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,
170 JH/01 Avoid costly startup code when not strictly needed. This reduces time
171 for some exim process initialisations. It does mean that the logging
172 of TLS configuration problems is only done for the daemon startup.
174 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
176 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
177 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
180 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
181 DKIM signing is being done. Previously a transport_filter would always
182 disable CHUNKING, falling back to traditional DATA.
184 JH/05 Regard command-line receipients as tainted.
186 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
188 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
189 PAM library frees one of the arguments given to it, despite the
190 documentation. Therefore a plain malloc must be used.
192 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
193 on-stack buffers were used, resulting in a taint trap when DSN information
194 copied from a received message was written into the buffer.
196 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
197 the ordering of its ARC headers. This caused a crash.
199 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
200 a new record was being constructed with information from the peer, a trap
203 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
204 installation would get error messages from DMARC verify, when it hit the
205 nonexistent file indicated by the default. Distros wanting DMARC enabled
206 should both provide the file and set the option.
207 Also enforce no DMARC verification for command-line sourced messages.
209 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
210 could, depending on the platform, hang at the STARTTLS response.
212 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
213 message on a connection. Previously if one message had ARC headers and
214 the following one did not, a crash could result when adding an
215 Authentication-Results: header.
217 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
218 Exim main code and Exim-related utities. The introduction of taint
219 tracking also did many adjustments to string handling. Since then, eximon
220 frequently terminated with an assert failure.
222 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
223 check for 452 responses. This slightly helps the inefficieny of doing
224 a large alias-expansion into a recipient-limited target. The max_rcpt
225 transport option still applies (and at the current default, will override
226 the new feature). The check is done for either cause of synch, and forces
227 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
228 connection. The new facility is not tunable at this time.
230 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
231 library live data was being used, so the results became garbage. Make
232 copies while it is still usable.
234 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
235 on delivery deferred (==) and failed (**) lines (if a delivery was
236 attemtped). Previously it was only on completion (=>) lines.
238 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
239 for the expansion of the server_scram_iter and server_scram_salt options.
241 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
242 are now specifically given a NO_DATA response without hitting the system
243 resolver. The library goes on to do the now-standard TXT lookup.
244 Use of dnsdb lookups is not affected.
246 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
247 only retrieve the errormessage once. Previously two calls to dlerror()
248 were used, and the second one (for mainlog/paniclog) retrieved null
251 JH/20 Taint checking: disallow use of tainted data for
252 - the appendfile transport file and directory options
253 - the pipe transport command
254 - the autoreply transport file, log and once options
255 - file names used by the redirect router (including filter files)
257 - paths used by single-key lookups
258 Previously this was permitted.
260 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
261 adjusted the size of a major service buffer; this failed because the
262 buffer was in use at the time. Change to a compile-time increase in the
263 buffer size, when this authenticator is compiled into exim.
265 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
266 previous fast-mode was untenable in the face of glibs using mmap to
267 support larger malloc requests.
269 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
270 New values supported, if defined on system where compiled:
271 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
272 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
274 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
275 running a limited number of proceses in parallel, a benefit is gained. The
276 amount varies with the platform hardware and load. The use of the option
277 queue_run_in_order means we cannot do this, as ordering becomes
280 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
281 had introduced a string-copy (for ensuring NUL-termination) which was not
282 appropriate for that case, which can include embedded NUL bytes in the
283 block of data. Investigation showed the copy to actually be needless, the
284 data being length-specified.
286 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
287 done during a receiving connection, and both used TLS, global info was
288 used rather than per-connection info for tracking the state of data
289 queued for transmission. This could result in a connection hang.
291 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
292 Previously, when delivering serveral messages down a single connection
293 only the first would provide a SIZE. This was due to the size information
294 not being properly tracked.
296 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
297 TAI (at 37 seconds currently), pretend to be in UTC for time-related
298 expansion and logging. Previously, spurious values such as a future
299 minute could be seen.
301 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
302 it could crash from a null-deref. This could also affect the
303 ${addresses: } operator and ${readsock } item.
305 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
306 message following a mime one, the variable was not reset.
308 JH/30 When an pipelined-connect fails at the first response, assume incorrect
309 cached capability (perhaps the peer reneged?) and immediately retry in
312 JH/31 Fix spurious detection of timeout while writing to transport filter.
314 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
315 an attempt to copy the string was made before checking it.
317 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
318 taint of the lookup key was maintained; we now regard the presence in the
319 filesystem as sufficient validation.
321 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
322 argument is supplied.
324 JH/35 The dsearch lookup now requires that the directory is an absolute path.
325 Previously this was not checked, and nonempty relative paths made an
326 access under Exim's current working directory.
328 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
329 Previously no event was raised.
331 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
332 parameter supplied by the sender MAIL FROM command. Previously it was
333 ignored, and only the check_spool_space option value for the required
336 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
337 the size of the signing public-key. Previously it was instead giving
338 the size of the signature hash.
340 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
341 the default. See the (new) dkim_verify_min_keysizes option.
343 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
344 and an ACL use a lookup for checking either the local_part or domain,
345 stale data could be accessed. Ensure that variable references are
346 dropped between messages.
348 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
349 by the client was not checked as pointing within response data before
350 being used. A malicious client could thus cause an out-of-bounds read and
351 possibly gain authentication. Fix by adding the check.
353 JH/42 Internationalisation: change the default for downconversion in the smtp
354 transport to be "if needed". Previously it was "as previously set" for
355 the message, which usually meant "if needed" for message-submission but
356 "no" for everything else. However, MTAs have been seen using SMTPUTF8
357 even when the envelope addresses did not need it, resulting in forwarding
358 failures to non-supporting MTAs. A downconvert in such cases will be
359 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
360 transport. The change does mean that addresses needing conversion will
361 be converted when previously a delivery failure would occur.
363 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
364 response was received it would be used unchecked in a fail-DSN, violating
365 standards on line-length limits. Truncate if needed.
367 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
368 doesn't work. (Additionally add a new main config option to configure the
375 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
376 side, to decode a TLS 1.3 packet capture.
378 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
379 Previously the default library behaviour applied, sending two, each in
382 JH/03 Debug output for ACL now gives the config file name and line number for
385 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
387 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
389 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
390 buffer overrun for (non-chunking) other transports.
392 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
393 TLS1.3, means that a server rejecting a client certificate is not visible
394 to the client until the first read of encrypted data (typically the
395 response to EHLO). Add detection for that case and treat it as a failed
396 TLS connection attempt, so that the normal retry-in-clear can work (if
397 suitably configured).
399 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
400 and/or domain. Found and fixed by Jason Betts.
402 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
403 configuration). If a CNAME target was not a wellformed name pattern, a
406 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
407 the OS reports them interleaved with other addresses.
409 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
410 used both for input and for a verify callout, both encrypted, SMTP
411 responses being sent by the server could be lost. This resulted in
412 dropped connections and sometimes bounces generated by a peer sending
415 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
416 string. Previously it was possible to cause undefined behaviour in a
417 library routine (usually a crash). Found by "zerons".
419 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
422 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
423 API was removed, so update to use the newer ones.
425 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
426 any timeout set, is taking a long time. Previously we would hang on to a
427 rotated logfile "forever" if the input was arriving with long gaps
428 (a previous attempt to fix addressed lack, for a long time, of initial
431 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
432 shared (NFS) environment. The length of the tempfile name is now
433 4 + 16 ("hdr.$message_exim_id") which might break on file
434 systems which restrict the file name length to lower values.
437 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
438 shared (NFS) environment.
440 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
441 did for all versions <4.90). Notably -M, -m, --invert, -I may be
444 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
445 on some platforms for bit 31.
447 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
448 to changes apparently associated with TLS1.3 handling some of the APIs
449 previously used were either nonfunctional or inappropriate. Strings
450 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
451 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
452 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
453 This affects log line X= elements, the $tls_{in,out}_cipher variables,
454 and the use of specific cipher names in the encrypted= ACL condition.
456 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
458 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
459 verification result was not updated unless hosts_require_ocsp applied.
461 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
462 queue_list_requires_admin set to false, non-admin users were denied the
465 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
466 directory-of-certs mode. Previously they were advertised despite the
469 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
470 A single TCP connection by a client will now hold a TLS connection open
471 for multiple message deliveries, by default. Previously the default was to
474 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
475 default. If built with the facility, DANE will be used. The facility
476 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
478 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
479 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
480 must be defined and you must still, unless you define DISABLE_TLS, manage
481 the the include-dir and library-file requirements that go with that
482 choice. Non-TLS builds are still supported.
484 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
487 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
488 default. If the platform supports and has the facility enabled, it will
489 be requested on all coneections.
491 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
492 controlled by the build-time option SUPPORT_PIPE_CONNECT.
494 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
496 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
497 success-DSN messages. Previously the From: header was always the default
498 one for these; the option was ignored.
500 JH/28 Fix the timeout on smtp response to apply to the whole response.
501 Previously it was reset for every read, so a teergrubing peer sending
502 single bytes within the time limit could extend the connection for a
503 long time. Credit to Qualsys Security Advisory Team for the discovery.
505 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
506 delivery address, which leaked information of the results of local
507 forwarding. Change to the original envelope recipient address, per
510 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
511 requested. Previously not bounce was generated and a log entry of
512 error ignored was made.
514 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
516 JH/32 Introduce a general tainting mechanism for values read from the input
517 channel, and values derived from them. Refuse to expand any tainted
518 values, to catch one form of exploit.
520 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
521 was unused and the unexpanded text used for the test. Found and
522 fixed by Ruben Jenster.
524 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
525 an attempt to use a TLS library read routine dereffed a nul pointer,
528 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
529 them in our smtp response.
531 JH/36 Have the general router option retry_use_local_part default to true when
532 any of the restrictive preconditions are set (to anything). Previously it
533 was only for check_local user. The change removes one item of manual
534 configuration which is required for proper retries when a remote router
535 handles a subset of addresses for a domain.
537 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
538 link count into consideration.
540 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
541 caused the extension of big_buffer, the following lines were ignored.
543 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
544 accordance with RFC 2308. Previously there was no expiry, so a longlived
545 receive process (eg. due to ACL delays) versus a short SOA value could
548 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
550 JH/39 Promote DMARC support to mainline.
552 JH/40 Bug 2452: Add a References: header to DSNs.
554 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
555 parameters. The relevant library call is documented as "Deprecated: This
556 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
557 3.6.0, DH parameters are negotiated following RFC7919."
559 HS/06 Change the default of dnssec_request_domains to "*"
561 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
562 carried on and emitted a BDAT command, even when PIPELINING was not
565 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
566 buffer was used for the filename, resulting in a trap when tainted
567 arguments (eg. $domain) were used.
569 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
570 recommended to avoid a possible server-load attack. The feature can be
571 re-enabled via the openssl_options main cofiguration option.
573 JH/45 local_scan API: documented the current smtp_printf() call. This changed
574 for version 4.90 - adding a "more data" boolean to the arguments.
575 Bumped the ABI version number also, this having been missed previously;
576 release versions 4.90 to 4.92.3 inclusive were effectively broken in
577 respect of usage of smtp_printf() by either local_scan code or libraries
578 accessed via the ${dlfunc } expansion item. Both will need coding
579 adjustment for any calls to smtp_printf() to match the new function
580 signature; a FALSE value for the new argument is always safe.
582 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
583 the file-offset (which the Linux syscall does, and exim expects); this
584 resulted in an indefinite loop.
586 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
587 to do ARC verification. The Authentication-Results: header line added
588 by the configuration then had no ARC item.
594 JH/01 Remove code calling the customisable local_scan function, unless a new
595 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
597 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
598 non-signal-safe functions being used.
600 JH/03 Bug 2269: When presented with a received message having a stupidly large
601 number of DKIM-Signature headers, disable DKIM verification to avoid
602 a resource-consumption attack. The limit is set at twenty.
604 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
605 report of oldest_pass in ${authres } in consequence, and separate out
606 some descriptions of reasons for verification fail.
608 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
609 files in the spool were present and unlocked. A queue-runner could spot
610 them, resulting in a duplicate delivery. Fix that by doing the unlock
611 after the unlink. Investigation by Tim Stewart. Take the opportunity to
612 add more error-checking on spoolfile handling while that code is being
615 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
616 No known attacks, no CVE, this is defensive hardening.
618 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
619 a queue-runner could start a delivery while other operations were ongoing.
620 Cutthrough delivery was a common victim, resulting in duplicate delivery.
621 Found and investigated by Tim Stewart. Fix by using the open message data
622 file handle rather than opening another, and not locally closing it (which
623 releases a lock) for that case, while creating the temporary .eml format
624 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
626 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
627 $sender_verify_failure/$recipient_verify_failure to "random".
629 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
632 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
633 Previously this would segfault.
635 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
638 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
639 like zero, since the resolver should be doing this for us, But we need one
640 as a CNAME but no MX presence gets the CNAME returned; we need to check
641 that doesn't point to an MX to declare it "no MX returned" rather than
642 "error, loop". A new main option is added so the older capability of
643 following some limited number of chain links is maintained.
645 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
647 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
648 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
649 now-deprecated earlier definitions used only specified the range up to TLS
650 1.2 (in the older-version library docs).
652 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
654 JH/15 Rework TLS client-side context management. Stop using a global, and
655 explicitly pass a context around. This enables future use of TLS for
656 connections to service-daemons (eg. malware scanning) while a client smtp
657 connection is using TLS; with cutthrough connections this is quite likely.
659 JH/16 Fix ARC verification to do AS checks in reverse order.
661 JH/17 Support a "tls" option on the ${readsocket } expansion item.
663 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
664 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
665 Previously the "utf8" would be re-prepended for every additional message.
667 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
668 Previously thery were accepted, resulting in issues when attempting to
669 forward messages to a non-supporting MTA.
671 PP/02 Let -n work with printing macros too, not just options.
673 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
674 one parent address was copied, and bogus data was used at delivery-logging
675 time. Either a crash (after delivery) or bogus log data could result.
676 Discovery and analysis by Tim Stewart.
678 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
679 Previously if the string ended mid-character, we did not insert the
680 promised '?' replacement.
682 PP/04 Documentation: current string operators work on bytes, not codepoints.
684 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
685 should pack well giving a smaller memory footprint so better caching and
686 therefore performance. Group the declarations where this can't be done so
687 that the byte-sized flag variables are not interspersed among pointer
688 variables, giving a better chance of good packing by the compiler.
690 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
691 non-null, to avoid issues with sites running BATV. Previously reports were
692 sent with an empty envelope sender so looked like bounces.
694 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
695 The ignore_error flag wasn't being returned from the filter subprocess so
696 was not set for later routers. Investigation and fix by Matthias Kurz.
698 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
699 and a msg:complete for the whole, when a message is manually removed using
700 -Mrm. Developement by Matthias Kurz, hacked on by JH.
702 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
703 a "Gnu special" function, asprintf() in the DB utility binary builds; I
704 hope that is portable enough.
706 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
707 requiring a known-CA anchor certificate; make it now rely entirely on the
708 TLSA as an anchor. Checking the name on the leaf cert against the name
709 on the A-record for the host is still done for TA (but not for EE mode).
711 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
712 included in delivery lines for non-proxied connections, when compiled with
713 SUPPORT_SOCKS and running with proxy logging enabled.
715 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
716 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
717 move the existing event to fire before the normal logging of message
718 failure so that custom logging is bracketed by normal logging.
720 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
721 msg:fail:internal event. Developement by Matthias Kurz.
723 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
724 far too small for todays use of crypto signatures stored there. Go all
725 the way to the max DNS message size of 64kB, even though this might be
726 overmuch for IOT constrained device use.
728 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
729 copy a string over itself. The library routine is documented as not
730 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
732 JH/32 For main options check_spool_space and check_inode_space, where the
733 platform supports 64b integers, support more than the previous 2^31 kB
734 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
735 the previous G, M, k.
737 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
738 $authenticated_fail_id variable on authentication failure. Previously
741 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
742 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
743 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
744 GNUTLS_SEC_PARAM_MEDIUM.
746 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
747 side. Previously we would continue as if no SNI had been received.
749 JH/36 Harden the handling of string-lists. When a list consisted of a sole
750 "<" character, which should be a list-separator specification, we walked
751 off past the nul-terimation.
753 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
754 causes) even when the retry time is not yet met. Previously they were
755 not, meaning that when (say) an account was over-quota and temp-rejecting,
756 and multiple senders' messages were queued, only one sender would get
757 notified on each configured delay_warning cycle.
759 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
761 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
762 cipher-suites, an error can be left on the stack even for a succeeding
763 accept; this results in impossible error messages when a later operation
766 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
767 return error codes indicating retry. Under TLS1.3 this becomes required.
769 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
770 it only wrote the new authenticators, resulting in a lack of tracking of
771 peer changes of ESMTP extensions until the next cache flush.
773 JH/41 Fix the loop reading a message header line to check for integer overflow,
774 and more-often against header_maxsize. Previously a crafted message could
775 induce a crash of the recive process; now the message is cleanly rejected.
777 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
778 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
784 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
785 When redis_servers is set to a list of > 1 element, and the Redis servers
786 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
787 case of MOVED into a DEFER case instead, thus moving the query onto the
788 next server in the list. For a cluster of N elements, all N servers must
789 be defined in redis_servers.
791 GF/02 Catch and remove uninitialized value warning in exiqsumm
792 Check for existence of @ARGV before looking at $ARGV[0]
794 JH/01 Replace the store_release() internal interface with store_newblock(),
795 which internalises the check required to safely use the old one, plus
796 the allocate and data copy operations duplicated in both (!) of the
797 extant use locations.
799 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
800 modifier. This matches the restriction on the commandline.
