt to fix a rare and obscure crash bug.
5786 15. Narrowed the window for a race in the daemon that could cause it to ignore
5787 SIGCHLD signals. This is not a major problem, because they are used only to
5788 wake it up if nothing else does.
5790 16. A malformed maildirsize file could cause Exim to calculate negative values
5791 for the mailbox size or file count. Odd effects could occur as a result.
5792 The maildirsize information is now recalculated if the size or filecount
5795 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5796 support for a long time. Removed HAVE_SYS_VFS_H.
5798 18. Updated exipick to current release from John Jetmore.
5800 19. Allow an empty sender to be matched against a lookup in an address list.
5801 Previously the only cases considered were a regular expression, or an
5804 20. Exim went into a mad DNS lookup loop when doing a callout where the
5805 host was specified on the transport, if the DNS lookup yielded more than
5808 21. The RFC2047 encoding function was originally intended for short strings
5809 such as real names; it was not keeping to the 75-character limit for
5810 encoded words that the RFC imposes. It now respects the limit, and
5811 generates multiple encoded words if necessary. To be on the safe side, I
5812 have increased the buffer size for the ${rfc2047: expansion operator from
5815 22. Failure to deliver a bounce message always caused it to be frozen, even if
5816 there was an errors_to setting on the router. The errors_to setting is now
5819 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5820 canonical form (fully expanded) before being placed in
5821 $sender_host_address.
5823 24. Updated eximstats to version 1.33
5825 25. Include certificate and key file names in error message when GnuTLS fails
5826 to set them up, because the GnuTLS error message doesn't include the name
5827 of the failing file when there is a problem reading it.
5829 26. Expand error message when OpenSSL has problems setting up cert/key files.
5832 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5833 (this can affect the format of dates).
5835 28. exim_tidydb, when checking for the continued existence of a message for
5836 which it has found a message-specific retry record, was not finding
5837 messages that were in split spool directories. Consequently, it was
5838 deleting retry records that should have stayed in existence.
5840 29. eximstats updated to version 1.35
5841 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5842 1.35 - bugfix such that pie charts by volume are generated correctly
5844 30. The SPA authentication driver was not abandoning authentication and moving
5845 on to the next authenticator when an expansion was forced to fail,
5846 contradicting the general specification for all authenticators. Instead it
5847 was generating a temporary error. It now behaves as specified.
5849 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5850 (the order specifies the preference for clients). The order is now AES256,
5851 AES128, 3DES, ARCFOUR128.
5853 31. Small patch to Sieve code - explicitly set From: when generating an
5856 32. Exim crashed if a remote delivery caused a very long error message to be
5857 recorded - for instance if somebody sent an entire SpamAssassin report back
5858 as a large number of 550 error lines. This bug was coincidentally fixed by
5859 increasing the size of one of Exim's internal buffers (big_buffer) that
5860 happened as part of the Exiscan merge. However, to be on the safe side, I
5861 have made the code more robust (and fixed the comments that describe what
5864 33. Some experimental protocols are using DNS PTR records for new purposes. The
5865 keys for these records are domain names, not reversed IP addresses. The
5866 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5867 leaves it alone. Component reversal etc. now happens only for IP addresses.
5870 34. The host_aton() function is supposed to be passed a string that is known
5871 to be a valid IP address. However, in the case of IPv6 addresses, it was
5872 not checking this. This is a hostage to fortune. Exim now panics and dies
5873 if the condition is not met. A case was found where this could be provoked
5874 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5875 components; fortuitously, this particular loophole had already been fixed
5876 by change 4.50/55 or 4.44/33 above.
5878 If there are any other similar loopholes, the new check in host_aton()
5879 itself should stop them being exploited. The report I received stated that
5880 data on the command line could provoke the exploit when Exim was running as
5881 exim, but did not say which command line option was involved. All I could
5882 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5883 running as the user.
5886 35. There was a buffer overflow vulnerability in the SPA authentication code
5887 (which came originally from the Samba project). I have added a test to the
5888 spa_base64_to_bits() function which I hope fixes it.
5891 36. The daemon start-up calls getloadavg() while still root for those OS that
5892 need the first call to be done as root, but it missed one case: when
5893 deliver_queue_load_max is set with deliver_drop_privilege. This is
5894 necessary for the benefit of the queue runner, because there is no re-exec
5895 when deliver_drop_privilege is set.
5897 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5898 was in use, was not putting the data itself into the right store pool;
5899 consequently, it could be overwritten for a subsequent message in the same
5900 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5903 38. Sometimes the final signoff response after QUIT could fail to get
5904 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5905 < 0 before doing a fflush(). This bug looks as though it goes back to the
5906 introduction of TLS in release 3.20, but "sometimes" must have been rare
5907 because the tests only now provoked it.
5913 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5914 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5915 Now it returns an integer. A number of calls were still expecting a T/F
5916 return. Fortuitously, in all cases, the tests worked in OK situations,
5917 which is the norm. However, things would have gone wrong on any write
5918 failures on the smtp file descriptor. This function is used when sending
5919 messages over SMTP and also when doing verify callouts.
5921 2. When Exim is called to do synchronous delivery of a locally submitted
5922 message (the -odf or -odi options), it no longer closes stderr before doing
5925 3. Implemented the mua_wrapper option.
5927 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5929 5. Implemented the functions header_remove(), header_testname(),
5930 header_add_at_position(), and receive_remove_recipient(), and exported them
5933 6. If an ACL "warn" statement specified the addition of headers, Exim already
5934 inserted X-ACL-Warn: at the start if there was no header name. However, it
5935 was not making this test for the second and subsequent header lines if
5936 there were newlines in the string. This meant that an invalid header could
5937 be inserted if Exim was badly configured.
5939 7. Allow an ACL "warn" statement to add header lines at the start or after all
5940 the Received: headers, as well as at the end.
5942 8. Added the rcpt_4xx retry error code.
5944 9. Added postmaster_mailfrom=xxx to callout verification option.
5946 10. Added mailfrom=xxxx to the callout verification option, for verify=
5949 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5950 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5952 12. Inserted some casts to stop certain compilers warning when using pointer
5953 differences as field lengths or precisions in printf-type calls (mostly
5954 affecting debugging statements).
5956 13. Added optional readline() support for -be (dynamically loaded).
5958 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5959 same clock tick as a message's arrival, so that its received time was the
5960 same as the "first fail" time on the retry record, and that message
5961 remained on the queue past the ultimate address timeout, every queue runner
5962 would try a delivery (because it was past the ultimate address timeout) but
5963 after another failure, the ultimate address timeout, which should have then
5964 bounced the address, did not kick in. This was a "< instead of <=" error;
5965 in most cases the first failure would have been in the next clock tick
5966 after the received time, and all would be well.
5968 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5969 being recognized when the domain list was tested by the match_domain
5970 condition in an expansion string.
5972 16. Added the ${str2b64: operator.
5974 17. Exim was always calling setrlimit() to set a large limit for the number of
5975 processes, without checking whether the existing limit was already
5976 adequate. (It did check for the limit on file descriptors.) Furthermore,
5977 errors from getrlimit() and setrlimit() were being ignored. Now they are
5978 logged to the main and panic logs, but Exim does carry on, to try to do its
5979 job under whatever limits there are.
5981 18. Imported PCRE 5.0.
5983 19. Trivial typo in log message " temporarily refused connection" (the leading
5986 20. If the log selector return_path_on_delivery was set and an address was
5987 redirected to /dev/null, the delivery process crashed because it assumed
5988 that a return path would always be set for a "successful" delivery. In this
5989 case, the whole delivery is bypassed as an optimization, and therefore no
5992 21. Internal re-arrangement: the function for sending a challenge and reading
5993 a response while authentication was assuming a zero-terminated challenge
5994 string. It's now changed to take a pointer and a length, to allow for
5995 binary data in such strings.
5997 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5999 23. Exim was not respecting finduser_retries when seeking the login of the
6000 uid under which it was called; it was always trying 10 times. (The default
6001 setting of finduser_retries is zero.) Also, it was sleeping after the final
6002 failure, which is pointless.
6004 24. Implemented tls_on_connect_ports.
6006 25. Implemented acl_smtp_predata.
6008 26. If the domain in control=submission is set empty, Exim assumes that the
6009 authenticated id is a complete email address when it generates From: or
6010 Sender: header lines.
6012 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6013 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6014 chown and chgrp in /bin and hostname in /usr/bin.
6016 28. Exim was keeping the "process log" file open after each use, just as it
6017 does for the main log. This opens the possibility of it remaining open for
6018 long periods when the USR1 signal hits a daemon. Occasional processlog
6019 errors were reported, that could have been caused by this. Anyway, it seems
6020 much more sensible not to leave this file open at all, so that is what now
6023 29. The long-running daemon process does not normally write to the log once it
6024 has entered its main loop, and it closes the log before doing so. This is
6025 so that log files can straightforwardly be renamed and moved. However,
6026 there are a couple of unusual error situations where the daemon does write
6027 log entries, and I had neglected to close the log afterwards.
6029 30. The text of an SMTP error response that was received during a remote
6030 delivery was being truncated at 512 bytes. This is too short for some of
6031 the long messages that one sometimes sees. I've increased the limit to
6034 31. It is now possible to make retry rules that apply only when a message has a
6035 specific sender, in particular, an empty sender.
6037 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6038 it possible to be selective about when SMTP synchronization is enforced.
6040 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6042 32. Implemented hosts_connection_nolog.
6044 33. Added an ACL for QUIT.
6046 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6049 35. Added mailbox_size and mailbox_filecount to appendfile.