802 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
803 Previously only the last row was returned.
805 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
806 we assumed that tags in the header were well-formed, and parsed the
807 element content after inspecting only the first char of the tag.
808 Assumptions at that stage could crash the receive process on malformed
811 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
812 While running the DKIM ACL we operate on the Permanent memory pool so that
813 variables created with "set" persist to the DATA ACL. Also (at any time)
814 DNS lookups that fail create cache records using the Permanent pool. But
815 expansions release any allocations made on the current pool - so a dnsdb
816 lookup expansion done in the DKIM ACL releases the memory used for the
817 DNS negative-cache, and bad things result. Solution is to switch to the
818 Main pool for expansions.
819 While we're in that code, add checks on the DNS cache during store_reset,
820 active in the testsuite.
821 Problem spotted, and debugging aided, by Wolfgang Breyha.
823 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
824 When none of the hosts presented to a transport match an already-open
825 connection, close it and proceed with the list. Previously we would
826 queue the message. Spotted by Lena with Yahoo, probably involving
829 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
830 Previously a spurious "250 OK id=" response was appended to the proper
833 JH/08 The "support for" informational output now, which built with Content
834 Scanning support, has a line for the malware scanner interfaces compiled
835 in. Interface can be individually included or not at build time.
837 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
838 by the template makefile "src/EDITME". The "STREAM" support for an older
839 ClamAV interface method is removed.
841 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
842 rows affected is given instead).
844 JH/11 The runtime Berkeley DB library version is now additionally output by
845 "exim -d -bV". Previously only the compile-time version was shown.
847 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
848 SMTP connection. Previously, when one had more receipients than the
849 first, an abortive onward connection was made. Move to full support for
850 multiple onward connections in sequence, handling cutthrough connection
851 for all multi-message initiating connections.
853 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
854 routers. Previously, a multi-recipient message would fail to match the
855 onward-connection opened for the first recipient, and cause its closure.
857 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
858 a timeout on read on a GnuTLS initiating connection, resulting in the
859 initiating connection being dropped. This mattered most when the callout
860 was marked defer_ok. Fix to keep the two timeout-detection methods
863 JH/15 Relax results from ACL control request to enable cutthrough, in
864 unsupported situations, from error to silently (except under debug)
865 ignoring. This covers use with PRDR, frozen messages, queue-only and
868 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
870 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
871 metadata, resulting in a crash in free().
873 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
874 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
875 Broken also in d185889f4, with init system revamp.
877 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
878 Previously we abruptly closed the connection after reading a malware-
879 found indication; now we go on to read the "scan ok" response line,
882 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
883 ACL. Previously, a crash would result.
885 JH/19 Speed up macro lookups during configuration file read, by skipping non-
886 macro text after a replacement (previously it was only once per line) and
887 by skipping builtin macros when searching for an uppercase lead character.
889 JH/20 DANE support moved from Experimental to mainline. The Makefile control
890 for the build is renamed.
892 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
893 was allocated for every new TLS startup, meaning one per message. Fix
894 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
896 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
897 reported the original. Fix to report (as far as possible) the ACL
898 result replacing the original.
900 JH/23 Fix memory leak during multi-message connections using STARTTLS under
901 OpenSSL. Certificate information is loaded for every new TLS startup,
902 and the resources needed to be freed.
904 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
906 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
909 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
910 DATA response info to the (existing) per-recipient response info for
911 the "C=" log element. It can have useful tracking info from the
912 destination system. Patch from Simon Arlott.
914 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
915 length value. Previously this would segfault.
917 HS/02 Support Avast multiline protoocol, this allows passing flags to
918 newer versions of the scanner.
920 JH/28 Ensure that variables possibly set during message acceptance are marked
921 dead before release of memory in the daemon loop. This stops complaints
922 about them when the debug_store option is enabled. Discovered specifically
923 for sender_rate_period, but applies to a whole set of variables.
924 Do the same for the queue-runner and queue-list loops, for variables set
925 from spool message files. Do the same for the SMTP per-message loop, for
926 certain variables indirectly set in ACL operations.
928 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
929 as a multi-recipient message from a mailinglist manager). The coding had
930 an arbitrary cutoff number of characters while checking for more input;
931 enforced by writing a NUL into the buffer. This corrupted long / fast
932 input. The problem was exposed more widely when more pipelineing of SMTP
933 responses was introduced, and one Exim system was feeding another.
934 The symptom is log complaints of SMTP syntax error (NUL chars) on the
935 receiving system, and refused recipients seen by the sending system
936 (propating to people being dropped from mailing lists).
937 Discovered and pinpointed by David Carter.
939 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
940 replaced by the ${authresults } expansion.
942 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
944 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
945 allows proper process termination in container environments.
947 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
948 Previously the "final dot" had a newline after it; ensure it is CR,LF.
950 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
951 and "err_perm", deprecated since 4.83 when the RFC-defined words
952 "temperror" and "permerror" were introduced.
954 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
955 transport-filters or DKIM-signing. The restriction was lost in the
956 consolidation of verify-callout and delivery SMTP handling.
957 Extend the restriction to also cover ARC-signing.
959 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
960 in defer=pass mode supply a 450 to the initiator. Previously the message
963 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
964 tls_require_ciphers is used as before.
966 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
967 "pass_unscanned". Only tmpfails from the scanner are written to
968 the paniclog, as they may require admin intervention (permission
969 denied, license issues). Other scanner errors (like decompression
970 bombs) do not cause a paniclog entry.
972 JH/36 Fix reinitialisation of DKIM logging variable between messages.
973 Previously it was possible to log spurious information in receive log
976 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
977 triggered odd behaviour from Outlook Express clients.
979 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
982 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
983 since the IETF WG has not yet settled on that versus the original
984 "bare" representation.
986 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
987 Previously the millisecond value corrupted the output.
988 Fix also for syslog_pid=no and log_selector +pid, for which the pid
989 corrupted the output.
995 JH/01 Rework error string handling in TLS interface so that the caller in
996 more cases is responsible for logging. This permits library-sourced
997 string to be attached to addresses during delivery, and collapses
998 pairs of long lines into single ones.
1000 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1001 during configuration. Wildcards are allowed and expanded.
1003 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1004 This permits better logging.
1006 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1007 do not close it down and have the child transport start it up again on
1008 the passed-on TCP connection. Instead, proxy the child (and any
1009 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1010 affected: the continued delivery log lines do not have any DNSSEC, TLS
1011 Certificate or OCSP information. TLS cipher information is still logged.
1013 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1014 identical IP addresses on different listening ports. Will also affect
1017 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1018 add noisy ifdef guards to special-case this sillyness.
1019 Patch from Bernd Kuhls.
1021 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1022 than 255 are no longer allowed.
1024 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1025 Disable the session-cache too, which might reduce our load. Since we
1026 currrectly use a new context for every connection, both as server and
1027 client, there is no benefit for these.
1028 GnuTLS appears to not support tickets server-side by default (we don't
1029 call gnutls_session_ticket_enable_server()) but client side is enabled
1030 by default on recent versions (3.1.3 +) unless the PFS priority string
1033 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1034 <https://reproducible-builds.org/specs/source-date-epoch/>.
1036 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1037 the check for any unsuccessful recipients did not notice the limit, and
1038 erroneously found still-pending ones.
1040 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1041 MSG_MORE. Only in-clear (not on TLS connections).
1043 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1044 transport-filter is involved we can buffer the headers in memory for
1045 creating the signature, and read the spool data file once for the
1046 signature and again for transmission.
1048 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1049 4.77 as the kernel support then wasn't solid, having issues in 64bit
1050 mode. Now, it's been long enough. Add support for FreeBSD also.
1052 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1053 case where the routing stage had gathered several addresses to send to
1054 a host before calling the transport for the first, we previously failed
1055 to close down TLS in the old transport process before passing the TCP
1056 connection to the new process. The new one sent a STARTTLS command
1057 which naturally failed, giving a failed delivery and bloating the retry
1058 database. Investigation and fix prototype from Wolfgang Breyha.
1060 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1061 false-negatives in the check that the sender had not preempted a response
1062 or prompt from Exim (running as a server), due to that code's lack of
1063 awareness of the SMTP input buffering.
1065 PP/04 Add commandline_checks_require_admin option.
1066 Exim drops privileges sanely, various checks such as -be aren't a
1067 security problem, as long as you trust local users with access to their
1068 own account. When invoked by services which pass untrusted data to
1069 Exim, this might be an issue. Set this option in main configuration
1070 AND make fixes to the calling application, such as using `--` to stop
1073 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1074 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1075 responses to those, into a single TLS record each way (this usually means
1076 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1079 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1080 affects you only if you're dancing at the edge of the param size limits.
1081 If you are, and this message makes sense to you, then: raise the
1082 configured limit or use OpenSSL 1.1. Nothing we can do for older
1085 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1086 cmdline element to get the documented default one. Previously it was
1089 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1090 are unsuable for non-message-specific reasons - eg. connection timeout,
1091 banner-time rejection.
1093 JH/16 Fix logging of delivery remote port, when specified by router, under
1096 PP/06 Repair manualroute's ability to take options in any order, even if one
1097 is the name of a transport.
1100 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1102 JH/17 Change the list-building routines interface to use the expanding-string
1103 triplet model, for better allocation and copying behaviour.
1105 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1106 Previously it was constructed the first time a possibly-matching string
1107 was met in the configuration file input during startup; now it is done
1110 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1111 compatible one, to avoid the (poorly documented) possibility of a config
1112 file in the working directory redirecting the DB files, possibly correpting
1113 some existing file. CVE-2017-10140 assigned for BDB.
1115 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1116 cache-hot. Previously, although the result was properly cached, the
1117 initial verify call returned a defer.
1119 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1120 the main verify for receipient in uncached-mode.
1122 JH/22 Retire historical build files to an "unsupported" subdir. These are
1123 defined as "ones for which we have no current evidence of testing".
1125 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1126 if present. Previously it was ignored.
1128 JH/24 Start using specified-initialisers in C structure init coding. This is
1129 a C99 feature (it's 2017, so now considered safe).
1131 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1132 if was a fixed-sized field and bitmask ops via macros; it is now more
1135 PP/07 GitHub PR 56: Apply MariaDB build fix.
1136 Patch provided by Jaroslav Škarvada.
1138 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1139 during Coverity cleanups [4.87 JH/47]
1140 Diagnosis and fix provided by Michael Fischer v. Mollard.
1142 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1143 the right size to place the terminating semicolon on its own folded
1144 line, the header hash was calculated to an incorrect value thanks to
1145 the (relaxed) space the fold became.
1147 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1148 and confused the parent.
1150 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1151 which could crash as a result. This could lead to undeliverable messages.
1153 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1156 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1157 matters on fast-turnover and PID-randomising systems, which were getting
1158 out-of-order delivery.
1160 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1161 a possibly-overlapping copy. The symptom was that "Remote host closed
1162 connection in response to HELO" was logged instead of the actual 4xx
1165 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1166 Previously only that bufferd was discarded, resulting in SYMTP command
1169 JH/32 DKIM: when a message has multiple signatures matching an identity given
1170 in dkim_verify_signers, run the dkim acl once for each. Previously only
1171 one run was done. Bug 2189.
1173 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1174 "panic the current process" to "deliberately defer". The panic log is
1175 still written with the problem list name; the mail and reject logs now
1176 get a temp-reject line for the message that was being handled, saying
1177 something like "domains check lookup or other defer". The SMTP 451
1178 message is still "Temporary local problem".
1180 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1181 A crafted sequence of BDAT commands could result in in-use memory beeing
1182 freed. CVE-2017-16943.
1184 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1185 from SMTP input. Previously it was always done; now only done for DATA
1186 and not BDAT commands. CVE-2017-16944.
1188 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1189 to the message (such as an overlong header line). Previously this was
1190 not done and we did not exit BDAT mode. Followon from the previous item
1191 though a different problem.
1197 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1198 than -2003 did; needs libidn2 in addition to libidn.
1200 JH/02 The path option on a pipe transport is now expanded before use.
1202 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1203 Patch provided by "Björn", documentation fix added too.
1205 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1206 missing a wire-to-host endian conversion.
1208 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1209 close after a BDAT command line could be taken as a following command,
1210 giving a synch failure. Fix by only checking for synch immediately
1211 before acknowledging the chunk.
1213 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1214 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1215 macro. Patches provided by Josh Soref.
1217 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1218 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1219 by IANA as a service extension. However, John Klensin suggests that we
1222 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1223 the dkim code may be unix-mode line endings rather than smtp wire-format
1224 CRLF, so prepend a CR to any bare LF.
1226 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1227 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1229 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1230 the first were themselves being wrongly included in the feed into dkim
1231 processing; with most chunk sizes in use this resulted in an incorrect
1232 body hash calculated value.
1234 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1235 DKIM signature block, for verification. Although advised against by
1236 standards it is specifically not ruled illegal.
1238 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1240 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1241 missing a body hash (the bh= tag).
1243 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1244 It seems that HAProxy sends the Proxy Protocol information in clear and
1245 only then does a TLS startup, so do the same.
1247 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1248 TCP connections (such as for Spamd) unless the daemon successfully set
1249 Fast Open mode on its listening sockets. This fixes breakage seen on
1250 too-old kernels or those not configured for Fast Open, at the cost of
1251 requiring both directions being enabled for TFO, and TFO never being used
1252 by non-daemon-related Exim processes.
1254 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1255 endings, at least on the first header line. Try to canonify any that get
1256 past that check, despite the cost.
1258 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1259 now limited to an arbitrary five deep, while parsing addresses with the
1260 strip_excess_angle_brackets option enabled.
1262 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1263 instead leave the unprompted TLS handshake in socket buffer for the
1264 TLS library to consume.
1266 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1268 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1270 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1271 region in one operation, for speed, and this leaves assigned pointers
1272 dangling. Add checks run only under the testsuite which checks all
1273 variables at a store-reset and panics on a dangling pointer; add code
1274 explicitly nulling out all the variables discovered. Fixes one known
1275 bug: a transport crash, where a dangling pointer for $sending_ip_address
1276 originally assigned in a verify callout, is re-used.
1278 PP/06 Drop '.' from @INC in various Perl scripts.
1280 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1282 PP/08 Reduce a number of compilation warnings under clang; building with
1283 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1284 should be warning-free.
1286 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1288 HS/01 Fix portability problems introduced by PP/08 for platforms where
1289 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1291 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1292 chunk. This allows us to accept broken chunked messages. We need a more
1293 general solution here.
1295 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1296 already-broken messages in the queue.
1298 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1300 JH/19 Fix reference counting bug in routing-generated-address tracking.
1306 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1307 supports it and a size is available (ie. the sending peer gave us one).
1309 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1310 years of being deprecated). The replacements are the ACLs
1311 acl_smtp_mime and acl_not_smtp_mime.
1313 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1314 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1315 or even an in-clear connection were permitted. Now, if the host lookup
1316 was dnssec and dane was requested then the host is only used if the
1317 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1318 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1319 if one fails this test.
1320 This means that a poorly-configured remote DNS will make it incommunicado;
1321 but it protects against a DNS-interception attack on it.
1323 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1324 non-noisy when a race steals the message being considered.
1326 JH/05 If main configuration option tls_certificate is unset, generate a
1327 self-signed certificate for inbound TLS connections.
1329 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1330 in rewrites and routers.
1332 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1333 and logged a warning sing 4.83; now they are a configuration file error.
1335 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1336 (lacking @domain). Apply the same qualification processing as RCPT.
1338 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1340 JH/10 Support ${sha256:} applied to a string (as well as the previous
1343 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1344 a cutthrough deliver is pending, as we always want to make a connection.
1345 This also avoids re-routing the message when later placing the cutthrough
1346 connection after a verify cache hit.
1347 Do not update it with the verify result either.
1349 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1350 when routing results in more than one destination address.
1352 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1353 signing (which inhibits the cutthrough capability). Previously only
1354 the presence of an option was tested; now an expansion evaluating as
1355 empty is permissible (obviously it should depend only on data available
1356 when the cutthrough connection is made).
1358 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1359 the relevant preceding SMTP command did not note the pipelining mode.
1361 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1362 Previously they were not counted.
1364 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1365 as one having no matching records. Previously we deferred the message
1366 that needed the lookup.
1368 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1369 distinguished as "(=".
1371 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1372 for missing MX records. Previously it only worked for missing A records.
1374 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1376 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1377 after the data-go-ahead and data-ack. Patch from Jason Betts.
1379 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1380 even for a "none" policy. Patch from Tony Meyer.
1382 JH/22 Fix continued use of a connection for further deliveries. If a port was
1383 specified by a router, it must also match for the delivery to be
1386 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1387 When one of the recipients of a message was unsuitable for the connection
1388 (has no matching addresses), we lost track of needing to mark it
1389 deferred. As a result mail would be lost.
1391 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1393 JH/25 Decoding ACL controls is now done using a binary search; the source code
1394 takes up less space and should be simpler to maintain. Merge the ACL
1395 condition decode tables also, with similar effect.
1397 JH/26 Fix problem with one_time used on a redirect router which returned the
1398 parent address unchanged. A retry would see the parent address marked as
1399 delivered, so not attempt the (identical) child. As a result mail would
1402 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1403 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1404 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1407 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1408 Main config options check_{log,spool}_{inodes,space} are now
1409 100 inodes, 10MB unless set otherwise in the configuration.
1411 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1412 Previously it only applied to the main-section connection policy
1415 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1417 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1418 by me. Added RFC7919 DH primes as an alternative.
1420 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1421 are not in the system include path.