6051 36. Added control = no_multiline_responses to ACLs.
6053 37. There was a bug in the logic of the code that waits for the clock to tick
6054 in the case where the clock went backwards by a substantial amount such
6055 that the microsecond fraction of "now" was more than the microsecond
6056 fraction of "then" (but the whole seconds number was less).
6058 38. Added support for the libradius Radius client library this is found on
6059 FreeBSD (previously only the radiusclient library was supported).
6065 1. When certain lookups returned multiple values in the form name=value, the
6066 quoting of the values was not always being done properly. Specifically:
6067 (a) If the value started with a double quote, but contained no whitespace,
6069 (b) If the value contained whitespace other than a space character (i.e.
6070 tabs or newlines or carriage returns) it was not quoted.
6071 This fix has been applied to the mysql and pgsql lookups by writing a
6072 separate quoting function and calling it from the lookup code. The fix
6073 should probably also be applied to nisplus, ibase and oracle lookups, but
6074 since I cannot test any of those, I have not disturbed their existing code.
6076 2. A hit in the callout cache for a specific address caused a log line with no
6077 reason for rejecting RCPT. Now it says "Previous (cached) callout
6078 verification failure".
6080 3. There was an off-by-one bug in the queryprogram router. An over-long
6081 return line was truncated at 256 instead of 255 characters, thereby
6082 overflowing its buffer with the terminating zero. As well as fixing this, I
6083 have increased the buffer size to 1024 (and made a note to document this).
6085 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6086 when Exim is waiting for an SMTP response from a remote server, Exim
6087 restarts its select() call on the socket, thereby resetting its timeout.
6088 This is not a problem when such interrupts are rare. Somebody set up a cron
6089 job to run exiwhat every 2 minutes, which is less than the normal select()
6090 timeout (5 or 10 minutes). This meant that the select() timeout never
6091 kicked in because it was always reset. I have fixed this by comparing the
6092 time when an interrupt arrives with the time at the start of the first call
6093 to select(). If more time than the timeout has elapsed, the interrupt is
6094 treated as a timeout.
6096 5. Some internal re-factoring in preparation for the addition of Sieve
6097 extensions (by MH). In particular, the "personal" test is moved to a
6098 separate function, and given an option for scanning Cc: and Bcc: (which is
6099 not set for Exim filters).
6101 6. When Exim created an email address using the login of the caller as the
6102 local part (e.g. when creating a From: or Sender: header line), it was not
6103 quoting the local part when it contained special characters such as @.
6105 7. Installed new OpenBSD configuration files.
6107 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6108 try to make them clearer.
6110 9. Callout options, other than the timeout value, were being ignored when
6111 verifying sender addresses in header lines. For example, when using
6113 verify = header_sender/callout=no_cache
6115 the cache was (incorrectly) being used.
6117 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6118 only the Cygwin environment.
6120 11. When return_path_on_delivery was set as a log selector, if different remote
6121 addresses in the same message used different return paths and parallel
6122 remote delivery occurred, the wrong values would sometimes be logged.
6123 (Whenever a remote delivery process finished, the return path value from
6124 the most recently started remote delivery process was logged.)
6126 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6127 lines when AUTH and/or TLS are in use. This is the "received protocol"
6128 field. Exim used to use "asmtp" for authenticated SMTP, without any
6129 indication (in the protocol name) for TLS use. Now it follows the RFC and
6130 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6131 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6132 names appear in log lines as well as in Received: header lines.
6134 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6135 extensions, and comparison tests, and to fix some bugs.
6137 14. Changes to the "personal" filter test:
6139 (1) The test was buggy in that it was just doing the equivalent of
6140 "contains" tests on header lines. For example, if a user's address was
6141 anne@some.where, the "personal" test would incorrectly be true for
6143 To: susanne@some.where
6145 This test is now done by extracting each address from the header in turn,
6146 and checking the entire address. Other tests that are part of "personal"
6147 are now done using regular expressions (for example, to check local parts
6148 of addresses in From: header lines).
6150 (2) The list of non-personal local parts in From: addresses has been
6151 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6152 taken from the Sieve specification recommendations.
6154 (3) If the message contains any header line starting with "List-" it is
6155 treated as non-personal.
6157 (4) The test for "circular" in the Subject: header line has been removed
6158 because it now seems ill-conceived.
6160 15. Minor typos in src/EDITME comments corrected.
6162 16. Installed latest exipick from John Jetmore.
6164 17. If headers_add on a router specified a text string that was too long for
6165 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6166 of string_sprintf() is now avoided.
6168 18. $message_body_size was not set (it was always zero) when running the DATA
6169 ACL and the local_scan() function.
6171 19. For the "mail" command in an Exim filter, no default was being set for
6172 the once_repeat time, causing a random time value to be used if "once" was
6173 specified. (If the value happened to be <= 0, no repeat happened.) The
6174 default is now 0s, meaning "never repeat". The "vacation" command was OK
6175 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6176 (I found it when inspecting the code).
6178 20. There is now an overall timeout for performing a callout verification. It
6179 defaults to 4 times the callout timeout, which applies to individual SMTP
6180 commands during the callout. The overall timeout applies when there is more
6181 than one host that can be tried. The timeout is checked before trying the
6182 next host. This prevents very long delays if there are a large number of
6183 hosts and all are timing out (e.g. when the network connections are timing
6184 out). The value of the overall timeout can be changed by specifying an
6185 additional sub-option for "callout", called "maxwait". For example:
6187 verify = sender/callout=5s,maxwait=20s
6189 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6190 to the end before writing, but this should make it even safer).
6192 22. Exim was forgetting that it had advertised PIPELINING for the second and
6193 subsequent messages on an SMTP connection. It was also not resetting its
6194 memory on STARTTLS and an internal HELO.
6196 23. When Exim logs an SMTP synchronization error within a session, it now
6197 records whether PIPELINING has been advertised or not.
6199 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6200 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6201 rather than long int.
6203 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6205 26. Added the never_mail option to autoreply.
6211 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6212 crash if the getsockname() call failed; this can happen if a connection is
6213 closed very soon after it is established. The problem was simply in the
6214 order in which certain operations were done, causing Exim to try to write
6215 to the SMTP stream before it had set up the file descriptor. The bug has
6216 been fixed by making things happen in the correct order.
6222 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6223 before the connection was closed, thus losing the rejection response.
6225 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6226 some early Solaris releases, but causes trouble in current releases where
6227 socklen_t is defined.
6229 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6232 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6235 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6236 that releasing the top of it at the end releases what was used for sub-
6237 expansions (unless the block got too big). However, discard this block if
6238 the first thing is a variable or header, so that we can use its block when
6239 it is dynamic (useful for very large $message_headers, for example).
6241 6. Lookups now cache *every* query, not just the most recent. A new, separate
6242 store pool is used for this. It can be recovered when all lookup caches are
6243 flushed. Lookups now release memory at the end of their result strings.
6244 This has involved some general refactoring of the lookup sources.
6246 7. Some code has been added to the store_xxx() functions to reduce the amount
6247 of flapping under certain conditions.
6249 8. log_incoming_interface used to affect only the <= reception log lines. Now
6250 it causes the local interface and port to be added to several more SMTP log
6251 lines, for example "SMTP connection from", and rejection lines.
6253 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6255 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6257 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6258 could be overwritten at the end of the current message (or the start of a
6259 new message if it was set in a HELO ACL). The value is now preserved for
6260 the duration of the SMTP connection.
6262 12. If a transport had a headers_rewrite setting, and a matching header line
6263 contained an unqualified address, that address was qualified, even if it
6264 did not match any rewriting rules. The underlying bug was that the values
6265 of the flags that permit the existence of unqualified sender and recipient
6266 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6267 non-local messages, and by -bnq for local messages) were not being
6268 preserved with the message after it was received.
6270 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6271 "next input=" as part of the text comprising the host identity instead of
6272 the correct text. The code was using the same buffer for two different
6273 strings. However, depending on which order the printing function evaluated
6274 its arguments, the bug did not always show up. Under Linux, for example, my
6275 test suite worked just fine.
6277 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6278 doesn't work with some older versions of Perl. It has been changed to "my",
6279 which in any case is probably the better facility to use.
6281 15. A really picky compiler found some instances of statements for creating
6282 error messages that either had too many or two few arguments for the format
6285 16. The size of the buffer for calls to the DNS resolver has been increased
6286 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6287 for addresses that have a lot of PTR records. This alleviates a problem; it
6288 does not fully solve it.
6290 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6291 buffer now truncates the list and logs the incident, which is the same
6292 action as happens when Exim is looking up a host name and its aliases.
6293 Previously in this situation something unpredictable would happen;
6294 sometimes it was "internal error: store_reset failed".
6296 18. If a server dropped the connection unexpectedly when an Exim client was
6297 using GnuTLS and trying to read a response, the client delivery process
6298 crashed while trying to generate an error log message.
6300 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6301 string, for example:
6303 warn message = H1: something\nH2: something
6305 the text was added as a single header line from Exim's point of view
6306 though it ended up OK in the delivered message. However, searching for the
6307 second and subsequent header lines using $h_h2: did not work. This has been
6308 fixed. Similarly, if a system filter added multiple headers in this way,
6309 the routers could not see them.
6311 20. Expanded the error message when iplsearch is called with an invalid key to
6312 suggest using net-iplsearch in a host list.
6314 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6315 ACLs are no longer actually imposed (and a message to that effect is
6318 22. If a "gecos" field in a passwd entry contained escaped characters, in
6319 particular, if it contained a \" sequence, Exim got it wrong when building
6320 a From: or a Sender: header from that name. A second bug also caused
6321 incorrect handling when an unquoted " was present following a character
6322 that needed quoting.
6324 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6325 was not being matched caselessly.
6327 24. Arranged for all hyphens in the exim.8 source to be escaped with
6330 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6331 properly. Recipient callout cache records were still being keyed to include
6332 the sender, even when use_sender was set false. This led to far more
6333 callouts that were necessary. The sender is no longer included in the key
6334 when use_sender is false.
6336 26. Added "control = submission" modifier to ACLs.
6338 27. Added the ${base62d: operator to decode base 62 numbers.
6340 28. dnsdb lookups can now access SRV records.
6342 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6343 the configuration file.
6345 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6346 (-v) mode. This makes the output for a verbose queue run more intelligible.
6348 31. Added a use_postmaster feature to recipient callouts.
6350 32. Added the $body_zerocount variable, containing the number of binary zero
6351 bytes in the message body.
6353 33. The time of last modification of the "new" subdirectory is now used as the
6354 "mailbox time last read" when there is a quota error for a maildir
6357 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6359 35. Added +ignore_unknown as a special item in host lists.
6361 36. Code for decoding IPv6 addresses in host lists is now included, even if
6362 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6363 address was recognized as an IP address, but was then not correctly decoded
6364 into binary, causing unexpected and incorrect effects when compared with
6371 1. Very minor rewording of debugging text in manualroute to say "list of
6372 hosts" instead of "hostlist".
6374 2. If verify=header_syntax was set, and a header line with an unqualified
6375 address (no domain) and a large number of spaces between the end of the
6376 name and the colon was received, the reception process suffered a buffer
6377 overflow, and (when I tested it) crashed. This was caused by some obsolete
6378 code that should have been removed. The fix is to remove it!
6380 3. When running in the test harness, delay a bit after writing a bounce
6381 message to get a bit more predictability in the log output.
6383 4. Added a call to search_tidyup() just before forking a reception process. In
6384 theory, someone could use a lookup in the expansion of smtp_accept_max_
6385 per_host which, without the tidyup, could leave open a database connection.
6387 5. Added the variables $recipient_data and $sender_data which get set from a
6388 lookup success in an ACL "recipients" or "senders" condition, or a router
6389 "senders" option, similar to $domain_data and $local_part_data.
6391 6. Moved the writing of debug_print from before to after the "senders" test
6394 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6395 problems for message scanning, either using a data ACL, or using
6396 local_scan() because the Received: header was not generated till after they
6397 were called (in order to set the time as the time of reception completion).
6398 I have revised the way this works. The header is now generated after the
6399 body is received, but before the ACL or local_scan() are called. After they
6400 are run, the timestamp in the header is updated.
6406 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6407 before starting a queue runner without re-exec. This happened only when
6408 deliver_drop_privilege was set or when the Exim user was set to root. The
6409 effect of the bug was that timeouts during subsequent deliveries caused
6410 crashes instead of being properly handled. The handler is now left at its
6411 default (and expected) setting.
6413 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6414 message, again when deliver_drop_privilege is set or Exim is run as root.
6415 The bug described in (1) was not present in this case, but the tidying up
6416 of the other signals was missing. I have made the two cases consistent.
6418 3. The ignore_target_hosts setting on a manualroute router was being ignored
6419 for hosts that were looked up using the /MX notation.
6421 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6424 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6425 operated on the sender address. After changing the $sender_address to <>
6426 for the sender address verify, Exim was re-instated it as the original
6427 (before rewriting) address, but remembering that it had rewritten it, so it
6428 wasn't rewriting it again. This bug also had the effect of breaking the
6429 sender address verification caching when the sender address was rewritten.
6431 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6432 This has been changed so that if the ip literal address matches
6433 ignore_target_hosts, the router declines.
6435 7. Added expansion conditions match_domain, match_address, and match_local_
6436 part (NOT match_host).
6438 8. The placeholder for the Received: header didn't have a length field set.
6440 9. Added code to Exim itself and to exim_lock to test for a specific race
6441 condition that could lead to file corruption when using MBX delivery. The
6442 issue is with the lockfile that is created in /tmp. If this file is removed
6443 after a process has opened it but before that process has acquired a lock,
6444 there is the potential for a second process to recreate the file and also
6445 acquire a lock. This could lead to two Exim processes writing to the file
6446 at the same time. The added code performs the same test as UW imapd; it
6447 checks after acquiring the lock that its file descriptor still refers to
6448 the same named file.
6450 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6451 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6454 11. Added the smtp_active_hostname option. If used, this will typically be made
6455 to depend on the incoming interface address. Because $interface_address is
6456 not set up until the daemon has forked a reception process, error responses
6457 that can happen earlier (such as "too many connections") no longer contain
6460 12. If an expansion in a condition on a "warn" statement fails because a lookup
6461 defers, the "warn" statement is abandoned, and the next ACL statement is
6462 processed. Previously this caused the whole ACL to be aborted.
6464 13. Added the iplsearch lookup type.
6466 14. Added ident_timeout as a log selector.
6468 15. Added tls_certificate_verified as a log selector.
6470 16. Added a global option tls_require_ciphers (compare the smtp transport
6471 option of the same name). This controls incoming TLS connections.
6473 17. I finally figured out how to make tls_require_ciphers do a similar thing
6474 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6475 before starting the TLS session.
6477 18. Tabs are now shown as \t in -bP output.
6479 19. If the log selector return_path_on_delivery was set, Exim crashed when
6480 bouncing a message because it had too many Received: header lines.
6482 20. If two routers both had headers_remove settings, and the first one included
6483 a superfluous trailing colon, the final name in the first list and the
6484 first name in the second list were incorrectly joined into one item (with a
6485 colon in the middle).
6491 1. Added -C and -D options to the exinext utility, mainly to make it easier
6492 to include in the automated testing, but these could be helpful when
6493 multiple configurations are in use.
6495 2. The exinext utility was not formatting the output nicely when there was
6496 an alternate port involved in the retry record key, nor when there was a
6497 message id as well (for retries that were specific to a specific message
6498 and a specific host). It was also confused by IPv6 addresses, because of
6499 the additional colons they contain. I have fixed the IPv4 problem, and
6500 patched it up to do a reasonable job for IPv6.
6502 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6503 delivery, the log line now contains "pipelined" if PIPELINING was used.
6505 4. An SMTP transport process used to panic and die if the bind() call to set
6506 an explicit outgoing interface failed. This has been changed; it is now
6507 treated in the same way as a connect() failure.
6509 5. A reference to $sender_host_name in the part of a conditional expansion
6510 that was being skipped was still causing a DNS lookup. This no longer
6513 6. The def: expansion condition was not recognizing references to header lines
6514 that used bh_ and bheader_.
6516 7. Added the _cache feature to named lists.
6518 8. The code for checking quota_filecount in the appendfile transport was
6519 allowing one more file than it should have been.
6521 9. For compatibility with Sendmail, the command line option
6529 and sets the incoming protocol and host name (for trusted callers). The
6530 host name and its colon can be omitted when only the protocol is to be set.
6531 Note the Exim already has two private options, -pd and -ps, that refer to
6532 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6533 "s", but I don't think that's a major issue.
6535 10. A number of refactoring changes to the code, none of which should affect
6538 (a) The number of logging options was getting close to filling up the
6539 32-bit word that was used as a bit map. I have split them into two classes:
6540 those that are passed in the argument to log_write(), and those that are
6541 only ever tested independently outside of that function. These are now in
6542 separate 32-bit words, so there is plenty of room for expansion again.
6543 There is no change in the user interface or the logging behaviour.
6545 (b) When building, for example, log lines, the code previously used a
6546 macro that called string_cat() twice, in order to add two strings. This is
6547 not really sufficiently general. Furthermore, there was one instance where
6548 it was actually wrong because one of the argument was used twice, and in
6549 one call a function was used. (As it happened, calling the function twice
6550 did not affect the overall behaviour.) The macro has been replaced by a
6551 function that can join an arbitrary number of extra strings onto a growing
6554 (c) The code for expansion conditions now uses a table and a binary chop
6555 instead of a serial search (which was left over from when there were very
6556 few conditions). Also, it now recognizes conditions like "pam" even when
6557 the relevant support is not compiled in: a suitably worded error message is
6558 given if an attempt is made to use such a condition.
6560 11. Added ${time_interval:xxxxx}.
6562 12. A bug was causing one of the ddress fields not to be passed back correctly
6563 from remote delivery subprocesses. The field in question was not being
6564 subsequently used, so this caused to problems in practice.
6566 13. Added new log selectors queue_time and deliver_time.
6568 14. Might have fixed a bug in maildirsizefile handling that threw up
6569 "unexpected character" debug warnings, and recalculated the data
6570 unnecessarily. In any case, I expanded the warning message to give more
6573 15. Added the message "Restricted characters in address" to the statements in
6574 the default ACL that block characters like @ and % in local parts.
6576 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6577 Three changes have been made:
6579 (a) There was a serious bug; a negative response to MAIL caused the whole
6580 recipient domain to be cached as invalid, thereby blocking all messages
6581 to all local parts at the same domain, from all senders. This bug has
6582 been fixed. The domain is no longer cached after a negative response to
6583 MAIL if the sender used is not empty.
6585 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6588 (c) A new callout option, "use_sender" has been added for people who want
6589 the modified behaviour.
6595 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6598 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6599 indeed breaks things for older releases.