1423 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1424 GnuTLS, when a session startup failed (eg because the client disconnected)
1425 Exim did stdio operations after fclose. This was exposed by a recent
1426 change which nulled out the file handle after the fclose.
1428 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1429 signed directly by the cert-signing cert, rather than an intermediate
1430 OCSP-signing cert. This is the model used by LetsEncrypt.
1432 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1434 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1435 an incoming connection.
1437 HS/02 Bug 1802: Do not half-close the connection after sending a request
1440 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1441 fallback to "prime256v1".
1443 JH/34 SECURITY: Use proper copy of DATA command in error message.
1444 Could leak key material. Remotely exploitable. CVE-2016-9963.
1450 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1451 and 3.4.4 - once the server is enabled to respond to an OCSP request
1452 it does even when not requested, resulting in a stapling non-aware
1453 client dropping the TLS connection.
1455 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1456 support variable-length bit vectors. No functional change.
1458 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1459 The I= interface field on outgoing lines is now after the H= remote
1460 host field, same as incoming lines. There is a separate
1461 outgoing_interface log selector which allows you to disable the
1464 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1465 If not running log_selector +smtp_connection the mainlog would be held
1466 open indefinitely after a "too many connections" event, including to a
1467 deleted file after a log rotate. Leave the per net connection logging
1468 leaving it open for efficiency as that will be quickly detected by the
1469 check on the next write.
1471 HS/01 Bug 1671: Fix post transport crash.
1472 Processing the wait-<transport> messages could crash the delivery
1473 process if the message IDs didn't exist for some reason. When
1474 using 'split_spool_directory=yes' the construction of the spool
1475 file name failed already, exposing the same netto behaviour.
1477 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1478 mime_regex ACL conditions.
1480 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1481 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1482 to HELO, local diagnostic string.
1484 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1485 log line to debug. Even when configured with a tls authenticator many
1486 client connections are expected to not authenticate in this way, so
1487 an authenticate fail is not an error.
1489 HS/02 Add the Exim version string to the process info. This way exiwhat
1490 gives some more detail about the running daemon.
1492 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1493 matter for fast-change records such as DNSBLs.
1495 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1496 retry record, even if constant. There may be multiple transports with
1497 different interface settings and the retry behaviour needs to be kept
1500 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1502 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1504 JH/10 Bug 840: fix log_defer_output option of pipe transport
1506 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1507 hosts_randomize. This matters a lot when combined with mua_wrapper.
1509 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1510 ${quote_pgsql:<string>} operator.
1512 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1514 JH/14 Change header file name for freeradius-client. Relevant if compiling
1515 with Radius support; from the Gentoo tree and checked under Fedora.
1517 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1519 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1520 option transport_filter, meaning no filtering.
1522 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1524 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1525 defaults to "*" (all hosts). The variable is now available when not built
1526 with TLS, default unset, mainly to enable keeping the testsuite sane.
1527 If a server certificate is not supplied (via tls_certificate) an error is
1528 logged, and clients will find TLS connections fail on startup. Presumably
1529 they will retry in-clear.
1530 Packagers of Exim are strongly encouraged to create a server certificate
1531 at installation time.
1533 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1534 with the $config_file variable.
1536 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1537 in transport context, after the attempt, and per-recipient. The latter type
1538 is per host attempted. The event data is the error message, and the errno
1539 information encodes the lookup type (A vs. MX) used for the (first) host,
1540 and the trailing two digits of the smtp 4xx response.
1542 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1543 to write to mainlog (or rejectlog, paniclog) in the window between file
1544 creation and permissions/ownership being changed. Particularly affects
1545 installations where exicyclog is run as root, rather than exim user;
1546 result is that the running daemon panics and dies.
1548 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1550 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1551 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1552 "pri" and "weight". Note that the previous implicit priority given by the
1553 list order is no longer honoured.
1555 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1556 for DKIM processing.
1558 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1559 by defining SUPPORT_SOCKS.
1561 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1562 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1563 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1564 variables are renamed to proxy_{local,external}_{address,port}.
1566 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1567 for a build by defining SUPPORT_I18N
1569 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1570 of the query string, and make ${quote_redis:} do that quoting.
1572 JH/29 Move Events support from Experimental to mainline, enabled by default
1573 and removable for a build by defining DISABLE_EVENT.
1575 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1577 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1578 cached by the daemon.
1580 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1581 by defining LOOKUP_REDIS. The libhiredis library is required.
1583 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1584 keys are given for lookup.
1586 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1587 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1588 only supported when built with TLS support. The PolarSSL SHA routines
1589 are still used when the TLS library is too old for convenient support.
1591 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1592 openssl_options), for security. OpenSSL forces this from version 1.1.0
1593 server-side so match that on older versions.
1595 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1596 allocation for $value could be released as the expansion processing
1597 concluded, but leaving the global pointer active for it.
1599 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1600 and to use the domains and local_parts ACL conditions.
1602 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1603 incorrectly not doubled on cutthrough transmission, hence seen as a
1604 body-termination at the receiving system - resulting in truncated mails.
1605 Commonly the sender saw a TCP-level error, and retransmitted the message
1606 via the normal store-and-forward channel. This could result in duplicates
1607 received - but deduplicating mailstores were liable to retain only the
1608 initial truncated version.
1610 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1612 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1614 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1615 we're in there, support oversigning also; bug 1309.
1617 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1619 HS/04 Add support for keep_environment and add_environment options.
1621 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1622 either intentional arithmetic overflow during PRNG, or testing config-
1625 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1626 delivery resulted in actual delivery. Cancel cutthrough before DATA
1629 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1630 rejects a recipient: pass the reject to the originator.
1632 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1633 Many were false-positives and ignorable, but it's worth fixing the
1636 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1637 for the new environment-manipulation done at startup. Move the routines
1638 from being local to tls.c to being global via the os.c file.
1640 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1641 an extract embedded as result-arg for a map, the first arg for extract
1642 is unavailable so we cannot tell if this is a numbered or keyed
1643 extraction. Accept either.
1649 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1652 JH/02 The smtp transport option "multi_domain" is now expanded.
1654 JH/03 The smtp transport now requests PRDR by default, if the server offers
1657 JH/04 Certificate name checking on server certificates, when exim is a client,
1658 is now done by default. The transport option tls_verify_cert_hostnames
1659 can be used to disable this per-host. The build option
1660 EXPERIMENTAL_CERTNAMES is withdrawn.
1662 JH/05 The value of the tls_verify_certificates smtp transport and main options
1663 default to the word "system" to access the system default CA bundle.
1664 For GnuTLS, only version 3.0.20 or later.
1666 JH/06 Verification of the server certificate for a TLS connection is now tried
1667 (but not required) by default. The verification status is now logged by
1668 default, for both outbound TLS and client-certificate supplying inbound
1671 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1674 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1675 Status Notification (bounce) messages are now MIME format per RFC 3464.
1676 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1677 under the control of the dsn_advertise_hosts option, and routers may
1678 have a dsn_lasthop option.
1680 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1681 default, modifiable by a malware= option. The list separator for
1682 the options can now be changed in the usual way. Bug 68.
1684 JH/10 The smtp_receive_timeout main option is now expanded before use.
1686 JH/11 The incoming_interface log option now also enables logging of the
1687 local interface on delivery outgoing connections.
1689 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1690 if the interface and destination host and port all match.
1692 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1695 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1696 Patch from Andrew Lewis.
1698 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1699 now supports optional time-restrictions, weighting, and priority
1700 modifiers per server. Patch originally by <rommer@active.by>.
1702 JH/16 The spamd_address main option now supports a mixed list of local
1703 and remote servers. Remote servers can be IPv6 addresses, and
1704 specify a port-range.
1706 JH/17 Bug 68: The spamd_address main option now supports an optional
1707 timeout value per server.
1709 JH/18 Bug 1581: Router and transport options headers_add/remove can
1710 now have the list separator specified.
1712 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1715 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1718 JH/21 Support for the A6 type of dns record is withdrawn.
1720 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1721 rather than the verbs used.
1723 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1724 from 255 to 1024 chars.
1726 JH/24 Verification callouts now attempt to use TLS by default.
1728 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1729 are generic router options now. The defaults didn't change.
1731 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1732 Original patch from Alexander Shikoff, worked over by JH.
1734 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1735 files if ZCAT_COMMAND is not executable.
1737 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1739 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1741 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1742 Normally benign, it bites when the pair was led to by a CNAME;
1743 modern usage is to not canonicalize the domain to a CNAME target
1744 (and we were inconsistent anyway for A-only vs AAAA+A).
1746 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1748 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1749 when evaluating $sender_host_dnssec.
1751 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1752 $sender_helo_dnssec variable.
1754 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1756 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1758 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1760 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1761 documented as working, but never had. Support all but $spam_report.
1763 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1764 added for tls authenticator.
1766 HS/03 Add perl_taintmode main config option
1772 TL/01 When running the test suite, the README says that variables such as
1773 no_msglog_check are global and can be placed anywhere in a specific
1774 test's script, however it was observed that placement needed to be near
1775 the beginning for it to behave that way. Changed the runtest perl
1776 script to read through the entire script once to detect and set these
1777 variables, reset to the beginning of the script, and then run through
1778 the script parsing/test process like normal.
1780 TL/02 The BSD's have an arc4random API. One of the functions to induce
1781 adding randomness was arc4random_stir(), but it has been removed in
1782 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1783 function when detected.
1785 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1786 cause callback expansion.
1788 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1789 syntax errors in an expansion can be treated as a string instead of
1790 logging or causing an error, due to the internal use of bool_lax
1791 instead of bool when processing it.
1793 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1794 server certificates when making smtp deliveries.
1796 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1798 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1800 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1802 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1803 Merged patch from Sebastian Wiedenroth.
1805 JH/05 Fix results-pipe from transport process. Several recipients, combined
1806 with certificate use, exposed issues where response data items split
1807 over buffer boundaries were not parsed properly. This eventually
1808 resulted in duplicates being sent. This issue only became common enough
1809 to notice due to the introduction of connection certificate information,
1810 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1812 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1813 size buffer was used, resulting in syntax errors when an expansion
1816 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1817 version 3.3.6 or later.
1819 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1820 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1821 both become "event_action", the variables become $event_name, $event_data
1822 and $event_defer_errno. There is a new variable $verify_mode, usable in
1823 routers, transports and related events. The tls:cert event is now also
1824 raised for inbound connections, if the main configuration event_action
1827 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1828 early OCSP support, but no stapling (appears to be less than 1.0.0).
1830 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1831 server certificate names available under the smtp transport option
1832 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1835 JH/10 Time-related extraction expansions from certificates now use the main
1836 option "timezone" setting for output formatting, and are consistent
1837 between OpenSSL and GnuTLS compilations. Bug 1541.
1839 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1840 encoded parameter in the incoming message. Bug 1558.
1842 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1843 include certificate info, eximon was claiming there were spoolfile
1846 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1848 JH/14 Log delivery-related information more consistently, using the sequence
1849 "H=<name> [<ip>]" wherever possible.
1851 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1852 are problematic for Debian distribution, omit them from the release
1855 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1857 JH/16 Fix string representation of time values on 64bit time_t architectures.
1860 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1861 output list separator was used.
1866 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1867 checkers that were complaining about end of non-void function with no
1870 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1871 This was a regression introduced in 4.83 by another bugfix.
1873 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1875 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1876 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1882 TF/01 Correctly close the server side of TLS when forking for delivery.
1884 When a message was received over SMTP with TLS, Exim failed to clear up
1885 the incoming connection properly after forking off the child process to
1886 deliver the message. In some situations the subsequent outgoing
1887 delivery connection happened to have the same fd number as the incoming
1888 connection previously had. Exim would try to use TLS and fail, logging
1889 a "Bad file descriptor" error.
1891 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1892 utilities have not been installed.
1894 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1895 temporary space as the ACL may create new global variables.
1897 TL/01 LDAP support uses per connection or global context settings, depending
1898 upon the detected version of the libraries at build time.
1900 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1901 to extract and use the src ip:port in logging and expansions as if it
1902 were a direct connection from the outside internet. PPv2 support was
1903 updated based on HAProxy spec change in May 2014.
1905 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1907 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1908 Properly escape header and check for NULL return.
1910 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1913 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1915 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1916 characters in header names, implemented as a verify condition.
1917 Contributed by Michael Fischer v. Mollard.
1919 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1920 results permerror and temperror. Previous values are deprecated but
1921 still accepted. In a future release, err_perm and err_temp will be
1922 completely removed, which will be a backward incompatibility if the
1923 ACL tests for either of these two old results. Patch contributed by
1924 user bes-internal on the mailing list.
1926 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1928 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1929 selectors, in both main and reject logs.
1931 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1934 JH/07 Add malware type "sock" for talking to simple daemon.
1936 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1938 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1939 routers/transports under cutthrough routing.
1941 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1942 numbers. Touch up "bool" conditional to keep the same definition.
1944 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1946 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1948 JH/12 Expand items in router/transport headers_add or headers_remove lists
1949 individually rather than the list as a whole. Bug 1452.
1951 Required for reasonable handling of multiple headers_ options when
1952 they may be empty; requires that headers_remove items with embedded
1953 colons must have them doubled (or the list-separator changed).
1955 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1956 view the policy declared in the DMARC record. Currently, $dmarc_status
1957 is a combined value of both the record presence and the result of the
1960 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1962 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1963 dnslookup router and the smtp transport (applying to the forward
1966 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1967 of ldap servers used for a specific lookup. Patch provided by Heiko
1970 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1971 New variable $lookup_dnssec_authenticated for observability.
1973 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1974 Patch submitted by Lars Timman.
1976 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1978 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1979 Requires trusted mode and valid format message id, aborts otherwise.
1980 Patch contributed by Heiko Schlichting.
1982 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1983 certextract with support for various fields. Bug 1358.
1985 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1986 is requested by default, modifiable by smtp transport option
1989 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1990 operate on certificate variables to give certificate fingerprints
1991 Also new ${sha256:cert_variable}.
1993 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1995 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1996 Christian Aistleitner.
1998 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2000 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2001 file. Patch from Wolfgang Breyha.
2003 JH/25 Expand the coverage of the delivery $host and $host_address to
2004 client authenticators run in verify callout. Bug 1476.
2006 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2007 align with daemon_smtp_ports. Bug 72.
2009 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2010 support and error reporting did not work properly.
2012 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2013 and is readable. Patch from Andrew Colin Kissa.
2015 TL/14 Enhance documentation of ${run expansion and how it parses the
2016 commandline after expansion, particularly in the case when an
2017 unquoted variable expansion results in an empty value.
2019 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2021 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2022 by renaming away from it. Addresses GH issue 15, reported by
2025 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2027 TL/15 SECURITY: prevent double expansion in math comparison functions
2028 (can expand unsanitized data). Not remotely exploitable.
2035 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2037 PP/02 Make -n do something, by making it not do something.
2038 When combined with -bP, the name of an option is not output.
2040 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2043 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2044 $sender_host_name and config options to manage this, and basic check
2047 PP/05 DSCP support for outbound connections and control modifier for inbound.
2049 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2050 (Only plugin which currently uses this is kerberos4, which nobody should
2051 be using, but we should make it available and other future plugins might
2052 conceivably use it, even though it would break NAT; stuff *should* be
2053 using channel bindings instead).
2055 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2056 name; added for Sendmail compatibility; requires admin caller.
2057 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2058 just ignore it); requires trusted caller.
2059 Also parse but ignore: -Ac -Am -X<logfile>
2062 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2064 TL/02 Add +smtp_confirmation as a default logging option.
2066 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2067 Patch by Magnus Holmgren from 2007-02-20.
2069 TL/04 Bugzilla 1281 - Spec typo.
2070 Bugzilla 1283 - Spec typo.
2071 Bugzilla 1290 - Spec grammar fixes.
2073 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2075 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2077 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2078 mailing list by by Dmitry Isaikin.
2080 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2082 JH/02 Support "G" suffix to numbers in ${if comparisons.
2084 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2086 NM/01 Bugzilla 1197 - Spec typo
2087 Bugzilla 1196 - Spec examples corrections
2089 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2091 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2092 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2095 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2096 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2098 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2099 "acl {{name}{arg}...}", and optional args on acl condition
2102 JH/05 Permit multiple router/transport headers_add/remove lines.
2104 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2106 JH/07 Avoid using a waiting database for a single-message-only transport.
2107 Performance patch from Paul Fisher. Bugzilla 1262.
2109 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2112 JH/09 Add $headers_added variable, with content from use of ACL modifier
2113 add_header (but not yet added to the message). Bugzilla 199.
2115 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2116 Pulled from Bugzilla 817 by Wolfgang Breyha.
2118 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2120 (nb: this is the same fix as in Exim 4.80.1)
2122 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2125 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2128 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2130 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2131 advertise SMTP AUTH mechanism to us, instead of a generic
2132 protocol violation error. Also, make Exim more robust to bad
2133 data from the Dovecot auth socket.
2135 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2137 When a queue runner is handling a message, Exim first routes the
2138 recipient addresses, during which it prunes them based on the retry
2139 hints database. After that it attempts to deliver the message to
2140 any remaining recipients. It then updates the hints database using
2143 So if a recipient address works intermittently, it can get repeatedly
2144 deferred at routing time. The retry hints record remains fresh so the
2145 address never reaches the final cutoff time.
2147 This is a fairly common occurrence when a user is bumping up against
2148 their storage quota. Exim had some logic in its local delivery code
2149 to deal with this. However it did not apply to per-recipient defers
2150 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2152 This change adds a proper retry rule check during routing so that the
2153 final cutoff time is checked against the message's age. We only do
2154 this check if there is an address retry record and there is not a
2155 domain retry record; this implies that previous attempts to handle
2156 the address had the retry_use_local_parts option turned on. We use
2157 this as an approximation for the destination being like a local
2158 delivery, as in LMTP.