6601 3. Added additional logging to the case where there is a problem reading data
6602 from a filter that is running in a subprocess using a pipe, in order to
6603 try to track down a specific problem.
6605 4. Testing facility fudge: when running in the test harness and attempting
6606 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6607 getting "No route to host". Convert this to a timeout.
6609 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6612 6. Some OS don't have socklen_t but use size_t instead. This affects the
6613 fifth argument of getsockopt() amongst other things. This is now
6614 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6615 can be set for individual OS. I have set it for SunOS5, OSF1, and
6616 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6617 some earlier ones do not.
6619 7. Change 4.30/15 was not doing the test caselessly.
6621 8. The standard form for an IPv6 address literal was being rejected by address
6622 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6623 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6624 this, as well as the form without the "IPv6" on the front (but only when
6625 address literals are enabled, of course).
6627 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6629 10. Exim crashed if a message with an empty sender address specified by -f
6630 encountered a router with an errors_to setting. This could be provoked only
6631 by a command such as
6635 where an empty string was supplied; "<>" did not hit this bug.
6637 11. Installed PCRE release 4.5.
6639 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6640 remained set. It is now erased.
6642 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6643 times from message ids (which are base 36 rather than the normal 62).
6645 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6646 were being counted as actual protocol errors, and logged if the log
6647 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6648 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6649 and DATA following a set of rejected RCPTs do not count as protocol errors.
6650 In other words, Exim assumes they were pipelined, though this may not
6651 actually be the case. Of course, in all cases the client gets an
6652 appropriate error code.
6654 15. If a lookup fails in an ACL condition, a message about the failure may
6655 be available; it is used if testing the ACL cannot continue, because most
6656 such messages specify what the cause of the deferral is. However, some
6657 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6658 that caused an old message to be retained and used if a later statement
6659 caused a defer, replacing the real cause of the deferral.
6661 16. If an IP address had so many PTR records that the DNS lookup buffer
6662 was not large enough to hold them, Exim could crash while trying to process
6663 the truncated data. It now detects and logs this case.
6665 17. Further to 4.21/58, another change has been made: if (and only if) the
6666 first line of a message (the first header line) ends with CRLF, a bare LF
6667 in a subsequent header line has a space inserted after it, so as not to
6668 terminate the header.
6670 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6671 unnecessarily, used atoi() instead of strtol(), and didn't check the
6672 termination when getting file sizes from file names by regex.
6674 19. Completely re-implemented the support for maildirsize files, in the light
6675 of a number of problems with the previous contributed implementation
6676 (4.30/29). In particular:
6678 . If the quota is zero, the maildirsize file is maintained, but no quota is
6681 . If the maildir directory does not exist, it is created before any attempt
6682 to write a maildirsize file.
6684 . The quota value in the file is just a cache; if the quota is changed in
6685 the transport, the new value overrides.
6687 . A regular expression is available for excluding directories from the
6690 20. The autoreply transport checks the characters in options that define the
6691 message's headers; it allows continued headers, but it was checking with
6692 isspace() after an embedded newline instead of explicitly looking for a
6695 21. If all the "regular" hosts to which an address was routed had passed their
6696 expiry times, and had not reached their retry times, the address was
6697 bounced, even if fallback hosts were defined. Now Exim should go on to try
6700 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6701 equivalent code in the SMTP transport. Some hosts send humungous responses
6702 to HELO/EHLO, more than 1024 it seems.
6704 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6705 gives compiler warnings in some environments. I've now done it "properly",
6708 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6709 (because of problems with the built-in one) was declared to return uschar *
6710 instead of char *, causing compiler failure.
6712 25. Fixed a file descriptor leak when processing alias/forward files.
6714 26. Fixed a minor format string issue in dbfn.c.
6716 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6718 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6719 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6722 29. When a sender address is verified, it is cached, to save repeating the test
6723 when there is more than one recipient in a message. However, when the
6724 verification involves a callout, it is possible for different callout
6725 options to be set for different recipients. It is too complicated to keep
6726 track of this in the cache, so now Exim always runs a verification when a
6727 callout is required, relying on the callout cache for the optimization.
6728 The overhead is duplication of the address routing, but this should not be
6731 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6732 to be verified with callout=postmaster, and the main callout worked but the
6733 postmaster check failed, the verification correctly failed. However, if a
6734 subsequent RCPT command asked for sender verification *without* the
6735 postmaster check, incorrect caching caused this verification also to fail,
6738 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6739 it was not caching the DNS options (qualify_single, search_parents) that
6740 were used when the lookup failed. A subsequent lookup with different
6741 options therefore always gave the same answer, though there were cases
6742 where it should not have. (Example: a "domains = !$mx_any" option on a
6743 dnslookup router: the "domains" option is always processed without any
6744 widening, but the router might have qualify_single set.) Now Exim uses the
6745 cached value only when the same options are set.
6747 32. Added John Jetmore's "exipick" utility to the distribution.
6749 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6750 than a timeout (e.g. a certificate is required, and is not provided), an
6751 Exim server now closes the connection immediately. Previously it waited for
6752 the client to close - but if the client is SSL, it seems that they each
6753 wait for each other, leading to a delay before one of them times out.
6755 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6756 maintained 0.8.x compatibility because I don't think many are using it, and
6757 it is clearly obsolete.
6759 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6762 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6763 client certificate was expired. A simple patch fixes this, though I don't
6764 understand the full logic of why the verify callback is called multiple
6767 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6768 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6769 which causes problems with some clients (such as the Certicom SSL Plus
6770 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6771 disables the coutermeasure allowing Eudora to connect."
6773 38. Exim was not checking that a write() to a log file succeeded. This could
6774 lead to Bad Things if a log got too big, in particular if it hit a file
6775 size limit. Exim now panics and dies if it cannot write to a log file, just
6776 as it does if it cannot open a log file.
6778 39. Modified OS/Makefile-Linux so that it now contains
6780 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6782 The two -D definitions ensure that Exim is compiled with large file
6783 support, which makes it possible to handle log files that are bigger than
6786 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6787 instance) a domain was checked against a named list that involved a lookup,
6788 causing $domain_data to be set, then another domain was checked against the
6789 same list, then the first domain was re-checked, the value of $domain_data
6790 after the final check could be wrong. In particular, if the second check
6791 failed, it could be set empty. This bug probably also applied to
6794 41. The strip_trailing_dot option was not being applied to the address given
6795 with the -f command-line option.
6797 42. The code for reading a message's header from the spool was incrementing
6798 $received_count, but never initializing it. This meant that the value was
6799 incorrect (doubled) while delivering a message in the same process in which
6800 it was received. In the most common configuration of Exim, this never
6801 happens - a fresh exec is done - but it can happen when
6802 deliver_drop_privilege is set.
6804 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6805 it now includes up to 150 characters of the unexpected data in the log
6808 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6809 and building data strings. The size of both of these buffers was 10 000
6810 bytes - far larger than anybody would *ever* want, thought I. Needless to
6811 say, somebody hit the limit. I have increased the maximum line length to
6812 20 000 and the maximum data length of concatenated lines to 100 000. I have
6813 also fixed two bugs, because there was no checking on these buffers. Tsk,
6814 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6815 buffer is too small.
6817 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6818 lsearch lookups. Now it does.
6820 46. When parsing a route_list item in a manualroute router, a fixed-length
6821 buffer was used for the list of hosts. I made this 1024 bytes long,
6822 thinking that nobody would ever have a list of hosts that long. Wrong.
6823 Somebody had a whole pile of complicated expansion conditions, and the
6824 string was silently truncated, leading to an expansion error. It turns out
6825 that it is easier to change to an unlimited length (owing to other changes
6826 that have happened since this code was originally written) than to build
6827 structure for giving a limitation error. The length of the item that
6828 expands into the list of hosts is now unlimited.
6830 47. The lsearch lookup could not handle data where the length of text line was
6831 more than 4095 characters. Such lines were truncated, leading to shortened
6832 data being returned. It should now handle lines of any length.
6834 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6835 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6838 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6839 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6840 no longer happens in comment lines. A list of replacements is now placed
6841 at the head of all of the source files, except those whose only change is
6842 to replace PERL_COMMAND in the very first #! line.
6844 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6845 messages on the queue, with a bottom-up merge sort, using code contributed
6846 by Michael Haardt. This should make operations like -bp somewhat faster on
6847 large queues. It won't affect queue runners, except when queue_run_in_order
6850 51. Installed eximstats 1.31 in the distribution.
6852 52. Added support for SRV lookups to the dnslookup router.
6854 53. If an ACL referred to $message_body or $message_body_end, the value was not
6855 reset for any messages that followed in the same SMTP session.
6857 54. The store-handling optimization for building very long strings was not
6858 differentiating between the different store pools. I don't think this
6859 actually made any difference in practice, but I've tidied it.
6861 55. While running the routers to verify a sender address, $sender_address
6862 was still set to the sender address. This is wrong, because when routing to
6863 send a bounce to the sender, it would be empty. Therefore, I have changed
6864 it so that, while verifying a sender address, $sender_address is set to <>.
6865 (There is no change to what happens when verifying a recipient address.)
6867 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6868 target A or AAAA records (if not already returned) without resetting the
6869 qualify_single or search_parents options of the DNS resolver. These are
6870 inappropriate in this case because the targets of MX and SRV records must
6871 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6872 target that, when qualified, matched something in the local domain. These
6873 two options are now turned off when doing these lookups.
6875 57. It seems that at least some releases of Reiserfs (which does not have the
6876 concept of a fixed number of inodes) returns zero and not -1 for the
6877 number of available inodes. This interacted badly with check_spool_inodes,
6878 which assumed that -1 was the "no such thing" setting. What I have done is
6879 to check that the total number of inodes is greater than zero before doing
6880 the test of how many are available.
6882 58. When a "warn" ACL statement has a log_message modifier, the message is
6883 remembered, and not repeated. This is to avoid a lot of repetition when a
6884 message has many recipients that cause the same warning to be written.
6885 However, Exim was preserving the list of already written lines for an
6886 entire SMTP session, which doesn't seem right. The memory is now reset if a
6887 new message is started.
6889 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6890 debugging output unless log_rewrite was also set.
6892 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6893 of (int)(handle) when we know that handle contains (void *)(-1).
6895 61. The Exim daemon panic-logs an error return when it closes the incoming
6896 connection. However "connection reset by peer" seems to be common, and
6897 isn't really an error worthy of noting specially, so that particular error
6900 62. When Exim is trying to find all the local interfaces, it used to panic and
6901 die if the ioctl to get the interface flags failed. However, it seems that
6902 on at least one OS (Solaris 9) it is possible to have an interface that is
6903 included in the list of interfaces, but for which you get a failure error
6904 for this call. This happens when the interface is not "plumbed" into a
6905 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6906 failure of the "get flags" call assumes that the interface is down.
6908 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6909 makes life easier for people who are doing arithmetic on fields extracted
6910 from dates, where you often get leading zeros that should not be
6911 interpreted as octal.
6913 64. Added qualify_domain to the redirect router, to override the global
6916 65. If a pathologically long header line contained very many addresses (the
6917 report of this problem mentioned 10 000) and each of them was rewritten,
6918 Exim could use up a very large amount of memory. (It kept on making new
6919 copies of the header line as it rewrote, and never released the old ones.)
6920 At the expense of a bit more processing, the header rewriting function has
6921 been changed so that it no longer eats memory in this way.
6923 66. The generation of the Received: header has been moved from the time that a
6924 message starts to be received, to the time that it finishes. The timestamp
6925 in the Received: header should now be very close to that of the <= log
6926 line. There are two side-effects of this change:
6928 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6929 logged header lines no longer include the local Received: line, because
6930 it has not yet been created. The same applies to a copy of the message
6931 that is returned to a non-SMTP sender when a message is rejected.
6933 (b) When a filter file is tested using -bf, no additional Received: header
6934 is added to the test message. After some thought, I decided that this
6937 This change does not affect the value of $received_for. It is still set
6938 after address rewriting, but before local_scan() is called.
6940 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6942 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6943 gave an unhelpful panic error message, and a defer error. I have managed to
6944 change this behaviour so that it now rejects any supplied certificate,
6945 which seems right, as the list of acceptable certificates is empty.
6947 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6948 gave an unhelpful defer error. I have not managed to make this reject any
6949 supplied certificates, but the error message it gives is "no certificate
6950 supplied", which is not helpful.
6952 70. exigrep's output now also includes lines that are not associated with any
6953 message, but which match the given pattern. Implemented by a patch from
6954 Martin Sluka, which also tidied up the Perl a bit.
6956 71. Recipient callout verification, like sender verification, was using <> in
6957 the MAIL FROM command. This isn't really the right thing, since the actual
6958 sender may affect whether the remote host accepts the recipient or not. I
6959 have changed it to use the actual sender in the callout; this means that
6960 the cache record is now keyed on a recipient/sender pair, not just the
6961 recipient address. There doesn't seem to be a real danger of callout loops,
6962 since a callout by the remote host to check the sender would use <>.
6963 [SEE ABOVE: changed after hitting problems.]
6965 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6966 temporary errors. However, in the case of such a code being given after
6967 the end of a data transmission (i.e. after ".") Exim was failing to write
6968 a retry record for the message. (Yes, there was some broken host that was
6969 actually sending 8xx at this point.)
6971 73. An unknown lookup type in a host list could cause Exim to panic-die when
6972 the list was checked. (An example that provoked this was putting <; in the
6973 middle of a list instead of at the start.) If this happened during a DATA
6974 ACL check, a -D file could be left lying around. This kind of configuration
6975 error no longer causes Exim to die; instead it causes a defer error. The
6976 incident is still logged to the main and panic logs.
6978 74. Buglet left over from Exim 3 conversion. The message "too many messages
6979 in one connection" was written to the rejectlog but not the mainlog, except
6980 when address rewriting (yes!) was being logged.
6982 75. Added write_rejectlog option.
6984 76. When a system filter was run not as root (that is, when system_filter_user
6985 was set), the values of the $n variables were not being returned to the
6986 main process; thus, they were not subsequently available in the $sn
6989 77. Added +return_path_on_delivery log selector.
6991 78. A connection timeout was being treated differently from recipients deferred
6992 when testing hosts_max_try with a message that was older than the host's
6993 retry timeout. (The host should not be counted, thus allowing all hosts to
6994 be tried at least once before bouncing.) This may have been the cause of an
6995 occasionally reported bug whereby a message would remain on the queue
6996 longer than the retry timeout, but would be bounced if a delivery was
6997 forced. I say "may" because I never totally pinned down the problem;
6998 setting up timeout/retry tests is difficult. See also the next item.
7000 79. The ultimate address timeout was not being applied to errors that involved
7001 a combination of host plus message (for example, a timeout on a MAIL
7002 command). When an address resolved to a number of possible hosts, and they
7003 were not all tried for each delivery (e.g. because of hosts_max_try), a
7004 message could remain on the queue longer than the retry timeout.
7006 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7009 81. Fixed an obscure SMTP outgoing bug which required at least the following
7010 conditions: (a) there was another message waiting for the same server;
7011 (b) the server returned 5xx to all RCPT commands in the first message so
7012 that the message was not completed; (c) the server dropped the connection
7013 or gave a negative response to the RSET that Exim sends to abort the
7014 transaction. The observed case was a dropped connection after DATA that had
7015 been sent in pipelining mode. That is, the server had advertised PIPELINING
7016 but was not implementing it correctly. The effect of the bug was incorrect
7017 behaviour, such as trying another host, and this could lead to a crash.
7023 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7024 and daemon.c were passed as pointers to ints; they should have been
7025 pointers to socklen_t variables (which are typically unsigned ints).
7027 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7030 3. Fixed a really odd bug that affected only the testing scheme; patching a
7031 certain fixed string in the binary changed the value of another string that
7032 happened to be identical to the end of the original first string.
7034 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7035 name", it returns that address as the IP address. On some operating
7036 systems (e.g. Solaris), it also passes back the IP address string as the
7037 "host name". However, on others (e.g. Linux), it passes back an empty
7038 string. Exim wasn't checking for this, and was changing the host name to an
7039 empty string, assuming it had been canonicalized.
7041 5. Although rare, it is permitted to have more than one PTR record for a given
7042 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7043 all the names associated with an address, because they do in Solaris.
7044 However, it seems that they do not in Linux for data that comes from the
7045 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7046 I found this out when I moved to a new Linux workstation and tried to run
7047 the Exim test suite.
7049 To get round this problem I have changed the code so that it now does its
7050 own call to the DNS to look up PTR records when searching for a host name.
7051 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7052 addresses that are only in /etc/hosts are still found.
7054 This behaviour is, however, controlled by an option called host_lookup_
7055 order, which defaults to "bydns:byaddr". If people want to use the other
7056 order, or indeed, just use one or the other means of lookup, they can
7057 specify it in this variable.
7059 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7060 some operating systems, this comes back from gethostbyaddr() as an empty
7061 string, and this is what Exim used to test for. However, it seems that in
7062 other systems, "." is yielded. Exim now tests for this case too.
7064 7. The values of check_spool_space and check_log_space are now held internally
7065 as a number of kilobytes instead of an absolute number of bytes. If a
7066 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7067 kilobyte. This means that much larger values can be stored.
7069 8. Exim monitor: an attempt to get the action menu when not actually pointing
7070 at a message produces an empty menu entitled "No message selected". This
7071 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7072 no entries in it ("Shell widget menu has zero width and/or height"). So I
7073 have added a single, blank menu entry in this case.
7075 9. Added ${quote_local_part.
7077 10. MIME decoding is now applied to the contents of Subject: header lines when
7080 11. Now that a reference to $sender_host_address automatically causes a reverse
7081 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7082 host lookup before query-style lookups in lists that might use this
7083 variable. This has therefore been abolished, and the "net-" prefix is no
7084 longer necessary for query-style lookups.
7086 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7087 have been a typo for LFLAGS, so it has been changed.
7089 13. The install script calls Exim with "-C /dev/null" in order to find the
7090 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7091 to be output. However, since Exim outputs its version number before the
7092 error, it didn't break the script. It just looked ugly. I fixed this by
7093 always allowing "-C /dev/null" if the caller is root.
7095 14. Ignore overlarge ACL variable number when reading spool file - insurance
7096 against a later release with more variables having written the file.
7098 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7099 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7100 this, as well as the form without the "IPv6" on the front.
7102 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7103 OS/Makefile-Darwin file.
7105 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7106 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7108 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7109 variable. However, some OS use ints for inodes. I've added cast to long int
7110 to get rid of the compiler warning.
7112 19. I had forgotten to lock out "/../" in configuration file names when
7113 ALT_CONFIG_PREFIX was set.
7115 20. Routers used for verification do not need to specify transports. However,
7116 if such a router generated a host list, and callout was configured, Exim
7117 crashed, because it could not find a port number from the (non-existent)
7118 transport. It now assumes port 25 in this circumstance.