2160 I suspect this new check makes the old local delivery cutoff check
2161 redundant, but I have not verified this so I left the code in place.
2163 TF/02 Correct gecos expansion when From: is a prefix of the username.
2165 Test 0254 submits a message to Exim with the header
2169 When I ran the test suite under the user fanf2, Exim expanded
2170 the header to contain my full name, whereas it should have added
2171 a Resent-Sender: header. It erroneously treats any prefix of the
2172 username as equal to the username.
2174 This change corrects that bug.
2176 GF/01 DCC debug and logging tidyup
2177 Error conditions log to paniclog rather than rejectlog.
2178 Debug lines prefixed by "DCC: " to remove any ambiguity.
2180 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2182 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2183 Bug spotted by Jeremy Harris; was flawed since initial commit.
2184 Would have resulted in OCSP responses post-SNI triggering an Exim
2185 NULL dereference and crash.
2187 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2189 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2190 Bug detection, analysis and fix by Samuel Thibault.
2191 Bugzilla 1331, Debian bug #698092.
2193 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2195 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2196 Server implementation by Todd Lyons, client by JH.
2197 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2198 config variable "prdr_enable" controls whether the server
2199 advertises the facility. If the client requests PRDR a new
2200 acl_data_smtp_prdr ACL is called once for each recipient, after
2201 the body content is received and before the acl_smtp_data ACL.
2202 The client is controlled by both of: a hosts_try_prdr option
2203 on the smtp transport, and the server advertisement.
2204 Default client logging of deliveries and rejections involving
2205 PRDR are flagged with the string "PRDR".
2207 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2208 fclose(). Diagnosis by Todd Lyons.
2210 PP/17 Update configure.default to handle IPv6 localhost better.
2211 Patch by Alain Williams (plus minor tweaks).
2214 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2215 This is now consistent with GnuTLS, and is now documented: the
2216 previous undocumented portable approach to treating the option as
2217 unset was to force an expansion failure. That still works, and
2218 an empty string is now equivalent.
2220 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2221 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2222 not performing validation itself.
2224 PP/20 Added force_command boolean option to pipe transport.
2225 Patch from Nick Koston, of cPanel Inc.
2227 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2230 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2232 PP/21 Fix eximon continuous updating with timestamped log-files.
2233 Broken in a format-string cleanup in 4.80, missed when I repaired the
2234 other false fix of the same issue.
2235 Report and fix from Heiko Schlichting.
2238 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2239 Report from Prashanth Katuri.
2241 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2242 It's SecureTransport, so affects any MacOS clients which use the
2243 system-integrated TLS libraries, including email clients.
2245 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2246 using a MIME ACL for non-SMTP local injection.
2247 Report and assistance in diagnosis by Warren Baker.
2249 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2251 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2253 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2254 last id that failed so it may be referenced in subsequent ACL's.
2256 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2259 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2260 ldap library initialization, allowing self-signed CA's to be
2261 used. Also properly sets require_cert option later in code by
2262 using NULL (global ldap config) instead of ldap handle (per
2263 session). Bug diagnosis and testing by alxgomz.
2265 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2266 the src/util/ subdirectory.
2268 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2269 renamed to Transport Post Delivery Action by Jeremy Harris, as
2272 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2273 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2274 redis_servers = needs to be configured which will be used by the redis
2275 lookup. Patch from Warren Baker, of The Packet Hub.
2277 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2279 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2280 hostname or reverse DNS when processing a host list. Used suggestions
2281 from multiple comments on this bug.
2283 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2285 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2286 Missed a few lines, added it to make the runtest require no keyboard
2289 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2290 contains upper case chars. Make router use caseful_local_part.
2292 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2293 support when GnuTLS has been built with p11-kit.
2299 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2301 This, or similar/improved, will also be change PP/11 of 4.82.
2307 PP/01 Handle short writes when writing local log-files.
2308 In practice, only affects FreeBSD (8 onwards).
2309 Bugzilla 1053, with thanks to Dmitry Isaikin.
2311 NM/01 Bugzilla 949 - Documentation tweak
2313 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2316 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2318 PP/02 Implemented gsasl authenticator.
2320 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2322 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2323 `pkg-config foo` for cflags/libs.
2325 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2326 with rest of GSASL and with heimdal_gssapi.
2328 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2329 `pkg-config foo` for cflags/libs for the TLS implementation.
2331 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2332 properties get this fed in as external SSF. A number of robustness
2333 and debugging improvements to the cyrus_sasl authenticator.
2335 PP/08 cyrus_sasl server now expands the server_realm option.
2337 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2338 Patch by Jeremy Harris.
2340 PP/10 Added dbmjz lookup type.
2342 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2344 PP/12 MAIL args handles TAB as well as SP, for better interop with
2345 non-compliant senders.
2346 Analysis and variant patch by Todd Lyons.
2348 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2349 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2350 Patch from Dirk Mueller <dmueller@suse.com>
2352 PP/13 tls_peerdn now print-escaped for spool files.
2353 Observed some $tls_peerdn in wild which contained \n, which resulted
2354 in spool file corruption.
2356 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2357 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2358 or write after TLS renegotiation, which otherwise led to messages
2361 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2362 as a tracking header (ie: a signed header comes before the signature).
2363 Patch from Wolfgang Breyha.
2365 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2366 comma-sep list; embedded commas doubled.
2368 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2370 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2372 Report and patch from Dmitry Banschikov.
2374 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2375 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2376 failures. We appear to now support TLS1.1+ with Exim.
2378 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2379 lets Exim select keys and certificates based upon TLS SNI from client.
2380 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2381 before an outbound SMTP session. New log_selector, +tls_sni.
2383 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2384 NULL dereference. Report and patch from Alun Jones.
2386 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2387 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2388 Not seeing resolver debug output on NetBSD, but suspect this is a
2389 resolver implementation change.
2391 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2392 Left warnings. Added "eximon gdb" invocation mode.
2394 PP/21 Defaulting "accept_8bitmime" to true, not false.
2396 PP/22 Added -bw for inetd wait mode support.
2398 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2399 locate the relevant includes and libraries. Made this the default.
2401 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2402 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2404 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2405 This may cause build issues on older platforms.
2407 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2408 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2409 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2410 Added SNI support via GnuTLS too.
2411 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2413 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2415 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2416 Applied second patch from Janne, implementing suggestion to default
2417 multiple-strings-in-record handling to match SPF spec.
2419 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2421 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2422 read-only, out of scope).
2423 Patch from Wolfgang Breyha, report from Stuart Northfield.
2425 PP/29 Fix three issues highlighted by clang analyser static analysis.
2426 Only crash-plausible issue would require the Cambridge-specific
2427 iplookup router and a misconfiguration.
2428 Report from Marcin Mirosław.
2430 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2432 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2433 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2434 As part of this, removing so much warning spew let me fix some minor
2435 real issues in debug logging.
2437 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2438 assignment on my part. Fixed.
2440 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2441 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2442 Janne Snabb (who went above and beyond: thank you).
2444 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2445 string otherwise requires a connection and a bunch more work and it's
2446 relatively easy to get wrong. Should also expose TLS library linkage
2449 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2450 64-bit ${eval} (JH/03).
2452 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2453 GNU libc to support some of the 64-bit stuff, should not lead to
2454 conflicts. Defined before os.h is pulled in, so if a given platform
2455 needs to override this, it can.
2457 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2458 protection layer was required, which is not implemented.
2459 Bugzilla 1254, patch from Wolfgang Breyha.
2461 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2462 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2463 tls_dhparam take prime identifiers. Also unbreak combination of
2464 OpenSSL+DH_params+TLSSNI.
2466 PP/39 Disable SSLv2 by default in OpenSSL support.
2472 PP/01 Solaris build fix for Oracle's LDAP libraries.
2473 Bugzilla 1109, patch from Stephen Usher.
2475 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2477 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2480 TF/02 Fix a couple more cases where we did not log the error message
2481 when unlink() failed. See also change 4.74-TF/03.
2483 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2484 lock up or crash if it happened to be inside a call to libc when it
2485 got a SIGUSR1 from exiwhat.
2487 The SIGUSR1 handler appends the current process status to the process
2488 log which is later printed by exiwhat. It used to use the general
2489 purpose logging code to do this, but several functions it calls are
2490 not safe for signals.
2492 The new output code in the SIGUSR1 handler is specific to the process
2493 log, and simple enough that it's easy to inspect for signal safety.
2494 Removing some special cases also simplifies the general logging code.
2495 Removing the spurious timestamps from the process log simplifies
2498 TF/04 Improved ratelimit ACL condition.
2500 The /noupdate option has been deprecated in favour of /readonly which
2501 has clearer semantics. The /leaky, /strict, and /readonly update modes
2502 are mutually exclusive. The update mode is no longer included in the
2503 database key; it just determines when the database is updated. (This
2504 means that when you upgrade Exim will forget old rate measurements.)
2506 Exim now checks that the per_* options are used with an update mode that
2507 makes sense for the current ACL. For example, when Exim is processing a
2508 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2509 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2510 must specify per_mail/readonly. If you omit the update mode it defaults to
2511 /leaky where that makes sense (as before) or /readonly where required.
2513 The /noupdate option is now undocumented but still supported for
2514 backwards compatibility. It is equivalent to /readonly except that in
2515 ACLs where /readonly is required you may specify /leaky/noupdate or
2516 /strict/noupdate which are treated the same as /readonly.
2518 A useful new feature is the /count= option. This is a generalization
2519 of the per_byte option, so that you can measure the throughput of other
2520 aggregate values. For example, the per_byte option is now equivalent
2521 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2523 The per_rcpt option has been generalized using the /count= mechanism
2524 (though it's more complicated than the per_byte equivalence). When it is
2525 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2526 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2527 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2528 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2529 Note that using per_rcpt with a non-readonly update mode in more than
2530 one ACL will cause the recipients to be double-counted. (The per_mail
2531 and per_byte options don't have this problem.)
2533 The handling of very low rates has changed slightly. If the computed rate
2534 is less than the event's count (usually one) then this event is the first
2535 after a long gap. In this case the rate is set to the same as this event's
2536 count, so that the first message of a spam run is counted properly.
2538 The major new feature is a mechanism for counting the rate of unique
2539 events. The new per_addr option counts the number of different
2540 recipients that someone has sent messages to in the last time period. It
2541 behaves like per_rcpt if all the recipient addresses are different, but
2542 duplicate recipient addresses do not increase the measured rate. Like
2543 the /count= option this is a general mechanism, so the per_addr option
2544 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2545 example, measure the rate that a client uses different sender addresses
2546 with the options per_mail/unique=$sender_address. There are further
2547 details in the main documentation.
2549 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2551 TF/06 Removed a few PCRE remnants.
2553 TF/07 Automatically extract Exim's version number from tags in the git
2554 repository when doing development or release builds.
2556 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2557 Bugzilla 879. Patch from Paul Fisher.
2559 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2560 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2563 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2565 PP/05 Variable $av_failed, true if the AV scanner deferred.
2566 Bugzilla 1078. Patch from John Horne.
2568 PP/06 Stop make process more reliably on build failure.
2569 Bugzilla 1087. Patch from Heiko Schlittermann.
2571 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2572 Bugzilla 1089. Patch from Heiko Schlittermann.
2574 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2575 Bugzilla 1131. Patch from Holger Weiß.
2577 PP/09 Handle IPv6 addresses with SPF.
2578 Bugzilla 860. Patch from Wolfgang Breyha.
2580 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2582 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2585 PP/11 match_* no longer expand right-hand-side by default.
2586 New compile-time build option, EXPAND_LISTMATCH_RHS.
2587 New expansion conditions, "inlist", "inlisti".
2589 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2591 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2593 PP/14 fix log_write() format string regression from TF/03.
2594 Bugzilla 1152. Patch from Dmitry Isaikin.
2600 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2602 PP/02 Harmonised TLS library version reporting; only show if debugging.
2603 Layout now matches that introduced for other libraries in 4.74 PP/03.
2605 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2607 PP/04 New "dns_use_edns0" global option.
2609 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2612 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2613 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2615 TK/01 Updated PolarSSL code to 0.14.2.
2616 Bugzilla 1097. Patch from Andreas Metzler.
2618 PP/07 Catch divide-by-zero in ${eval:...}.
2619 Fixes bugzilla 1102.
2621 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2624 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2625 format-string attack -- SECURITY: remote arbitrary code execution.
2627 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2628 time unintentionally subject to list matching rules, letting the header
2629 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2630 arbitrary string expansion). This allowed for information disclosure.
2632 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2633 INT_MIN/-1 -- value coerced to INT_MAX.
2639 NM/01 Workaround for PCRE version dependency in version reporting
2642 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2643 This fixes portability to compilers other than gcc, notably
2644 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2646 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2647 makefiles for portability to HP-UX and POSIX correctness.
2649 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2650 Also via indented variable definition in the Makefile.
2651 (Debugging by Oliver Heesakkers).
2653 PP/02 Restore caching of spamd results with expanded spamd_address.
2654 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2656 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2657 Improves build reliability. Fix from: Frank Elsner
2659 NM/02 Fix wide character breakage in the rfc2047 coding
2660 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2662 NM/03 Allow underscore in dnslist lookups
2663 Fixes bug 1026. Patch from Graeme Fowler
2665 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2666 Code patches from Adam Ciarcinski of NetBSD.
2668 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2671 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2672 is logged, to avoid truncation. Patch from John Horne.
2674 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2675 Patch from Jakob Hirsch.
2677 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2678 SQL string expansion failure details.
2679 Patch from Andrey Oktyabrski.
2681 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2682 Patch from Simon Arlott.
2684 PP/09 New lookups functionality failed to compile on old gcc which rejects
2685 extern declarations in function scope.
2686 Patch from Oliver Fleischmann
2688 PP/10 Use sig_atomic_t for flags set from signal handlers.
2689 Check getgroups() return and improve debugging.
2690 Fixed developed for diagnosis in bug 927 (which turned out to be
2693 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2694 Patch from Mark Zealey.
2696 PP/12 Bugzilla 1056: Improved spamd server selection.
2697 Patch from Mark Zealey.
2699 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2700 Based on patch from Heiko Schlittermann.
2702 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2703 Patch from Uwe Doering, sign-off by Michael Haardt.
2705 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2706 variable declaration deep within a block. Bug and patch from
2709 PP/15 lookups-Makefile IRIX compatibility coercion.
2711 PP/16 Make DISABLE_DKIM build knob functional.
2713 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2714 Patch by Simon Arlott
2716 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2717 variable argument macros. Our copy now differs from upstream.
2723 TF/01 Failure to get a lock on a hints database can have serious
2724 consequences so log it to the panic log.
2726 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2727 controlled using the smtp_confirmation log selector.
2729 TF/03 Include the error message when we fail to unlink a spool file.
2731 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2732 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2733 for maintaining out-of-tree patches for some time.
2735 PP/01 Bugzilla 139: Documentation and portability issues.
2736 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2737 Handle per-OS dynamic-module compilation flags.
2739 PP/02 Let /dev/null have normal permissions.
2740 The 4.73 fixes were a little too stringent and complained about the
2741 permissions on /dev/null. Exempt it from some checks.
2742 Reported by Andreas M. Kirchwitz.
2744 PP/03 Report version information for many libraries, including
2745 Exim version information for dynamically loaded libraries. Created
2746 version.h, now support a version extension string for distributors
2747 who patch heavily. Dynamic module ABI change.
2749 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2750 privilege escalation vulnerability whereby the Exim run-time user
2751 can cause root to append content of the attacker's choosing to
2754 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2757 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2758 If dropping privileges for untrusted macros, we disabled normal logging
2759 on the basis that it would fail; for the Exim run-time user, this is not
2760 the case, and it resulted in successful deliveries going unlogged.
2761 Fixed. Reported by Andreas Metzler.
2767 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2768 only prepend for the Resent-* case. Fixes regression introduced in
2769 Exim 4.70 by NM/22 for Bugzilla 607.
2771 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2772 increasing numbers of administrators be bitten by this.
2774 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2776 PP/03 Bugzilla 994: added openssl_options main configuration option.
2778 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2780 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2782 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2784 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2785 without a peer certificate, leading to a segfault because of an
2786 assumption that peers always have certificates. Be a little more
2787 paranoid. Problem reported by Martin Tscholak.
2789 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2790 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2791 NB: ClamAV planning to remove STREAM in "middle of 2010".
2792 CL also introduces -bmalware, various -d+acl logging additions and
2793 more caution in buffer sizes.
2795 PP/09 Implemented reverse_ip expansion operator.
2797 PP/10 Bugzilla 937: provide a "debug" ACL control.
2799 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2801 PP/12 Bugzilla 973: Implement --version.
2803 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2805 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2807 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2809 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2810 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2811 ignore trailing whitespace.
2813 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2815 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2818 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2819 Notification from Dr Andrew Aitchison.
2821 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2822 ExtendedDetectionInfo response format.
2823 Notification from John Horne.
2825 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2828 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2829 XSL and documented dependency on system catalogs, with examples of how
2832 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2835 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2836 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2837 configuration file which is writeable by the Exim user or group.
2839 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2840 of configuration files to cover files specified with the -C option if
2841 they are going to be used with root privileges, not just the default
2844 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2845 option (effectively making it always true).
2847 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2848 files to be used while preserving root privileges.
2850 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2851 that rogue child processes cannot use them.
2853 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2854 run-time user, instead of root.
2856 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2857 Exim run-time user without dropping privileges.