7120 21. Added the -t option to exigrep.
7122 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7123 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7124 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7125 them is set, LOOKUP_LSEARCH is forced.
7127 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7129 24. Added sender and host information to the "rejected by local_scan()" log
7130 line; previously there was no indication of these.
7132 25. Added .include_if_exists.
7134 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7135 Linux. It turns out that not all file systems support this. Apparently some
7136 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7137 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7138 not supported on the file descriptor, is now ignored when Exim is trying to
7139 sync a directory. This applies only to Linux.
7141 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7143 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7144 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7147 29. Added support for maildirsize files from supplied patch (modified a bit).
7149 30. The use of :fail: followed by an empty string could lead Exim to respond to
7150 sender verification failures with (e.g.):
7152 550 Verification failed for <xxx>
7153 550 Sender verify failed
7155 where the first response line was missing the '-' that indicates it is not
7156 the final line of the response.
7158 31. The loop for finding the name of the user that called Exim had a hardwired
7159 limit of 10; it now uses the value of finduser_retries, which is used for
7160 all other user lookups.
7162 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7165 33. Exim was neglecting to zero errno before one call of strtol() when
7166 expanding a string and expecting an integer value. On some systems this
7167 resulted in spurious "integer overflow" errors. Also, it was casting the
7168 result into an int without checking.
7170 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7171 did not work. The code looks as if it has *never* worked, though it appears
7172 to have been documented since at least release 1.62. I have made it work.
7174 35. The "timeout_DNS" error in retry rules, also documented since at least
7175 1.62, also never worked. As it isn't clear exactly what this means, and
7176 clearly it isn't a major issue, I have abolished the feature by treating it
7177 as "timeout", and writing a warning to the main and panic logs.
7179 36. The display of retry rules for -brt wasn't always showing the error code
7182 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7183 timeout_connect_A, timeout_connect_MX.
7185 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7186 to the empty sender.
7188 39. The daemon was not analysing the content of -oX till after it had closed
7189 stderr and disconnected from the controlling terminal. This meant that any
7190 syntax errors were only noted on the panic log, and the return code from
7191 the command was 0. By re-arranging the code a little, I've made the
7192 decoding happen first, so such errors now appear on stderr, and the return
7193 code is 1. However, the actual setting up of the sockets still happens in
7194 the disconnected process, so errors there are still only recorded on the
7197 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7198 connections (as happens on some IP stacks) was logged at start up time as
7199 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7200 it from "IPv6 and IPv4", which means that two separate sockets are being
7203 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7204 says whether AF_INET or AF_INET6 was passed as an argument.
7206 42. Exiwhat output was messed up when time zones were included in log
7209 43. Exiwhat now gives more information about the daemon's listening ports,
7210 and whether -tls-on-connect was used.
7212 44. The "port" option of the smtp transport is now expanded.
7214 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7215 silently ignored. Now an error message is written to the main and panic
7218 46. There's a new ACL modifier called "logwrite" which writes to a log file
7219 as soon as it is encountered.
7221 47. Added $local_user_uid and $local_user_gid at routing time.
7223 48. Exim crashed when trying to verify a sender address that was being
7226 49. Exim was recognizing only a space character after ".include". It now also
7227 recognizes a tab character.
7229 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7230 extracting the relevant information from the specification. The man page no
7231 longer contains scrambled data for the -d option, and I've added a section
7232 at the front about calling Exim under different names.
7234 51. Added "extra_headers" argument to the "mail" command in filter files.
7236 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7239 53. Installed eximstats 1.29.
7241 54. Added transport_filter_timeout as a generic transport option.
7243 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7244 Cc: header lines. This was required by RFC 822, but it not required by RFC
7247 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7248 incoming messages that did not have them. Now it does so only if the
7249 message originates locally, that is, if there is no associated remote host
7250 address. When Resent- header lines are present, this applies to the Resent-
7251 lines rather than the non-Resent- lines.
7253 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7254 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7256 58. Messages for configuration errors now include the name of the main
7257 configuration file - useful now that there may be more than one file in a
7258 list (.included file names were always shown).
7260 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7261 for those rare installations that do not start the daemon as root or run it
7262 setuid root. I've cut out the call to initgroups() if the daemon is not
7265 60. The Exim user and group can now be bound into the binary as text strings
7266 that are looked up at the start of Exim's processing.
7268 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7270 62. Added $mailstore_basename variable.
7272 63. Installed patch to sieve.c from Michael Haardt.
7274 64. When Exim failed to open the panic log after failing to open the main log,
7275 the original message it was trying to log was written to stderr and debug
7276 output, but if they were not available (the usual case in production), it
7277 was lost. Now it is written to syslog before the two lines that record the
7278 failures to open the logs.
7280 65. Users' Exim filters run in subprocesses under the user's uid. It is
7281 possible for a "deliver" command or an alias in a "personal" command to
7282 provoke an address rewrite. If logging of address rewriting is configured,
7283 this fails because the process is not running as root or exim. There may be
7284 a better way of dealing with this, but for the moment (because 4.30 needs
7285 to be released), I have disabled address rewrite logging when running a
7286 filter in a non-root, non-exim process.
7292 1. The buildconfig auxiliary program wasn't quoting the value set for
7293 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7294 not defined. This bug was masked in 4.22 by the effect that was fixed in
7297 2. Some messages that were rejected after a message id was allocated were
7298 shown as "incomplete" by exigrep. It no longer does this for messages that
7299 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7301 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7302 have allow_domain_literals set, the ID did not get logged in the <= line.
7303 Domain literals are now always recognized in Message-ID: header lines.
7305 4. The first argument for a ${extract expansion item is the key name or field
7306 number. Leading and trailing spaces in this item were not being ignored,
7307 causing some misleading effects.
7309 5. When deliver_drop_privilege was set, single queue runner processes started
7310 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7311 same command in the process it spins off) were not dropping privilege.
7313 6. When the daemon running as "exim" started a queue runner, it always
7314 re-executed Exim in the spun-off process. This is a waste of effort when
7315 deliver_drop_privilege is set. The new process now just calls the
7316 queue-runner function directly.
7322 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7325 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7326 ignored. Though the use of -oP was forcing the writing of a pid file, it
7327 was always written to the default place.
7329 3. If the message "no IP address found for host xxxx" is generated during
7330 incoming verification, it is now followed by identification of the incoming
7331 connection (so you can more easily find what provoked it).
7333 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7335 5. Added some features to "harden" Exim a bit more against certain attacks:
7337 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7338 be put in Local/Makefile. This is like the never_users runtime option,
7339 but it cannot be overridden. The default setting is "root".
7341 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7342 prefix string with which any file named in a -C command line option
7345 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7346 is retained for -C and -D only if the caller of Exim is root. Without
7347 it, the exim user may also use -C and -D and retain privilege.
7349 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7350 command line option is disabled.
7352 6. Macro names set by the -D option must start with an upper case letter, just
7353 like macro names defined in the configuration file.
7355 7. Added "dereference=" facility to LDAP.
7357 8. Two instances of the typo "uknown" in the source files are fixed.
7359 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7360 the Configure-Makefile script screwed up while processing it.
7362 10. Incorporated PCRE 4.4.
7364 11. The SMTP synchronization check was not operating right at the start of an
7365 SMTP session. For example, it could not catch a HELO sent before the client
7366 waited for the greeting. There is now a check for outstanding input at the
7367 point when the greeting is written. Because of the duplex, asynchronous
7368 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7369 way, but not yet received, when the check is performed.
7371 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7372 on TCP/IP sockets, because this apparently causes some broken clients to
7375 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7376 unchanged) from the Cygwin maintainer.
7378 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7379 was supported instead of testing for mbx. Effectively a typo.
7381 15. The spa authenticator server code was not checking that the input it
7382 received was valid base64.
7384 16. The debug output line for the "set" modifier in ACLs was not showing the
7385 name of the variable that was being set.
7387 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7389 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7390 lookup on its own. Something else was needed to trigger the lookup. For
7391 example, a match in host_lookup or the need for a host name in a host list.
7392 Now, if $sender_host_name is referenced and the host name has not yet been
7393 looked up, a lookup is performed. If the lookup fails, the variable remains
7394 empty, and $host_lookup_failed is set to "1".
7396 19. Added "eqi" as a case-independent comparison operator.
7398 20. The saslauthd authentication condition could segfault if neither service
7399 nor realm was specified.
7401 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7402 error message that was logged was misleading, and incoming SMTP
7403 connections were dropped. The message is now more accurate, and temporary
7404 errors are given to SMTP connections.
7406 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7407 (except RSET) with a 421 error, until QUIT is received. However, it was
7408 failing to send a response to QUIT.
7410 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7411 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7412 ACL is now run after the helo_try_verify_hosts code.
7414 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7415 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7416 case-independent, so other case variants are also recognized). Apparently
7417 some systems use these upper case variants.
7419 25. If more than two messages were waiting for the same host, and a transport
7420 filter was specified for the transport, Exim sent two messages over the
7421 same TCP/IP connection, and then failed with "socket operation on non-
7422 socket" when it tried to send the third.
7424 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7426 27. The extern definition of crypt16() in expand.c was not being excluded when
7427 the OS had its own crypt16() function.
7429 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7430 as a preferred synonym for return_size_limit, both as an option and as an
7433 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7435 30. Changed the default configuration ACL to relax the local part checking rule
7436 for addresses that are not in any local domains. For these addresses,
7437 slashes and pipe symbols are allowed within local parts, but the sequence
7438 /../ is explicitly forbidden.