2859 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2860 result string, instead of calling string_vformat() twice with the same
2863 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2864 for other users. Others should always drop root privileges if they use
2865 -C on the command line, even for a whitelisted configure file.
2867 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2869 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2875 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2876 $data_path, and $header_path variables; fixed documentation bugs and
2879 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2880 exipick to access non-standard spools, including the "frozen" queue
2883 NM/01 Bugzilla 965: Support mysql stored procedures.
2884 Patch from Alain Williams
2886 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2888 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2889 Patch from Andreas Metzler
2891 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2892 Patch from Kirill Miazine
2894 NM/05 Bugzilla 671: Added umask to procmail example.
2896 JJ/03 installed exipick 20100323.0, fixing doc bug
2898 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2899 directory. Notification and patch from Dan Rosenberg.
2901 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2903 TK/02 Improve log output when DKIM signing operation fails.
2905 MH/01 Treat the transport option dkim_domain as a colon separated
2906 list, not as a single string, and sign the message with each element,
2907 omitting multiple occurences of the same signer.
2909 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2910 Bugzilla 985, 986. Patch by Simon Arlott
2912 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2913 Patch by Simon Arlott
2915 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2916 MBX locking. Notification from Dan Rosenberg.
2922 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2924 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2926 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2928 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2930 NM/04 Bugzilla 915: Flag broken perl installation during build.
2936 TK/01 Added patch by Johannes Berg that expands the main option
2937 "spamd_address" if it starts with a dollar sign.
2939 TK/02 Write list of recipients to X-Envelope-Sender header when building
2940 the mbox-format spool file for content scanning (suggested by Jakob
2943 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2944 (http://www.dcc-servers.net/) support via dccifd. Activated by
2945 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2947 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2948 by Mark Daniel Reidel <mr@df.eu>.
2950 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2951 When building exim an external PCRE library is now needed -
2952 PCRE is a system library on the majority of modern systems.
2953 See entry on PCRE_LIBS in EDITME file.
2955 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2956 conversation. Added nologin parameter to request.
2957 Patch contributed by Kirill Miazine.
2959 TF/01 Do not log submission mode rewrites if they do not change the address.
2961 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2963 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2964 log files in place. Contributed by Roberto Lima.
2966 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2968 TF/03 Bugzilla 615: When checking the local_parts router precondition
2969 after a local_part_suffix or local_part_prefix option, Exim now
2970 does not use the address's named list lookup cache, since this
2971 contains cached lookups for the whole local part.
2973 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2974 Robert Millan. Documentation is in experimental-spec.txt.
2976 TF/04 Bugzilla 668: Fix parallel build (make -j).
2978 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2980 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2981 Patch provided by Jan Srzednicki.
2983 TF/05 Leading white space used to be stripped from $spam_report which
2984 wrecked the formatting. Now it is preserved.
2986 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2987 that they are available at delivery time.
2989 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2991 TF/08 TLS error reporting now respects the incoming_interface and
2992 incoming_port log selectors.
2994 TF/09 Produce a more useful error message if an SMTP transport's hosts
2995 setting expands to an empty string.
2997 NM/06 Bugzilla 744: EXPN did not work under TLS.
2998 Patch provided by Phil Pennock.
3000 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3001 Patch provided by Richard Godbee.
3003 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3004 acl_smtp_notquit, added index entry.
3006 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3007 Patch provided by Eugene Bujak.
3009 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3010 accept(). Patch provided by Maxim Dounin.
3012 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3013 Patch provided by Phil Pennock.
3015 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3017 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3018 Patch provided by Brad "anomie" Jorsch.
3020 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3021 Patch provided by Dean Brooks.
3023 TK/05 Add native DKIM support (does not depend on external libraries).
3025 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3026 Patch provided by Graeme Fowler.
3028 NM/16 Bugzilla 851: Documentation example syntax fix.
3030 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3032 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3035 NM/19 Bugzilla 745: TLS version reporting.
3036 Patch provided by Phil Pennock.
3038 NM/20 Bugzilla 167: bool: condition support.
3039 Patch provided by Phil Pennock.
3041 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3042 clients. Patch provided by Phil Pennock.
3044 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3045 Patch provided by Brad "anomie" Jorsch.
3047 NM/23 Bugzilla 687: Fix misparses in eximstats.
3048 Patch provided by Heiko Schlittermann.
3050 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3051 Patch provided by Heiko Schlittermann.
3053 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3054 plus update to original patch.
3056 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3058 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3059 Patch provided by David Brownlee.
3061 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3063 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3065 NM/30 Bugzilla 888: TLS documentation bugfixes.
3067 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3069 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3070 Unlike the original bugzilla I have changed all shell scripts in src tree.
3072 NM/33 Bugzilla 898: Transport filter timeout fix.
3073 Patch by Todd Rinaldo.
3075 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3076 Patch by Serge Demonchaux.
3078 NM/35 Bugzilla 39: Base64 decode bug fixes.
3079 Patch by Jakob Hirsch.
3081 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3083 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3085 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3087 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3093 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3094 ALT-N's libdkim that I have put here:
3095 http://duncanthrax.net/exim-experimental/
3097 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3098 were called 'true' and it seems that C99 defines that as a reserved
3099 keyword to be used with 'bool' variable types. That means you could
3100 not include C99-style headers which use bools without triggering
3101 build errors in sieve.c.
3103 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3104 as mailq or other aliases. Changed the --help handling significantly
3105 to do whats expected. exim_usage() emits usage/help information.
3107 SC/01 Added the -bylocaldomain option to eximstats.
3109 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3111 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3113 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3119 PH/01 Another patch from the Sieve maintainer.
3121 PH/02 When an IPv6 address is converted to a string for single-key lookup
3122 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3123 dots are used instead of colons so that keys in lsearch files need not
3124 contain colons. This was done some time before quoting was made available
3125 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3126 (notated using the quote facility) so as to distinguish them from IPv4
3127 keys. This meant that lookups for IP addresses in host lists did not work
3128 for iplsearch lookups.
3130 This has been fixed by arranging for IPv6 addresses to be expressed with
3131 colons if the lookup type is iplsearch. This is not incompatible, because
3132 previously such lookups could never work.
3134 The situation is now rather anomalous, since one *can* have colons in
3135 ordinary lsearch keys. However, making the change in all cases is
3136 incompatible and would probably break a number of configurations.
3138 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3141 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3142 conversion specification without a maximum field width, thereby enabling
3143 a rogue spamd server to cause a buffer overflow. While nobody in their
3144 right mind would setup Exim to query an untrusted spamd server, an
3145 attacker that gains access to a server running spamd could potentially
3146 exploit this vulnerability to run arbitrary code as the Exim user.
3148 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3149 $primary_hostname instead of what libspf2 thinks the hosts name is.
3151 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3152 a directory entry by the name of the lookup key. Previously, if a
3153 symlink pointed to a non-existing file or a file in a directory that
3154 Exim lacked permissions to read, a lookup for a key matching that
3155 symlink would fail. Now it is enough that a matching directory entry
3156 exists, symlink or not. (Bugzilla 503.)
3158 PH/03 The body_linecount and body_zerocount variables are now exported in the
3161 PH/04 Added the $dnslist_matched variable.
3163 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3164 This means they are set thereafter only if the connection becomes
3167 PH/06 Added the client_condition to authenticators so that some can be skipped
3168 by clients under certain conditions.
3170 PH/07 The error message for a badly-placed control=no_multiline_responses left
3171 "_responses" off the end of the name.
3173 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3175 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3176 (without spaces) instead of just copying the configuration text.
3178 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3180 PH/11 Added $max_received_linelength.
3182 PH/12 Added +ignore_defer and +include_defer to host lists.
3184 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3185 way in which PCRE > 7.0 is built.
3187 PH/14 Implemented queue_only_load_latch.
3189 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3190 MAIL command. The effect was to mangle the value on 64-bit systems.
3192 PH/16 Another patch from the Sieve maintainer.
3194 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3196 PH/18 If a system quota error occurred while trying to create the file for
3197 a maildir delivery, the message "Mailbox is full" was not appended to the
3198 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3199 only to a quota excession during the actual writing of the file.
3201 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3202 characters?) which causes problems in log lines. The DN values are now
3203 passed through string_printing() before being added to log lines.
3205 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3206 and InterBase are left for another time.)
3208 PH/21 Added message_body_newlines option.
3210 PH/22 Guard against possible overflow in moan_check_errorcopy().
3212 PH/23 POSIX allows open() to be a macro; guard against that.
3214 PH/24 If the recipient of an error message contained an @ in the local part
3215 (suitably quoted, of course), incorrect values were put in $domain and
3216 $local_part during the evaluation of errors_copy.
3222 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3223 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3226 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3227 issue a MAIL command.
3229 PH/02 In an ACL statement such as
3231 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3233 if a client was not listed at all, or was listed with a value other than
3234 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3235 the condition was not true (as it should be), so access was not denied.
3236 The bug was that the ! inversion was incorrectly passed on to the second
3237 item. This has been fixed.
3239 PH/03 Added additional dnslists conditions == and =& which are different from
3240 = and & when the dns lookup returns more than one IP address.
3242 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3243 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3245 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3246 FSYNC, which compiles an option called disable_fsync that allows for
3247 bypassing fsync(). The documentation is heavily laced with warnings.
3249 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3251 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3252 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3253 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3254 including adding "make clean"; (3) Added -fPIC when compiling the test
3255 dynamically loaded module, to get rid of a warning.
3257 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3258 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3259 = 0s. The bug is otherwise harmless.
3261 PH/07 There was a bug in the dovecot authenticator such that the value of
3262 $auth1 could be overwritten, and so not correctly preserved, after a
3263 successful authentication. This usually meant that the value preserved by
3264 the server_setid option was incorrect.
3266 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3268 PH/09 Installed PCRE release 7.0.
3270 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3271 run for batched SMTP input. It is now run at the start of every message
3272 in the batch. While fixing this I discovered that the process information
3273 (output by running exiwhat) was not always getting set for -bs and -bS
3274 input. This is fixed, and it now also says "batched" for BSMTP.
3276 PH/11 Added control=no_pipelining.
3278 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3279 patch, slightly modified), and move the expansion of helo_data till after
3280 the connection is made in the smtp transport (so it can use these
3283 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3285 PH/14 Added log_selector = +pid.
3287 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3289 PH/16 Add ${if forany and ${if forall.
3291 PH/17 Added dsn_from option to vary the From: line in DSNs.
3293 PH/18 Flush SMTP output before performing a callout, unless control =
3294 no_callout_flush is set.
3296 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3297 was true (the default) a successful delivery failed to delete the retry
3298 item, thus causing premature timeout of the address. The bug is now
3301 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3303 PH/21 Long custom messages for fakedefer and fakereject are now split up
3304 into multiline responses in the same way that messages for "deny" and
3305 other ACL rejections are.
3307 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3308 with slight modification.
3310 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3311 draft, changing the syntax and factoring some duplicate code".
3313 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3314 for deliveries of the second and subsequent messages over the same SMTP
3317 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3318 ${reduce, with only minor "tidies".
3320 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3322 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3323 expansion side effects.
3325 PH/27 When a message times out after an over-quota error from an Exim-imposed
3326 quota, the bounce message says "mailbox is full". This message was not
3327 being given when it was a system quota that was exceeded. It now should
3330 MH/03 Made $recipients available in local_scan(). local_scan() already has
3331 better access to the recipient list through recipients_list[], but
3332 $recipients can be useful in postmaster-provided expansion strings.
3334 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3335 in the case of a MAIL command with additional options following the
3336 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3337 were accidentally chopped off.
3339 PH/29 SMTP synchronization checks are implemented when a command is read -
3340 there is a check that no more input is waiting when there shouldn't be
3341 any. However, for some commands, a delay in an ACL can mean that it is
3342 some time before the response is written. In this time, more input might
3343 arrive, invalidly. So now there are extra checks after an ACL has run for
3344 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3345 pipelining has not been advertised.
3347 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3349 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3350 returned in the SMTP response when smtp_return_error_details was false.
3351 This has been fixed.
3353 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3354 instead of the C I/O that was originally supplied, because problems were
3355 reported on Solaris.
3357 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3358 Exim which did not show up earlier: it was assuming that a call to
3359 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3360 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3361 was a macro that became an assignment, so it seemed to work. This has
3362 changed to a proper function call with a void return, hence the compile
3363 error. Exim's code has been fixed.
3365 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3368 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3370 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3372 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3373 because it turns out to be too big or there is a timeout) there is no
3374 "Completed" line in the log. When some messages of this type were
3375 selected by exigrep, they were listed as "not completed". Others were
3376 picked up by some special patterns. I have improved the selection
3377 criteria to be more general.
3379 PH/38 The host_find_failed option in the manualroute router can now be set
3380 to "ignore", to completely ignore a host whose IP address cannot be
3381 found. If all hosts are ignored, the behaviour is controlled by the new
3382 host_all_ignored option.
3384 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3385 homing or because of multiple MX records with /mx) generated more than
3386 one IP address, and the following item turned out to be the local host,
3387 all the secondary addresses of the first item were incorrectly removed
3388 from the list, along with the local host and any following hosts (which
3389 is what is supposed to happen).
3391 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3392 whoever called Exim into the -H file. In the case of the daemon it was
3393 behaving confusingly. When first started, it used values for whoever
3394 started the daemon, but after a SIGHUP it used the Exim user (because it
3395 calls itself on a restart). I have changed the code so that it now always
3398 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3399 message are rejected with the same error (e.g. no authentication or bad
3400 sender address), and a DATA command is nevertheless sent (as can happen
3401 with PIPELINING or a stupid MUA), the error message that was given to the
3402 RCPT commands is included in the rejection of the DATA command. This is
3403 intended to be helpful for MUAs that show only the final error to their
3406 PH/42 Another patch from the Sieve maintainer.
3408 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3409 Eximstats - Fixed some broken HTML links and added missing column headers
3411 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3412 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3414 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3416 PH/43 Yet another patch from the Sieve maintainer.
3418 PH/44 I found a way to check for a TCP/IP connection going away before sending
3419 the response to the final '.' that terminates a message, but only in the
3420 case where the client has not sent further data following the '.'
3421 (unfortunately, this is allowed). However, in many cases there won't be
3422 any further data because there won't be any more messages to send. A call
3423 to select() can be used: if it shows that the input is "ready", there is
3424 either input waiting, or the socket has been closed. An attempt to read
3425 the next input character can distinguish the two cases. Previously, Exim
3426 would have sent an OK response which the client would never have see.
3427 This could lead to message repetition. This fix should cure that, at
3428 least in a lot of common cases.
3430 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3431 advertised in response to EHLO.
3437 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3438 fixed by 4.65/MH/01 (is this a record?) are fixed:
3440 (i) An empty string was always treated as zero by the numeric comparison
3441 operators. This behaviour has been restored.
3443 (ii) It is documented that the numeric comparison operators always treat
3444 their arguments as decimal numbers. This was broken in that numbers
3445 starting with 0 were being interpreted as octal.
3447 While fixing these problems I realized that there was another issue that
3448 hadn't been noticed. Values of message_size_limit (both the global option
3449 and the transport option) were treated as octal if they started with 0.
3450 The documentation was vague. These values are now always treated as
3451 decimal, and I will make that clear in the documentation.
3457 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3458 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3461 MH/01 Don't check that the operands of numeric comparison operators are
3462 integers when their expansion is in "skipping" mode (fixes bug
3463 introduced by 4.64-PH/07).
3465 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3466 child addresses, Exim now panics and dies. Previously, because the count
3467 is held in a short int, deliveries were likely to be lost. As such a
3468 large number of recipients for a single message is ridiculous
3469 (performance will be very, very poor), I have chosen to impose a limit
3470 rather than extend the field.
3476 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3477 leftover -K file (the existence of which was triggered by #402).
3478 While we were at it, introduced process PID as part of the -K
3479 filename. This should rule out race conditions when creating
3482 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3483 processing considerably. Previous code took too long for large mails,
3484 triggering a timeout which in turn triggers #401.
3486 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3487 in the DK code in transports.c. sendfile() is not really portable,
3488 hence the _LINUX specificness.
3490 TF/01 In the add_headers option to the mail command in an Exim filter,
3491 there was a bug that Exim would claim a syntax error in any
3492 header after the first one which had an odd number of characters
3495 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3496 callout verification, Exim cached a "reject" for the entire domain. This
3497 is correct for most verifications, but it is not correct for a recipient
3498 verification with use_sender or use_postmaster set, because in that case
3499 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3500 case of MAIL FROM:<> rejection from other early rejections (e.g.
3501 rejection of HELO). When verifying a recipient using a non-null MAIL
3502 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3503 Whatever the result of the callout, the value of the domain cache is
3504 left unchanged (for any other kind of callout, getting as far as trying
3505 RCPT means that the domain itself is ok).
3507 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3510 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3511 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3512 session, the daemon ignores ECONNECT errors and logs others; it now
3513 ignores EPIPE as well.
3515 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3516 (quoted-printable decoding).
3518 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3519 later the small subsequent patch to fix an introduced bug.
3521 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3523 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3525 PH/08 An error is now given if message_size_limit is specified negative.
3527 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3528 to be given (somewhat) arbitrary names.
3530 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3533 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3534 miscellaneous code fixes
3536 PH/10 Added the log_reject_target ACL modifier to specify where to log
3539 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3540 hostname. This is wrong, because it relates to the incoming message (and
3541 probably the interface on which it is arriving) and not to the outgoing
3542 callout (which could be using a different interface). This has been
3543 changed to use the value of the helo_data option from the smtp transport
3544 instead - this is what is used when a message is actually being sent. If
3545 there is no remote transport (possible with a router that sets up host
3546 addresses), $smtp_active_hostname is used.
3548 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3549 tweaks were necessary in order to get it to work (see also 21 below):
3550 (a) The code assumed that strncpy() returns a negative number on buffer
3551 overflow, which isn't the case. Replaced with Exim's string_format()
3553 (b) There were several signed/unsigned issues. I just did the minimum
3554 hacking in of casts. There is scope for a larger refactoring.
3555 (c) The code used strcasecmp() which is not a standard C function.
3556 Replaced with Exim's strcmpic() function.
3557 (d) The code set only $1; it now sets $auth1 as well.
3558 (e) A simple test gave the error "authentication client didn't specify
3559 service in request". It would seem that Dovecot has changed its
3560 interface. Fortunately there's a specification; I followed it and
3561 changed what the client sends and it appears to be working now.
3563 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3566 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3567 address A is aliased to B and C, where B exists and C does not. Without
3568 -v the output is "A verified" because verification stops after a
3569 successful redirection if more than one address is generated. However,
3570 with -v the child addresses are also verified. Exim was outputting "A
3571 failed to verify" and then showing the successful verification for C,
3572 with its parentage. It now outputs "B failed to verify", showing B's
3573 parentage before showing the successful verification of C.
3575 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3576 look up a TXT record in a specific list after matching in a combined
3579 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3580 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3581 they consult the DNS. I had assumed they would set it the way they
3582 wanted; and indeed my experiments on Linux seem to show that in some
3583 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3584 To be on the safe side, however, I have now made the interface to
3585 host_find_byname() similar to host_find_bydns(), with an argument
3586 containing the DNS resolver options. The host_find_byname() function now
3587 sets these options at its start, just as host_find_bydns() does. The smtp
3588 transport options dns_qualify_single and dns_search_parents are passed to
3589 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3590 of host_find_byname() use the default settings of RES_DEFNAMES
3591 (qualify_single) but not RES_DNSRCH (search_parents).
3593 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3594 spool_read_header() do less string testing, by means of a preliminary
3595 switch on the second character of optional "-foo" lines. (This is
3596 overdue, caused by the large number of possibilities that now exist.
3597 Originally there were few.) While I was there, I also converted the
3598 str(n)cmp tests so they don't re-test the leading "-" and the first
3599 character, in the hope this might squeeze out yet more improvement.
3601 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3602 flag allowing group syntax was set by the header_syntax check but not
3603 turned off, possible causing trouble later; (2) The flag was not being
3604 set at all for the header_verify test, causing "group"-style headers to
3605 be rejected. I have now set it in this case, and also caused header_
3606 verify to ignore an empty address taken from a group. While doing this, I
3607 came across some other cases where the code for allowing group syntax
3608 while scanning a header line wasn't quite right (mostly, not resetting
3609 the flag correctly in the right place). These bugs could have caused
3610 trouble for malformed header lines. I hope it is now all correct.
3612 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3613 with the "reply" argument non-NULL. The code, however (which originally
3614 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3615 but it didn't always do it. This confused somebody who was copying the
3616 code for some other use. I have removed all the tests.
3618 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3619 feature that was used to support insecure browsers during the U.S. crypto
3620 embargo. It requires special client support, and Exim is probably the
3621 only MTA that supported it -- and would never use it because real RSA is
3622 always available. This code has been removed, because it had the bad
3623 effect of slowing Exim down by computing (never used) parameters for the
3624 RSA_EXPORT functionality.
3626 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3627 authenticator to fail if there's a tab character in the incoming data
3628 (there should never be unless someone is messing about, as it's supposed
3629 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3630 if the connection is using TLS or if the remote IP is the same as the
3631 local IP, and the "valid-client-cert option" if a client certificate has
3634 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3635 authenticators. This can be used for authorization after authentication
3636 succeeds. (In the case of plaintext, it servers for both authentication
3639 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3640 if any retry times were supplied.
3642 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3643 connection, where there is no client IP address to check. In this
3644 situation, the verify now always succeeds.
3646 PH/25 Applied John Jetmore's -Mset patch.
3648 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3650 PH/27 In a string expansion for a processed (not raw) header when multiple
3651 headers of the same name were present, leading whitespace was being
3652 removed from all of them, but trailing whitespace was being removed only
3653 from the last one. Now trailing whitespace is removed from each header
3654 before concatenation. Completely empty headers in a concatenation (as
3655 before) are ignored.
3657 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3658 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3660 PH/29 [Removed. This was a change that I later backed out, and forgot to
3661 correct the ChangeLog entry (that I had efficiently created) before
3662 committing the later change.]
3664 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3665 address errors (4xx response to RCPT) over the same connection as other
3666 messages routed to the same hosts. Such deliveries are always "forced",
3667 so retry times are not inspected. This resulted in far too many retries
3668 for the affected addresses. The effect occurred only when there were more
3669 hosts than the hosts_max_try setting in the smtp transport when it had
3670 the 4xx errors. Those hosts that it had tried were not added to the list
3671 of hosts for which the message was waiting, so if all were tried, there
3672 was no problem. Two fixes have been applied:
3674 (i) If there are any address or message errors in an SMTP delivery, none
3675 of the hosts (tried or untried) are now added to the list of hosts
3676 for which the message is waiting, so the message should not be a
3677 candidate for sending over the same connection that was used for a
3678 successful delivery of some other message. This seems entirely
3679 reasonable: after all the message is NOT "waiting for some host".
3680 This is so "obvious" that I'm not sure why it wasn't done
3681 previously. Hope I haven't missed anything, but it can't do any
3682 harm, as the worst effect is to miss an optimization.
3684 (ii) If, despite (i), such a delivery is accidentally attempted, the
3685 routing retry time is respected, so at least it doesn't keep
3686 hammering the server.
3688 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3689 in ${readsocket because some servers need this prod.
3691 PH/32 Added some extra debug output when updating a wait-xxx database.
3693 PH/33 The hint "could be header name not terminated by colon", which has been
3694 given for certain expansion errors for a long time, was not being given
3695 for the ${if def:h_colon_omitted{... case.
3697 PH/34 The spec says: "With one important exception, whenever a domain list is
3698 being scanned, $domain contains the subject domain." There was at least
3699 one case where this was not true.
3701 PH/35 The error "getsockname() failed: connection reset by peer" was being
3702 written to the panic log as well as the main log, but it isn't really
3703 panic-worthy as it just means the connection died rather early on. I have
3704 removed the panic log writing for the ECONNRESET error when getsockname()
3707 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3708 runs only) independently of the message's sender address. This meant
3709 that, if the 4xx error was in fact related to the sender, a different
3710 message to the same recipient with a different sender could confuse
3711 things. In particular, this can happen when sending to a greylisting
3712 server, but other circumstances could also provoke similar problems.
3713 I have changed the default so that the retry time for these errors is now
3714 based a combination of the sender and recipient addresses. This change
3715 can be overridden by setting address_retry_include_sender=false in the
3718 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3719 remote server are returned as part of bounce messages. This was not
3720 happening for LMTP over a pipe (the lmtp transport), but now it is the
3721 same for both kinds of LMTP.
3723 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3724 in header lines that were in fact CNAMEs. This is no longer the case.
3726 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3727 and queue runs started by the daemon processed all messages. This has
3728 been fixed so that -R and -S can now usefully be given with -q<time>.
3730 PH/40 Import PCRE release 6.7 (fixes some bugs).
3732 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3734 PH/42 Give an error if -q is specified more than once.
3736 PH/43 Renamed the variables $interface_address and $interface_port as
3737 $received_ip_address and $received_port, to make it clear that these
3738 values apply to message reception, and not to the outgoing interface when
3739 a message is delivered. (The old names remain recognized, of course.)
3741 PH/44 There was no timeout on the connect() call when using a Unix domain
3742 socket in the ${readsocket expansion. There now is.
3744 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3745 be meaningful with "accept".
3747 SC/01 Eximstats V1.43
3748 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3750 SC/02 Eximstats V1.44
3751 Use a glob alias rather than an array ref in the generated
3752 parser. This improves both readability and performance.
3754 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3755 Collect SpamAssassin and rejection statistics.
3756 Don't display local sender or destination tables unless
3757 there is data to show.
3758 Added average volumes into the top table text output.
3760 SC/04 Eximstats V1.46
3761 Collect data on the number of addresses (recipients)
3762 as well as the number of messages.
3764 SC/05 Eximstats V1.47
3765 Added 'Message too big' to the list of mail rejection
3766 reasons (thanks to Marco Gaiarin).
3768 SC/06 Eximstats V1.48
3769 Mainlog lines which have GMT offsets and are too short to
3770 have a flag are now skipped.
3772 SC/07 Eximstats V1.49 (Alain Williams)
3773 Added the -emptyok flag.
3775 SC/08 Eximstats V1.50
3776 Fixes for obtaining the IP address from reject messages.
3778 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3779 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3780 whitespace changes from 4.64-PH/27
3782 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3785 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3786 are found, allow negative numbers in numeric criteria)
3788 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3790 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3791 to match changes made in 4.64-PH/43
3793 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3795 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3796 "domain" is actually an IP address, and force a failure. This locks out
3797 those revolvers/nameservers that support "A-for-A" lookups, in
3798 contravention of the specifications.
3800 PH/48 When a host name was looked up from an IP address, and the subsequent
3801 forward lookup of the name timed out, the host name was left in
3802 $sender_host_name, contrary to the specification.
3804 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3805 restricted to single-key lookups, Exim was not diagnosing an error if
3806 * or *@ was used with a query-style lookup.
3808 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3810 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3811 long ago, but noone interested enough thought of it. Let's just say that
3812 the "1.1" means that there are some new functions that weren't there at
3813 some point in the past.
3815 PH/51 Error processing for expansion failure of helo_data from an smtp
3816 transport during callout processing was broken.
3818 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3819 tested/used via the -bh/-bhc/-bs options.
3821 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3822 bug, fixed in subsequent PCRE releases).
3824 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3825 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3827 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3833 SC/01 Use a glob alias rather than an array ref in eximstats generated
3834 parser. This improves both readability and performance.
3836 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3837 Don't display local sender or destination tables in eximstats unless
3838 there is data to show.
3839 Added average volumes into the eximstats top table text output.
3841 SC/03 Collect data on the number of addresses (recipients) as well
3842 as the number of messages in eximstats.
3844 TF/01 Correct an error in the documentation for the redirect router. Exim
3845 does (usually) call initgroups() when daemonizing.
3847 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3848 with consistent privilege compared to when running as a daemon.
3850 TF/03 Note in the spec that $authenticated_id is not set for local
3851 submissions from trusted users.
3853 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3854 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3856 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3857 by adding some example configuration directives to the default
3858 configuration file. A little bit of work is required to uncomment the
3859 directives and define how usernames and passwords are checked, but
3860 there is now a framework to start from.
3862 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3863 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3864 without this. I don't know how relevant this is to other LDAP libraries.
3866 PH/02 Add the verb name to the "unknown ACL verb" error.
3868 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3870 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3872 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3873 directory not expanded when it should be if an expanded home directory
3874 was set for the address (which is overridden by the transport).
3876 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3879 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3880 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3881 because it is too late at that time, and has no effect.
3883 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3884 security issue with \' (bugzilla #107). I could not use the
3885 PQescapeStringConn() function, because it needs a PGconn value as one of
3888 PH/08 When testing addresses using -bt, indicate those final addresses that
3889 are duplicates that would not cause an additional delivery. At least one
3890 person was confused, thinking that -bt output corresponded to deliveries.
3891 (Suppressing duplicates isn't a good idea as you lose the information
3892 about possibly different redirections that led to the duplicates.)
3894 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3895 systems where poll() doesn't work, in particular OS X.
3897 PH/10 Added more information to debugging output for retry time not reached.
3899 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3900 operations in malware.c.
3902 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3905 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3906 syslog_duplication set false, log lines that would normally be written
3907 both the the main log and to the reject log were not written to syslog at
3910 PH/14 In the default configuration, change the use of "message" in ACL warn
3911 statements to "add_header".
3913 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3914 not followed by a command (e.g. "seen endif").
3916 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3917 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3920 PH/17 Added extra conditions to the default value of delay_warning_condition
3924 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3925 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3926 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3929 The Auto-Submitted: and various List- headers are standardised, whereas I
3930 don't think Precedence: ever was.
3932 PH/18 Refactored debugging code in route_finduser() to show more information,
3933 in particular, the error code if getpwnam() issues one.
3935 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3936 This is apparently needed in addition to the PH/07 change above to avoid
3937 any possible encoding problems.
3939 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3940 but not after initializing Perl.
3942 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3943 output them only if debugging. By default they are written stderr,
3944 apparently, which is not desirable.
3946 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3949 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3952 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3954 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3955 authenticated or an ident call has been made. Suppress the default
3956 values for $authenticated_id and $authenticated_sender (but permit -oMai
3957 and -oMas) when testing with -bh.
3959 PH/24 Re-jigged the order of the tests in the default configuration so that the
3960 tests for valid domains and recipients precede the DNS black list and CSA
3961 tests, on the grounds that those ones are more expensive.
3963 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3964 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3965 This bug exists in every version of Exim that I still have, right back to
3968 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3969 However, an attempt to turn on case-sensitivity in a regex key by
3970 including (?-i) didn't work because the subject string was already
3971 lowercased, and the effects were non-intuitive. It turns out that a
3972 one-line patch can be used to allow (?-i) to work as expected.
3978 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3979 other effects) broke the use of negated acl sub-conditions.
3981 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3984 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3985 "Deny" causes Exim to reject the incoming connection with a 554 error.
3986 Unfortunately, if there is a major crisis, such as a disk failure,
3987 tcp-wrappers gives "deny", whereas what one would like would be some
3988 kind of temporary error. A kludge has been added to help with this.
3989 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3990 554 error is used if errno is still zero or contains ENOENT (which occurs
3991 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3994 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3996 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3997 errors. Otherwise a message that provokes a temporary error (when other
3998 messages do not) can cause a whole host to time out.
4000 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4001 addresses were routed directly to files or pipes from a redirect router.
4002 File deliveries just didn't batch; pipe deliveries might have suffered
4005 PH/06 A failure to get a lock for a hints database would erroneously always say
4006 "Failed to get write lock", even when it was really a read lock.
4008 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4009 of 0600. This has been changed to use the value of the lockfile_mode
4010 option (which defaults to 0600).
4012 PH/08 Applied small patch from the Sieve maintainer.
4014 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4015 folder from quota calculations, a direct delivery into this folder messed
4016 up the contents of the maildirsize file. This was because the regex was
4017 used only to exclude .Trash (or whatever) when the size of the mailbox
4018 was calculated. There was no check that a delivery was happening into an
4019 excluded directory. This bug has been fixed by ignoring all quota
4020 processing for deliveries into excluded directories.
4022 PH/10 Added the maildirfolder_create_regex option to appendfile.
4028 PH/01 The code for finding all the local interface addresses on a FreeBSD
4029 system running IPv6 was broken. This may well have applied to all BSD
4030 systems, as well as to others that have similar system calls. The broken
4031 code found IPv4 interfaces correctly, but gave incorrect values for the
4032 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4033 that it would not match correctly against @[] and not recognize the IPv6
4036 PH/02 The ipliteral router was not recognizing addresses of the form user@
4037 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4039 PH/03 Added disable_ipv6.
4041 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4042 decoded form, because it is most often used to construct To: headers
4043 lines in autoreplies, and the decoded form may well be syntactically
4044 invalid. However, $reply_address has leading white space removed, and all
4045 newlines turned into spaces so that the autoreply transport does not
4048 PH/05 If group was specified without a user on a router, and no group or user
4049 was specified on a transport, the group from the router was ignored.
4051 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4052 for visible compile-time settings that can be used to change these
4053 numbers, for those that want even more. Backwards compatibility with old
4054 spool files has been maintained. However, going back to a previous Exim
4055 release will lost any variables that are in spool files.
4057 PH/07 Two small changes when running in the test harness: increase delay when
4058 passing a TCP/IP connection to a new process, in case the original
4059 process has to generate a bounce, and remove special handling of
4060 127.0.0.2 (sic), which is no longer necessary.
4062 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4063 be the same on different OS.
4065 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4068 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4069 whether --show-vars was specified or not
4071 JJ/02 exipick: Added support for new ACL variable spool format introduced
4074 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4075 syntactically invalid From: or Reply-to: line, and a filter used this to
4076 generate an autoreply, and therefore failed to obtain an address for the
4077 autoreply, Exim could try to deliver to a non-existent relative file
4078 name, causing unrelated and misleading errors. What now happens is that
4079 it logs this as a hard delivery error, but does not attempt to create a
4082 PH/11 The exinext utility has a -C option for testing purposes, but although
4083 the given file was scanned by exinext itself; it wasn't being passed on
4084 when Exim was called.
4086 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4087 an end-of-file indication when reading a command response.
4089 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4090 compiled. In many other places in Exim, IPv6 addresses are always
4091 recognized, so I have changed this. It also means that IPv4 domain
4092 literals of the form [IPV4:n.n.n.n] are now always recognized.
4094 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4095 used if the router is not running as root, for example, when verifying at
4096 ACL time, or when using -bh. The debugging output from this situation was
4097 non-existent - all you got was a failure to exec. I have made two
4100 (a) Failures to set uid/gid, the current directory, or a process leader
4101 in a subprocess such as that created by queryprogram now generate
4102 suitable debugging output when -d is set.
4104 (b) The queryprogram router detects when it is not running as root,
4105 outputs suitable debugging information if -d is set, and then runs
4106 the subprocess without attempting to change uid/gid.
4108 PH/15 Minor change to Makefile for building test_host (undocumented testing
4111 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4112 additional section of a DNS packet that returns MX or SRV records.