7440 31. SPA server authentication was not clearing the challenge buffer before
7443 32. log_message in a "warn" ACL statement was writing to the reject log as
7444 well as to the main log, which contradicts the documentation and doesn't
7445 seem right (because no rejection is happening). So I have stopped it.
7447 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7448 However, I am unable to do any testing of this.
7450 34. Fixed an infelicity in the appendfile transport. When checking directories
7451 for a mailbox, to see if any needed to be created, it was accidentally
7452 using path names with one or more superfluous leading slashes; tracing
7453 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7455 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7456 contents are added to the log line that is written for every discarded
7457 recipient. (Previously a log_message setting was ignored.)
7459 36. The ${quote: operator now quotes the string if it is empty.
7461 37. The install script runs exim in order to find its version number. If for
7462 some reason other than non-existence or emptiness, which it checks, it
7463 could not run './exim', it was installing it with an empty version number,
7464 i.e. as "exim-". This error state is now caught, and the installation is
7467 38. An argument was missing from the function that creates an error message
7468 when Exim fails to connect to the socket for saslauthd authentication.
7469 This could cause Exim to crash, or give a corrupted message.
7471 39. Added isip, isip4, and isip6 to ${if conditions.
7473 40. The ACL variables $acl_xx are now saved with the message, and can be
7474 accessed later in routers, transports, and filters.
7476 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7477 strings in the file are not string-expanded.
7479 42. If a MAIL command specified a SIZE value that was too large to fit into an
7480 int variable, the check against message_size_limit failed. Such values are
7481 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7482 day this will have to be increased, but I don't think I want to be around
7483 when emails are that large.
7490 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7491 iconv() is not standard in FreeBSD.
7493 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7494 IPv6 enabled. The observed symptom was a segmentation fault on return from
7495 the function os_common_find_running_interfaces() in src/os.c.
7497 3. In the check_special_case() function in daemon.c I had used "errno" as an
7498 argument name, which causes warnings on some systems. This was basically a
7499 typo, since it was named "eno" in the comments!
7501 4. The code that waits for the clock to tick (at a resolution of some fraction
7502 of a second) so as to ensure message-id uniqueness was always waiting for
7503 at least one whole tick, when it could have waited for less. [This is
7504 almost certainly not relevant at current processor speeds, where it is
7505 unlikely to ever wait at all. But we try to future-proof.]
7507 5. The function that sleeps for a time interval that includes fractions of a
7508 second contained a race. It did not block SIGALRM between setting the
7509 timer, and suspending (a couple of lines later). If the interval was short
7510 and the sigsuspend() was delayed until after it had expired, the suspension
7511 never ended. On busy systems this could lead to processes getting stuck for
7514 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7515 process, before it forks any delivery processes. The open lookup caching
7516 mechanism meant that the open file or database connection was passed into
7517 the delivery process. The problem was that delivery processes always tidy
7518 up cached lookup data. This could cause a problem for the next delivery
7519 process started by the queue runner, because the external queue runner
7520 process does not know about the closure. So the next delivery process
7521 still has data in the lookup cache. In the case of a file lookup, there was
7522 no problem because closing a file descriptor in a subprocess doesn't affect
7523 the parent. However, if the lookup was caching a connection to a database,
7524 the connection was closed, and the second delivery process was likely to
7525 see errors such as "PGSQL: query failed: server closed the connection
7526 unexpectedly". The problem has been fixed by closing all cached lookups
7527 in a queue runner before running a delivery process.
7529 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7530 seem to have the "const" qualifier which it has on other OS. I've
7533 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7534 *of the same type* (client or server) with the same public name that an
7535 error should be diagnosed.
7537 9. When Exim looked up a host name for an IP address, but failed to find the
7538 original IP address when looking up the host name (a safety check), it
7539 output the message "<ip address> does not match any IP for NULL", which was
7540 confusing, to say the least. The bug was that the host name should have
7541 appeared instead of "NULL".
7543 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7544 user that is built into the binary, and the -C or -D options is used, root
7545 privilege is dropped before the configuration file is read. In addition,
7546 logging is switched to stderr instead of the normal log files. If the
7547 configuration then re-defines the Exim user, the unprivileged environment
7548 is probably not what is expected, so Exim logs a panic warning message (but
7551 However, if deliver_drop_privilege is set, the unprivileged state may well
7552 be exactly what is intended, so the warning has been cut out in that case,
7553 and Exim is allowed to try to write to its normal log files.
7559 1. smtp_return_error_details was not giving details for temporary sender
7560 or receiver verification errors.
7562 2. Diagnose a configuration error if two authenticators have the same public
7565 3. Exim used not to create the message log file for a message until the first
7566 delivery attempt. This could be confusing when incoming messages were held
7567 for policy or load reasons. The message log file is now created at the time
7568 the message is received, and an initial "Received" line is written to it.
7570 4. The automatically generated man page for command line options had a minor
7571 bug that caused no ill effects; however, a more serious problem was that
7572 the procedure for building the man page automatically didn't always
7573 operate. Consequently, release 4.20 contains an out-of-date version. This
7574 shouldn't happen again.
7576 5. When building Exim with embedded Perl support, the script that builds the
7577 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7578 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7580 6. The freeze_tell option was not being used for messages that were frozen on
7581 arrival, either by an ACL or by local_scan().
7583 7. Added the smtp_incomplete_transaction log selector.
7585 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7586 was accepting AUTH without a new EHLO.
7588 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7589 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7592 10. Logging of TCP/IP connections (when configured) now happens in the main
7593 daemon process instead of the child process, so that the TCP/IP connection
7594 count is more accurate (but it can never be perfect).
7596 11. The use of "drop" in a nested ACL was not being handled correctly in the
7597 outer ACL. Now, if condition failure induced by the nested "drop" causes
7598 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7599 or "require"), the connection is dropped.
7601 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7602 that yield "discard" can now be used with an "accept" or a "discard" verb,
7603 but an error is generated for any others (because I can't see a useful way
7604 to define what should happen).
7606 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7607 are now processed in the same way as lines in the Exim configuration file.
7608 In particular, continuation lines are supported.
7610 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7612 15. Added -ti meaning -t -i.
7614 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7615 domains, and warn by logging if others are found.
7617 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7618 returned from GIFCONF when Exim is trying to find the list of interfaces on
7619 a host. The code in os.c has been modified to copy each ifreq to an aligned
7620 structure in all cases.
7622 Also, in some cases, the returned ifreq's were being copied to a 'struct
7623 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7624 means the last couple of bytes of an IPv6 address could be chopped if the
7625 ifreq contained only a normal sockaddr (14 bytes storage).
7627 18. Named domain lists were not supported in the hosts_treat_as_local option.
7628 An entry such as +xxxx was not recognized, and was treated as a literal
7631 19. Ensure that header lines added by a DATA ACL are included in the reject log
7632 if the ACL subsequently rejects the message.
7634 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7635 MD5 (which is deprecated).
7637 21. When testing a filter file using -bf, Exim was writing a message when it
7638 took the sender from a "From " line in the message, but it was not doing so
7639 when it took $return_path from a Return-Path: header line. It now does.
7641 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7642 with a valid header line field name (a series of printing characters
7643 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7645 23. Changed "disc" in the source to "disk" to conform to the documentation and
7646 the book and for uniformity.
7648 24. Ignore Sendmail's -Ooption=value command line item.
7650 25. When execve() failed while trying to run a command in a pipe transport,
7651 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7652 could be confused with a return value of 69 from the command itself. This
7653 has been changed to 127, the value the shell returns if it is asked to run
7654 a non-existent command. The wording for the related log line suggests a
7655 non-existent command as the problem.
7657 26. If received_header_text expands to an empty string, do not add a Received:
7658 header line to the message. (Well, it adds a token one on the spool, but
7659 marks it "old" so that it doesn't get used or transmitted.)
7661 27. Installed eximstats 1.28 (addition of -nt option).
7663 28. There was no check for failure on the call to getsockname() in the daemon
7664 code. This can fail if there is a shortage of resources on the system, with
7665 ENOMEM, for example. A temporary error is now given on failure.
7667 29. Contrary to the C standard, it seems that in some environments, the
7668 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7669 program. Exim now does this explicitly; it affects the formatting of
7670 timestamps using strftime().
7672 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7673 complaints. I've now initialized all the variables, to avoid this.
7675 32. Header lines added by a system filter were not being "seen" during
7676 transport-time rewrites.
7678 33. The info_callback() function passed to OpenSSL is set up with type void
7679 (*)(SSL *, int, int), as described somewhere. However, when calling the
7680 function (actually a macro) that sets it up, the type void(*)() is
7681 expected. I've put in a cast to prevent warnings from picky compilers.
7683 34. If a DNS black list lookup found a CNAME record, but there were no A
7684 records associated with the domain it pointed at, Exim crashed.
7686 35. If a DNS black list lookup returned more than one A record, Exim ignored
7687 all but the first. It now scans all returned addresses if a particular IP
7688 value is being sought. In this situation, the contents of the
7689 $dnslist_value variable are a list of all the addresses, separated by a
7692 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7693 to accept a host name and all its aliases if the forward lookup for any of
7694 them yielded the IP address of the incoming connection. Now it accepts only
7695 those names whose forward lookup yields the correct IP address. Any other
7696 names are discarded. This closes a loophole whereby a rogue DNS
7697 administrator could create reverse DNS records to break through a
7698 wildcarded host restriction in an ACL.