4113 Instead, it always explicitly searches for A/AAAA records. This avoids
4114 major problems that occur when a DNS server includes only records of one
4115 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4116 fixed another bug: if SRV records were looked up and the corresponding
4117 address records were *not* found in the additional section, the port
4118 values from the SRV records were lost.
4120 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4121 using the correct key (the original address) when searching the retry
4122 rules in order to find which one to use for generating the retry hint.
4124 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4125 adding the default one. Similarly, if it contains a Reply-To: header, the
4126 errors_reply_to option, if set, is not used.
4128 PH/19 When calculating a retry time, Exim used to measure the "time since
4129 failure" by looking at the "first failed" field in the retry record. Now
4130 it does not use this if it is later than than the arrival time of the
4131 message. Instead it uses the arrival time. This makes for better
4132 behaviour in cases where some deliveries succeed, thus re-setting the
4133 "first failed" field. An example is a quota failure for a huge message
4134 when small messages continue to be delivered. Without this change, the
4135 "time since failure" will always be short, possible causing more frequent
4136 delivery attempts for the huge message than are intended.
4137 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4139 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4140 $1, $2, $3) because the numerical variables can be reset during some
4141 expansion items (e.g. "match"), thereby losing the authentication data.
4143 PH/21 Make -bV show the size of off_t variables so that the test suite can
4144 decide whether to run tests for quotas > 2G.
4146 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4147 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4148 filecount value is greater than 2G or if a quota value is greater than 2G
4149 on a system where the size of off_t is not greater than 4, a panic error
4152 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4153 never match. The debug and -bh output now contains an explicit error
4154 message indicating a malformed IPv4 address or mask.
4156 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4157 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4158 PH/23 above applies.
4160 PH/25 Do not write to syslog when running in the test harness. The only
4161 occasion when this arises is a failure to open the main or panic logs
4162 (for which there is an explicit test).
4164 PH/26 Added the /no_tell option to "control=freeze".
4166 PH/27 If a host name lookup failed very early in a connection, for example, if
4167 the IP address matched host_lookup and the reverse lookup yielded a name
4168 that did not have a forward lookup, an error message of the form "no IP
4169 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4170 could be logged. Now it outputs the IP address instead of "NULL".
4172 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4173 allows the sender and the authenticated sender to be set when
4174 submitting a message from within Exim. Since child_open_exim() is
4175 documented for local_scan(), the new function should be too.
4177 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4178 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4179 results in an empty string is now treated as unset.
4181 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4183 PH/31 Added sender_verify_fail logging option.
4185 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4186 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4187 tidied the source and removed it altogether.
4189 PH/33 When a queue run was abandoned because the load average was too high, a
4190 log line was always written; now it is written only if the queue_run log
4191 selector is set. In addition, the log line for abandonment now contains
4192 information about the queue run such as the pid. This is always present
4193 in "start" and "stop" lines but was omitted from the "abandon" line.
4195 PH/34 Omit spaces between a header name and the colon in the error message that
4196 is given when verify = headers_syntax fails (if there are lots of them,
4197 the message gets confusing).
4199 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4200 names, as there are now some PTR records that contain slashes. This check
4201 is only to protect against broken name servers that fall over on strange
4202 characters, so the fact that it applies to all lookups doesn't matter.
4204 PH/36 Now that the new test suite is complete, we can remove some of the
4205 special code in Exim that was needed for the old test suite. For example,
4206 sorting DNS records because real resolvers return them in an arbitrary
4207 order. The new test suite's fake resolver always returns records in the
4210 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4211 bounces) except when queue_only is set, to avoid logging races between
4212 the different processes.
4214 PH/38 Panic-die if .include specifies a non-absolute path.
4216 PH/39 A tweak to the "H" retry rule from its user.
4218 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4219 a label. They prevented compilation on older perls.
4221 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4222 a warning to be raised on newish perls.
4224 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4225 on queue. Changes to match documented behaviour of showing count of
4226 messages matching specified criteria.
4228 PH/40 Changed the default ident timeout from 30s to 5s.
4230 PH/41 Added support for the use of login_cap features, on those BSD systems
4231 that have them, for controlling the resources used by pipe deliveries.
4233 PH/42 The content-scanning code uses fopen() to create files in which to put
4234 message data. Previously it was not paying any attention to the mode of
4235 the files. Exim runs with umask(0) because the rest of the code creates
4236 files with open(), and sets the required mode explicitly. Thus, these
4237 files were ending up world-writeable. This was not a big issue, because,
4238 being within the spool directory, they were not world-accessible. I have
4239 created a function called modefopen, which takes an additional mode
4240 argument. It sets umask(777), creates the file, chmods it to the required
4241 mode, then resets the umask. All the relevant calls to fopen() in the
4242 content scanning code have been changed to use this function.
4244 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4245 to 24 hours. This avoids potential overflow problems when processing G
4246 and H retry rules. I suspect nobody ever tinkers with this value.
4248 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4250 PH/45 When the plaintext authenticator is running as a client, the server's
4251 challenges are checked to ensure they are valid base64 strings. By
4252 default, the authentication attempt is cancelled if an invalid string is
4253 received. Setting client_ignore_invalid_base64 true ignores these errors.
4254 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4255 they are received. Thus, the responses can be made to depend on the
4256 challenges. If an invalid string is ignored, an empty string is placed in
4259 PH/46 Messages that are created by the autoreply transport now contains a
4260 References: header, in accordance with RFCs 2822 and 3834.
4262 PH/47 Added authenticated_sender_force to the smtp transport.
4264 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4266 PH/49 Installed latest patch from the Sieve maintainer.
4268 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4269 was also set, the appendfile transport was unnecessarily scanning a
4270 directory of message files (e.g. for maildir delivery) to find the count
4271 of files (along with the size), even though it did not need this
4272 information. It now does the scan only if it needs to find either the
4273 size of the count of files.
4275 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4277 PH/52 Two bugs concerned with error handling when the smtp transport is
4280 (i) Exim was not creating retry information for temporary errors given
4281 for individual recipients after the DATA command when the smtp transport
4282 was used in LMTP mode. This meant that they could be retried too
4283 frequently, and not timed out correctly.
4285 (ii) Exim was setting the flag that allows error details to be returned
4286 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4287 recipients that were returned after the DATA command.
4289 PH/53 This is related to PH/52, but is more general: for any failing address,
4290 when detailed error information was permitted to be returned to the
4291 sender, but the error was temporary, then after the final timeout, only
4292 "retry timeout exceeded" was returned. Now it returns the full error as
4293 well as "retry timeout exceeded".
4295 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4296 do this, and (what is worse) MTAs that accept it.
4298 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4299 will now be deprecated.
4301 PH/56 New os.c-cygwin from the Cygwin maintainer.
4303 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4304 formats (previously only available in exim formats via -bpr, -bpru,
4305 and -bpra. Now also available in native and exiqgrep formats)
4307 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4308 with very large, slow to parse queues
4310 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4312 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4314 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4315 responses to authentication challenges, though it was showing the
4316 challenges; (ii) I've removed the CR characters from the debug output for
4319 PH/58 Allow for the insertion of a newline as well as a space when a string
4320 is turned into more than one encoded-word during RFC 2047 encoding. The
4321 Sieve code now uses this.
4323 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4324 data_4xx, lost_connection, tls_required.
4326 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4327 message was being sent as an SMTP response.
4329 PH/61 Add -l and -k options to exicyclog.
4331 PH/62 When verifying, if an address was redirected to one new address, so that
4332 verification continued, and the new address failed or deferred after
4333 having set something in $address_data, the value of $address_data was not
4334 passed back to the ACL. This was different to the case when no
4335 redirection occurred. The value is now passed back in both cases.
4337 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4338 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4339 use login_cap.h, so on its own it isn't the distinguishing feature. The
4340 new name refers directly to the setclassresources() function.
4342 PH/65 Added configuration files for NetBSD3.
4344 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4346 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4347 is preferred over IPv4.
4349 PH/68 The bounce_return_message and bounce_return_body options were not being
4350 honoured for bounces generated during the reception of non-SMTP messages.
4351 In particular, this applied to messages rejected by the ACL. This bug has
4352 been fixed. However, if bounce_return_message is true and bounce_return_
4353 body is false, the headers that are returned for a non-SMTP message
4354 include only those that have been read before the error was detected.
4355 (In the case of an ACL rejection, they have all been read.)
4357 PH/69 The HTML version of the specification is now built in a directory called
4358 spec_html instead of spec.html, because the latter looks like a path with
4359 a MIME-type, and this confuses some software.
4361 PH/70 Catch two compiler warnings in sieve.c.
4363 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4364 function verify_get_ident() calls ip_connect() to connect a socket, but
4365 if the "connect()" function timed out, ip_connect() used to close the
4366 socket. However, verify_get_ident() also closes the socket later, and in
4367 between Exim writes to the log, which may get opened at this point. When
4368 the socket was closed in ip_connect(), the log could get the same file
4369 descriptor number as the socket. This naturally causes chaos. The fix is
4370 not to close the socket in ip_connect(); the socket should be closed by
4371 the function that creates it. There was only one place in the code where
4372 this was missing, in the iplookup router, which I don't think anybody now
4373 uses, but I've fixed it anyway.
4375 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4376 well as to direct DNS lookups. Otherwise the handling of names in host
4377 lists is inconsistent and therefore confusing.
4383 PH/01 Two changes to the default runtime configuration:
4385 (1) Move the checks for relay_from_hosts and authenticated clients from
4386 after to before the (commented out) DNS black list checks.
4388 (2) Add control=submission to the relay_from_hosts and authenticated
4389 clients checks, on the grounds that messages accepted by these
4390 statements are most likely to be submissions.
4392 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4394 (1) Generate an error if the third argument for the ${prvs expansion is
4397 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4400 (3) Reset the variables that are obtained from the first argument of
4401 ${prvscheck and used in the second argument before leaving the code,
4402 because their memory is reclaimed, so using them afterwards may do
4405 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4406 one (it's much easier than Tom thought :-).
4408 (5) Because of (4), we can now allow for the use of $prvscheck_result
4409 inside the third argument.
4411 PH/03 For some reason, the default setting of PATH when running a command from
4412 a pipe transport was just "/usr/bin". I have changed it to
4415 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4416 anything to be listed in the output from -bV.
4418 PH/05 When a filter generated an autoreply, the entire To: header line was
4419 quoted in the delivery log line, like this:
4421 => >A.N.Other <ano@some.domain> <original@ddress> ...
4423 This has been changed so that it extracts the operative address. There
4424 may be more than one such address. If so, they are comma-separated, like
4427 => >ano@some.domain,ona@other.domain <original@ddress> ...
4429 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4430 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4431 not being looked up in the rDNS to get a host name, Exim was showing the
4432 IP address twice in Received: lines, even though the IP addresses were
4433 identical. For example:
4435 Received: from [1.2.3.4] (helo=[1.2.3.4])
4437 However, if the real host name was known, it was omitting the HELO data
4438 if it matched the actual IP address. This has been tidied up so that it
4439 doesn't show the same IP address twice.
4441 PH/07 When both +timestamp and +memory debugging was on, the value given by
4442 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4443 called by the debug printing, thereby overwriting the timestamp buffer.
4444 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4446 PH/08 When the original message was included in an autoreply transport, it
4447 always said "this is a copy of the message, including all the headers",
4448 even if body_only or headers_only was set. It now gives an appropriate
4451 PH/09 Applied a patch from the Sieve maintainer which:
4453 o fixes some comments
4454 o adds the (disabled) notify extension core
4455 o adds some debug output for the result of if/elsif tests
4456 o points to the current vacation draft in the documentation
4457 and documents the missing references header update
4461 o fixes a bug in processing the envelope test (when testing
4462 multiple envelope elements, the last element determined the
4465 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4466 Electronic Mail") by including:
4468 Auto-submitted: auto-generated
4470 in the messages that it generates (bounce messages and others, such as
4471 warnings). In the case of bounce messages for non-SMTP messages, there was
4472 also a typo: it was using "Auto_submitted" (underscore instead of
4473 hyphen). Since every message generated by Exim is necessarily in response
4474 to another message, thes have all been changed to:
4476 Auto-Submitted: auto-replied
4478 in accordance with these statements in the RFC:
4480 The auto-replied keyword:
4482 - SHOULD be used on messages sent in direct response to another
4483 message by an automatic process,
4485 - MUST NOT be used on manually-generated messages,
4487 - MAY be used on Delivery Status Notifications (DSNs) and Message
4488 Disposition Notifications (MDNs),
4490 - MUST NOT be used on messages generated by automatic or periodic
4491 processes, except for messages which are automatic responses to
4494 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4495 to the default Received: header definition.
4497 PH/12 Added log selector acl_warn_skipped (default on).
4499 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4500 variables because (a) they are in the wrong storage pool and (b) even if
4501 they were copied, it wouldn't work properly because of the caching.
4503 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4504 checking when decoding. Apparently there are clients that generate
4505 overlong encoded strings. Why am I not surprised?
4507 PH/15 If the first argument of "${if match_address" was not empty, but did not
4508 contain an "@" character, Exim crashed. Now it writes a panic log message
4509 and treats the condition as false.
4511 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4513 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4514 extension "envelope-auth". The code is finished and in agreement with
4515 other implementations, but there is no documentation so far and in fact,
4516 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4517 not changing the active code.
4519 Print executed "if" and "elsif" statements when debugging is used. This
4520 helps a great deal to understand what a filter does.
4522 Document more things not specified clearly in RFC3028. I had all this
4523 sorted out, when out of a sudden new issues came to my mind. Oops."
4525 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4528 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4529 canonical form (as documented). However, after a host name lookup from
4530 the IP address, check_host() was doing a simple string comparison with
4531 addresses acquired from the DNS when checking that the found name did
4532 have the original IP as one of its addresses. Since any found IPv6
4533 addresses are likely to be in abbreviated form, the comparison could
4534 fail. Luckily, there already exists a function for doing the comparison
4535 by converting both addresses to binary, so now that is used instead of
4536 the text comparison.
4538 PH/20 There was another similar case to PH/19, when a complete host name was
4539 given in a host list; looking up its IP address could give an abbreviated
4540 form, whereas the current host's name might or might not be abbreviated.
4541 The same fix has been applied.
4547 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4548 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4551 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4552 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4554 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4556 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4557 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4558 Also, for "Auto-Submitted", treat anything other than "no" as
4559 non-personal, in accordance with RFC 3834. (Previously it treated
4560 anything starting "auto-" as non-personal.)
4562 TF/01 The control=submission/name=... option had a problem with syntax
4563 errors if the name included a slash character. The /name= option
4564 now slurps the rest of the string, so it can include any characters
4565 but it must come last in the list of options (after /sender_retain
4568 PH/05 Some modifications to the interface to the fake nameserver for the new
4576 TK/01 Added the "success_on_redirect" address verification option. See
4577 NewStuff for rationale and an example.
4579 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4581 PH/02 Patch to exigrep to allow it to work on syslog lines.
4583 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4584 fread() to skip over the body file's header line, because in Cygwin the
4585 header line is locked and is inaccessible.
4587 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4588 co-exist for some time) to make it clear that it is the Exim ID that is
4589 referenced, not the Message-ID: header line.
4591 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4592 string_format() function, because snprintf() does not exist on all
4595 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4596 ${stat: expansion item.
4598 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4599 protocol synchronization error", to keep the pedants happy.
4601 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4602 well as for IRIX systems, when gcc is being used. See the host.c source
4605 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4607 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4610 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4611 transport and to the smtp transport in LMTP mode.
4613 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4615 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4616 run a filter in a subprocess. This could lead to confusion in subsequent
4617 lookups in the parent process. There should also be a search_tidyup() at
4618 the end of the subprocess.
4620 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4621 only if the host matched helo_try_verify_hosts, which caused the
4622 verification to occur when the EHLO/HELO command was issued. The ACL just
4623 tested the remembered result. Now, if a previous verification attempt has
4624 not happened, "verify = helo" does it there and then.
4626 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4628 TK/03 Fix log output including CR from clamd.
4630 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4631 exist provoked a memory error which could cause a segfault.
4633 PH/15 Installed PCRE 6.2
4635 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4637 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4638 of the problem. Specifically, suggested +O2 rather than +O1 for the
4641 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4643 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4644 record, the port was not correctly logged when the outgoing_port log
4645 selector was set (it logged the transort's default port).
4647 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4648 fallback_hosts, and "hosts" in the smtp transport.
4650 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4651 host errors such as "Connection refused".
4653 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4654 authentication with radiusclient 0.4.9:
4656 - Error returned from rc_read_config was caught wrongly
4657 - Username/password not passed on to radius server due to wrong length.
4659 The presumption is that some radiusclient API changes for 4.51/PH/17
4660 were not taken care of correctly. The code is still untested by me (my
4661 Linux distribution still has 0.3.2 of radiusclient), but it was
4662 contributed by a Radius user.
4664 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4665 expanding the "port" option of the smtp transport.
4667 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4668 while reading a MIME header. Thanks to Tom Hughes for a patch.
4670 PH/24 Include config.h inside local_scan.h so that configuration settings are
4673 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4674 that in an ACL for RCPT (for example), you can examine exactly what was
4677 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4678 commands, but it was not correctly comparing the address with the actual
4679 client host address. Thus, it would show the EHLO address in Received:
4680 header lines when this was not necessary.
4682 PH/27 Added the % operator to ${eval:}.
4684 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4685 it should be ignoring failures (because with -C, for example, it has lost
4686 privilege). It wasn't ignoring creation failures other than "already
4689 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4692 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4693 because an input line was too long, either on its own, or by virtue of
4694 too many continuations, the temporary file was not being removed, and the
4695 return code was incorrect.
4697 PH/31 Missing "BOOL" in function definition in filtertest.c.
4699 PH/32 Applied Sieve patches from the maintainer.
4701 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4703 PH/33 Added "verify = not_blind".
4705 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4706 Local/Makefile (with some defaults set). These are used in built scripts
4707 such as exicyclog, but they have never been used in the exim_install
4708 script (though there are many overriding facilities there). I have
4709 arranged that the exim_install script now takes note of these two
4712 PH/35 Installed configuration files for Dragonfly.
4714 PH/36 When a locally submitted message by a trusted user did not contain a
4715 From: header, and the sender address was obtained from -f or from an SMTP
4716 MAIL command, and the trusted user did not use -F to supply a sender
4717 name, $originator_name was incorrectly used when constructing a From:
4718 header. Furthermore, $originator_name was used for submission mode
4719 messages from external hosts without From: headers in a similar way,
4720 which is clearly wrong.
4722 PH/37 Added control=suppress_local_fixups.
4724 PH/38 When log_selector = +received_sender was set, and the addition of the
4725 sender made the log line's construction buffer exactly full, or one byte
4726 less than full, an overflow happened when the terminating "\n" was
4729 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4730 when the result of a list match is failure because a DNS lookup failed.
4732 PH/40 RM_COMMAND is now used in the building process.
4734 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4735 the "build-* directories that it finds.
4737 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4738 address in a domain literal was a prefix of an interface address.
4740 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4741 when verifying a sender address, unless rewrite_headers is false.
4743 PH/44 Wrote a long comment about why errors_to addresses are verified as
4744 recipients, not senders.
4746 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4747 the ratelimit ACL was added.
4749 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4751 PH/46 Added extra information about PostgreSQL errors to the error string.
4753 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4754 avoiding the need to install special zones in a real server. This is
4755 backwards compatible; if it can't find the fake resolver, it drops back.
4756 Thus, both old and new test suites can be run.
4758 TF/02 Added util/ratelimit.pl
4760 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4761 clock is set back in time.
4763 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4764 Candler <B.Candler@pobox.com>.
4766 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4767 OK for addresses that are the result of redirections.
4769 PH/48 A number of further additions for the benefit of the new test suite,
4770 including a fake gethostbyname() that interfaces to the fake DNS resolver
4773 TF/06 The fix for widen_domains has also been applied to qualify_single and
4774 search_parents which are the other dnslookup options that can cause
4777 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4780 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4782 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4783 service" if it encountered an MX record with an empty target hostname.
4784 The message is now "an MX or SRV record indicated no SMTP service".
4786 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4787 if the DNS of the sending site is misconfigured. This is quite a
4788 common situation. This change restores the behaviour of treating a
4789 helo verification defer as a failure.
4791 PH/51 If self=fail was set on a router, the bounce message did not include the
4792 actual error message.
4798 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4800 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4801 command itself ended in error, the underlying message about the transport
4802 filter timeout was being overwritten with the pipe command error. Now the
4803 underlying error message should be appended to the second error message.
4805 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4807 PH/02 Reduce the amount of output that "make" produces by default. Full output
4808 can still be requested.
4810 PH/03 The warning log line about a condition test deferring for a "warn" verb
4811 was being output only once per connection, rather than after each
4812 occurrence (because it was using the same function as for successful
4813 "warn" verbs). This seems wrong, so I have changed it.
4815 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4816 it should not have, which might have caused a crash in the right
4817 circumstances, but probably never did.
4819 PH/04 Installed a modified version of Tony Finch's patch to make submission
4820 mode fix the return path as well as the Sender: header line, and to
4821 add a /name= option so that you can make the user's friendly name appear
4824 TF/03 Added the control = fakedefer ACL modifier.
4826 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4827 Mark Lowes for thorough testing.
4829 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4831 TK/03 Merged latest SRS patch from Miles Wilton.
4833 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4834 with the definition in sysexits.h (which is #included earlier).
4835 Fortunately, Exim does not actually use EX_OK. The code used to try to
4836 preserve the sysexits.h value, by assuming that macro definitions were
4837 scanned for macro replacements. I have been disabused of this notion,
4838 so now the code just undefines EX_OK before #including unistd.h.
4840 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4841 in the smtp transport. When a block could not be written in a single
4842 write() function, the timeout was being re-applied to each part-write.
4843 This seems wrong - if the receiver was accepting one byte at a time it
4844 would take for ever. The timeout is now adjusted when this happens. It
4845 doesn't have to be particularly precise.
4847 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4848 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4850 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4851 without a domain if the check to <postmaster@domain> fails.
4853 SC/01 Eximstats: added -xls and the ability to specify output files
4854 (patch written by Frank Heydlauf).
4856 SC/02 Eximstats: use FileHandles for outputting results.
4858 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4860 SC/04 Eximstats: fixed display of large numbers with -nvr option
4862 SC/05 Eximstats: fixed merging of reports with empty tables.
4864 SC/06 Eximstats: added the -include_original_destination flag
4866 SC/07 Eximstats: removed tabs and trailing whitespace.
4868 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4870 TK/06 MBOX spool code: Add real "From " MBOX separator line
4871 so the .eml file is really in mbox format (even though
4872 most programs do not really care). Patch from Alex Miller.
4874 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4875 The latter is generated from $received_to and is only set if the
4876 message has one envelope recipient. SA can use these headers,
4877 obviously out-of-the-box. Patch from Alex Miller.
4879 PH/08 The ${def test on a variable was returning false if the variable's
4880 value was "0", contrary to what the specification has always said!
4881 The result should be true unless the variable is empty.
4883 PH/09 The syntax error of a character other than { following "${if
4884 def:variable_name" (after optional whitespace) was not being diagnosed.
4885 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4886 accidental colon was present, for example, could give incorrect results.
4888 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4889 result is used (not including appendfile, where other changes are about
4892 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4893 This involved changing a lot of size variables from int to off_t. It
4894 should work with maildirs and everything.
4896 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4897 spamd dying while we are connected to it.
4899 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4902 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4903 function for BDB 4.3.
4905 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4907 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4908 the sources for PCRE have been re-arranged and more files are now
4911 PH/15 The code I had for printing potentially long long variables in PH/11
4912 above was not the best (it lost precision). The length of off_t variables
4913 is now inspected at build time, and an appropriate printing format (%ld
4914 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4915 to be "long long int" or "long int". This is needed for the internal
4916 formatting function string_vformat().
4918 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4919 the configuration file to be ":syslog", then the script "guesses" where
4920 the logs files are, rather than using the compiled in default. In our
4921 case the guess is not the same as the compiled default, so the script
4922 suddenly stopped working when I started to use syslog. The patch checks
4923 to see if log_file_path is "". If so, it attempts to read it from exim
4924 with no configuration file to get the compiled in version, before it
4925 falls back to the previous guessing code."
4927 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4928 implementing BATV in an Exim configuration. See NewStuff for the gory
4931 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4932 Makefile that are specific to HP-UX.
4934 PH/18 If the "use_postmaster" option was set for a recipient callout together
4935 with the "random" option, the postmaster address was used as the MAIL
4936 FROM address for the random test, but not for the subsequent recipient
4937 test. It is now used for both.
4939 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4940 patch removes a few documentation additions to RFC 3028, because the
4941 latest draft now contains them. It adds the new en;ascii-case comparator
4942 and a new error check for 8bit text in MIME parts. Comparator and
4943 require names are now matched exactly. I enabled the subaddress
4944 extension, but it is not well tested yet (read: it works for me)."
4946 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4947 rework some of the code of TK/09 above to avoid the hardwired use of
4948 "%lld" and "long long". Replaced the call to snprintf() with a call to
4951 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4952 records point to non-existent hosts", "retry timeout exceeded", and
4953 "retry time not reached for any host after a long failure period".
4955 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4956 experimental DomainKeys support:
4958 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4959 (2) On an error such as an illegally used "control", the wrong name for
4960 the control was given.
4962 These problems did NOT occur unless DomainKeys support was compiled.
4964 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4966 PH/24 Added ${if match_ip condition.
4968 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4969 databases so that it will be absolutely obvious if a crash occurs in the
4970 DB library. This is a regular occurrence (often caused by mis-matched
4973 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4974 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4975 picked up on a user's system that detects such things. There doesn't seem
4976 to be a gcc warning option for this - only an attribute that has to be
4977 put on the function's prototype. It seems that in Fedora Core 4 they have
4978 set this on a number of new functions. No doubt there will be more in due
4981 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4982 specify a transport. However, if an address that was verified by such a
4983 router was the subject of a callout, Exim crashed because it tried to
4984 read the rcpt_include_affixes from the non-existent transport. Now it
4985 just assumes that the setting of that option is false. This bug was
4986 introduced by 4.51/PH/31.
4988 PH/28 Changed -d+all to exclude +memory, because that information is very
4989 rarely of interest, but it makes the output a lot bigger. People tend to
4990 do -d+all out of habit.
4992 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4993 code in os-type was giving problems when libc.so lives in lib64, like on
4996 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4997 aren't the modern standard, and it seems that some systems' include files
4998 don't always have them. Exim was already checking for some of the newer
4999 ones like T_AAAA, and defining it itself. I've added checks for all the
5000 record types that Exim uses.
5002 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5003 not automatically generating a new one, as it is supposed to. This
5004 prevented TLS from working. If the file did exist, but contained invalid
5005 data, a new version was generated, as expected. It was only the case of a
5006 non-existent file that was broken.
5008 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5009 with a change in libdomainkeys > 0.64.
5011 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5012 from DNS. If the selector record carries the flag, it now has
5013 precedence over the domain-wide flag.
5015 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5017 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5018 the use of a port name that isn't defined in /etc/services) occurred, the
5019 message was deferred as in a normal delivery, and thus remained on the
5020 spool, instead of being failed because of the mua_wrapper setting. This
5021 is now fixed, and I tidied up some of the mua_wrapper messages at the
5024 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5025 the messages in a hash of arrays rather than using individual hashes.
5026 This is a bit cleaner and results in dramatic memory savings, albeit
5027 at a slight CPU cost.
5029 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5030 as requested by Marc Sherman.
5032 SC/10 Eximstats: added histograms for user specified patterns as requested
5035 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5037 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5038 fopen() in the content-scanning modules that did not already have it.
5044 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5045 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5047 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5049 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5051 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5052 to test Sieve filters that use "vacation".
5054 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5055 that changes the way the GnuTLS parameters are stored in the cache file.
5056 The new format can be generated externally. For backward compatibility,
5057 if the data in the cache doesn't make sense, Exim assumes it has read an
5058 old-format file, and it generates new data and writes a new file. This
5059 means that you can't go back to an older release without removing the
5062 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5063 work if there are any delivery delays because "one_time" forces the
5064 parent to be marked "delivered", so its unseen clone is never tried
5065 again. For this reason, Exim now forbids the simultaneous setting of
5068 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5069 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5070 entry if you want to know the details. The fix, however, neglected to
5071 consider the case when local delivery batching is involved. The test for
5072 "previously delivered" was not happening when checking to see if an
5073 address could be batched with a previous (undelivered) one; under
5074 certain circumstances this could lead to multiple deliveries to the same
5077 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5078 in its include files, and this causes problems building Exim.
5080 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5081 header_syntax) but Exim was just ignoring anything given after a slash.
5082 In particular, this caused confusion with an attempt to use "verify =
5083 reverse_host_lookup/defer_ok". An error is now given when options are
5084 supplied for verify items that do not have them. (Maybe reverse_host_
5085 lookup should have a defer_ok option, but that's a different point.)
5087 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5088 defined by RFC 821) to 2048, because there were problems with some AUTH
5089 commands, and RFC 1869 says the size should be increased for extended
5090 SMTP commands that take arguments.
5092 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5095 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5096 "unknown" error; now it says that the functionality isn't in the binary.
5098 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5099 an address' error message when a string expansion fails (syntax or
5100 whatever). Otherwise the password may appear in the log. Following change
5101 PH/42 below, there is no longer a chance of it appearing in a bounce
5104 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5106 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5107 of its addresses was ever tried. (Bugzilla bug #2.)
5109 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5110 the result incorrectly in the debug output. (It correctly added a newline
5111 to what was transported.)
5113 TF/01 Added $received_time.
5115 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5116 commented out examples of how to interface to a virus scanner and to
5117 SpamAssassin. Also added commented examples of av_scanner and
5118 spamd_address settings.
5120 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5121 and controls are allowed in which ACLs. There were a couple of minor
5122 errors. Some of the entries in the conditions table (which is a table of
5123 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5124 negation of where the condition IS allowed.
5126 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5128 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5129 header file does not have a version number, so I've had to invent a new
5130 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5131 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5132 radiusclient), but it was contributed by a Radius user.
5134 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5135 files or directories, for OpenSSL.
5137 PH/19 When an Exim process that is running as root has to create an Exim log
5138 file, it does so in a subprocess that runs as exim:exim so as to get the
5139 ownership right at creation (otherwise, other Exim processes might see
5140 the file with the wrong ownership). There was no test for failure of this
5141 fork() call, which would lead to the process getting stuck as it waited
5142 for a non-existent subprocess. Forks do occasionally fail when resources
5143 run out. I reviewed all the other calls to fork(); they all seem to check
5146 PH/20 When checking for unexpected SMTP input at connect time (before writing
5147 the banner), Exim was not dealing correctly with a non-positive return
5148 from the read() function. If the client had disconnected by this time,
5149 the result was a log entry for a synchronization error with an empty
5150 string after "input=" when read() returned zero. If read() returned -1
5151 (an event I could not check), uninitialized data bytes were printed.
5152 There were reports of junk text (parts of files, etc) appearing after
5155 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5157 PH/22 Added support for macro redefinition, and (re)definition in between
5158 driver and ACL definitions.
5160 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5161 forgetting to use the resulting value; it was using the unexpanded value.
5163 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5164 hadn't been configured. The fix is from Juergen Kreileder, who
5165 understands it better than I do:
5167 "Here's what I see happening with three configured cyrus_sasl
5168 authenticators configured (plain, login, cram-md5):
5170 On startup auth_cyrus_sasl_init() gets called for each of these.
5171 This means three calls to sasl_listmech() without a specified mech_list.
5172 => SASL tests which mechs of all available mechs actually work
5173 => three warnings about OTP not working
5174 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5176 With the patch, sasl_listmech() also gets called three times. But now
5177 SASL's mech_list option is set to the server_mech specified in the the
5178 authenticator. Or in other words, the answer from sasl_listmech()
5179 gets limited to just the mech you're testing for (which is different
5181 => the return list contains just 'plain' or 'login', 'cram-md5' or
5182 nothing depending on the value of ob->server_mech.
5184 I've just tested the patch: Authentication still works fine,
5185 unavailable mechs specified in the exim configuration are still
5186 caught, and the auth.log warnings about OTP are gone."
5188 PH/25 When debugging is enabled, the contents of the command line are added
5189 to the debugging output, even when log_selector=+arguments is not
5192 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5193 answer is "GNU", and only if the return is "GNU/something" is the answer
5196 PH/27 $acl_verify_message is now set immediately after the failure of a
5197 verification in an ACL, and so is available in subsequent modifiers. In
5198 particular, the message can be preserved by coding like this:
5200 warn !verify = sender
5201 set acl_m0 = $acl_verify_message
5203 Previously, $acl_verify_message was set only while expanding "message"
5204 and "log_message" when a very denied access.
5206 PH/28 Modified OS/os.c-Linux with
5208 -#ifndef OS_LOAD_AVERAGE
5209 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5211 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5212 nomenclature these days.)
5214 PH/29 Installed patch from the Sieve maintainer that adds the options
5215 sieve_useraddress and sieve_subaddress to the redirect router.
5217 PH/30 In these circumstances:
5218 . Two addresses routed to the same list of hosts;
5219 . First host does not offer TLS;
5220 . First host accepts first address;
5221 . First host gives temporary error to second address;
5222 . Second host offers TLS and a TLS session is established;
5223 . Second host accepts second address.
5224 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5225 and peerdn, if requested) that were in fact used only for the second
5228 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5229 attention to any local part prefix or suffix that was matched by the
5230 router that accepted the address. It now behaves in the same way as it
5231 does for delivery: the affixes are removed from the local part unless
5232 rcpt_include_affixes is set on the transport.
5234 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5235 timeout during the DATA phase of an incoming message.
5237 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5238 applied a patch sanctioned by the Sieve maintainer.
5240 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5241 the uid or gid is negative. A case of a negative gid caused this to be
5242 noticed. The fix allows for either to be negative.
5244 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5245 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5248 PH/36 The change PH/12 above was broken. Fixed it.
5250 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5251 the grounds that routing the same address twice would always produce the
5252 same answer. This might have been true once, but it is certainly no
5253 longer true now. Routing a child address may depend on the previous
5254 routing that produced that child. Some complicated redirection strategies
5255 went wrong when messages had multiple recipients, and made Exim's
5256 behaviour dependent on the order in which the addresses were given.
5258 I have moved the duplicate checking until after the routing is complete.
5259 Exim scans the addresses that are assigned to local and remote
5260 transports, and removes any duplicates. This means that more work will be
5261 done, as duplicates will always all be routed, but duplicates are
5262 presumably rare, so I don't expect this is of any significance.
5264 For deliveries to pipes, files, and autoreplies, the duplicate checking
5265 still happens during the routing process, since they are not going to be
5268 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5269 It corrects a timeout issue with spamd. This is Ian's comment: "The
5270 background is that sometimes spamd either never reads data from a
5271 connection it has accepted, or it