7700 37. If a user filter or a system filter that ran in a subprocess used any of
7701 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7702 the wrong values were passed to the pipe command ($thisaddress had the
7703 value of $0, $0 had the value of $1, etc). This bug was introduced by
7704 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7706 38. Improved the line breaking for long SMTP error messages from ACLs.
7707 Previously, if there was no break point between 40 and 75 characters, Exim
7708 left the rest of the message alone. Two changes have been made: (a) I've
7709 reduced the minimum length to 35 characters; (b) if it can't find a break
7710 point between 35 and 75 characters, it looks ahead and uses the first one
7711 that it finds. This may give the occasional overlong line, but at least the
7712 remaining text gets split now.
7714 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7715 file descriptors might be low, and that setting 1000 would always raise it.
7716 It turns out that in some environments, the limit is already over 1000 and
7717 that lowering it causes trouble. So now Exim takes care not to decrease it.
7719 40. When delivering a message, the value of $return_path is set to $sender_
7720 address at the start of routing (routers may change the value). By an
7721 oversight, this default was not being set up when an address was tested by
7722 -bt or -bv, which affected the outcome if any router or filter referred to
7725 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7726 log only when "message" or "log_message" are set. However, if one of the
7727 conditions was an address verification, or a call to a nested ACL, the
7728 messages generated by the underlying test were being passed through. This
7729 no longer happens. The underlying message is available in $acl_verify_
7730 message for both "message" and "log_message" expansions, so it can be
7731 passed through if needed.
7733 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7734 new expansion $bh_ to give the encoded byte string without charset
7735 translation. Translation happens only if iconv() is available; HAVE_ICONV
7736 indicates this at build time. HEADERS_CHARSET gives the charset to
7737 translate to; headers_charset can change it in the configuration, and
7738 "headers charset" can change it in an individual filter file.
7740 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7741 that creates RFC 2047 encoded "words" labels them as that charset instead
7742 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7743 expansion operator; (ii) when Exim creates a From: line for a local
7744 message; (iii) when a header line is rewritten to include a "phrase" part.
7746 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7747 buggy, causing it to skip the first lines of messages whose message ID
7748 ended in 'D'. This would not have bitten before Exim release 4.14, because
7749 message IDs were unlikely to end in 'D' before then. The effect was to have
7750 incorrect size information for certain domains.
7752 45. #include "config.h" was missing at the start of the crypt16.c module. This
7753 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7756 46. If there was a timeout during a "random" callout check, Exim treated it as
7757 a failure of the random address, and carried on sending RSET and the real
7758 address. If the delay was just some slowness somewhere, the response to the
7759 original RCPT would be taken as a response to RSET and so on, causing
7760 mayhem of various kinds.
7762 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7763 when I implemented it. It didn't allow for the fact that some option values
7764 may legitimately be negative (e.g. size_addition), and it didn't even do
7765 the right test for positive values.
7767 48. Domain names in DNS records are case-independent. Exim always looks them up
7768 in lower case. Some resolvers return domain names in exactly the case they
7769 appear in the zone file, that is, they may contain uppercase letters. Not
7770 all resolvers do this - some return always lower case. Exim was treating a
7771 change of case by a resolver as a change of domain, similar to a widening
7772 of a domain abbreviation. This triggered its re-routing code and so it was
7773 trying to route what was effectively the same domain again. This normally
7774 caused routing to fail (because the router wouldn't handle the domain
7775 twice). Now Exim checks for this case specially, and just changes the
7776 casing of the domain that it ultimately uses when it transmits the message
7779 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7782 50. If a filter generated a file delivery with a non-absolute name (possible if
7783 no home directory exists for the router), the forbid_file option was not
7786 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7787 the existing equality matching.
7789 52. Exim was using ints instead of ino_t variables in some places where it was
7790 dealing with inode numbers.
7792 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7793 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7794 called TMPDIR, and if it finds it is different, it changes its value.
7796 54. The smtp_printf() function is now made available to local_scan() so
7797 additional output lines can be written before returning. There is also an
7798 smtp_fflush() function to enable the detection of a dropped connection.
7799 The variables smtp_input and smtp_batched_input are exported to
7802 55. Changed the default runtime configuration: the message "Unknown user"
7803 has been removed from the ACL, and instead placed on the localuser router,
7804 using the cannot_route_message feature. This means that any verification
7805 failures that generate their own messages won't get overridden. Similarly,
7806 the "Unrouteable address" message that was in the ACL for unverifiable
7807 relay addresses has also been removed.
7809 56. Added hosts_avoid_esmtp to the smtp transport.
7811 57. The exicyclog script was not checking for the esoteric option
7812 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7813 will work only if exicyclog is run under the appropriate euid.
7815 58. Following a discussion on the list, the rules by which Exim recognises line
7816 endings on incoming messages have been changed. The -dropcr and drop_cr
7817 options are now no-ops, retained only for backwards compatibility. The
7818 following line terminators are recognized: LF CRLF CR. However, special
7819 processing applies to CR:
7821 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7822 nor a local message in the state where . is a terminator.
7824 (ii) If a bare CR is encountered in a header line, an extra space is added
7825 after the line terminator so as not to end the header. The reasoning
7826 behind this is that bare CRs in header lines are most likely either
7827 to be mistakes, or people trying to play silly games.
7829 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7830 was being incorrectly given as 18 bytes larger than it should have been.
7831 This is a VOB (very old bug).
7833 60. This may never have affected anything current, but just in case it has:
7834 When the local host is found other than at the start of a list of hosts,
7835 the local host, those with the same MX, and any that follow, are discarded.
7836 When the list in question was part of a longer list of hosts, the following
7837 hosts (not currently being processed) were also being discarded. This no
7838 longer happens. I'm not sure if this situation could ever has previously
7841 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7844 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7845 header. This is something that is recommended in a new Internet Draft, and
7846 is something that is documented as being done by Sendmail. There are two
7847 possible values. For messages generated by the autoreply transport, Exim
7850 Auto-Submitted: auto-replied
7852 whereas for all other generated messages (e.g. bounces) it adds
7854 Auto-Submitted: auto-generated
7856 63. The "personal" condition in filters now includes a test for the
7857 Auto-Submitted: header. If it contains the string "auto-" the message it
7858 not considered personal.
7860 64. Added rcpt_include_affixes as a generic transport option.
7862 65. Added queue_only_override (default true).
7864 66. Added the syslog_duplication option.
7866 67. If what should have been the first header line of a message consisted of
7867 a space followed by a colon, Exim was mis-interpreting it as a header line.
7868 It isn't of course - it is syntactically invalid and should therefore be
7869 treated as the start of the message body. The misbehaviour could have
7870 caused a number of strange effects, including loss of data in subsequent
7871 header lines, and spool format errors.
7873 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7874 client host had authenticated. This control can now be exercised by an ACL
7875 for more flexibility.
7877 69. By default, callouts do not happen when testing with -bh. There is now a
7878 variant, -bhc, which does actually run the callout code, including
7879 consulting and updating the callout cache.
7881 70. Added support for saslauthd authentication, courtesy of Alexander
7884 71. If statvfs() failed on the spool or log directories while checking their
7885 size for availability, Exim confusingly gave the error "space shortage".
7886 Furthermore, in debugging mode it crashed with a floating point exception.
7887 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7888 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7889 serious problem, Exim now writes to the main and panic logs when this
7890 happens, with details of the failure. It then refuses to accept the
7891 incoming message, giving the message "spool directory problem" or "log
7892 directory problem" with a 421 code for SMTP messages.
7894 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7895 0, 1, and 2 exist, because some OS complain for execs without them (see
7896 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7897 descriptors. However, the code omitted to check that the open succeeded,
7898 causing mysterious errors if for some reason the permissions on /dev/null
7899 got screwed. Now Exim writes a message to the main and panic logs, and
7900 bombs out if it can't open /dev/null.
7902 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7903 interact so that it is all more flexible. It is supposed to remain
7904 backwards compatible. Also added extra_local_interfaces.
7906 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7907 to bomb out with an assertion failure - to the client this appears as a
7908 connection drop. This problem occurs in the part of the code that was taken
7909 from the Samba project. Fortunately, the assertion is in a very simple
7910 function, so I have fixed this by reproducing the function inline in the
7911 one place where it is called, and arranging for authentication to fail
7912 instead of killing the process with assert().
7914 75. The SPA client code was not working when the server requested OEM rather
7915 than Unicode encoding.
7917 76. Added code to make require_files with a specific uid setting more usable in
7918 the case where statting the file as root fails - usually a non-root-mounted
7919 NFS file system. When this happens and the failure is EACCES, Exim now
7920 forks a subprocess and does the per-uid checking as the relevant uid.
7922 77. Added process_log_path.
7924 78. If log_file_path was not explicitly set, a setting of check_log_space or
7925 check_log_inodes was ignored.
7927 79. If a space check for the spool or log partitions fails, the incident is now
7928 logged. Of course, in the latter case the data may get lost...
7930 80. Added the %p formatting code to string_format() so that it can be used to
7931 print addresses in debug_print(). Adjusted all the address printing in the
7932 debugging in store.c to use %p rather than %d.
7934 81. There was a concern that a line of code in smtp_in.c could overflow a
7935 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7936 initially expressed, the concern was not well-founded, because trailing
7937 spaces are removed early. However, if the trailing spaces were followed by
7938 a NULL, they did not get removed, so the overflow was possible. Two fixes
7941 (a) I re-wrote the offending code in a cleaner fashion.
7942 (b) If an incoming SMTP command contains a NULL character, it is rejected
7945 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7946 runs initgroups(), so that if the Exim user is in any additional groups,
7947 they will be used during message reception.
7953 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob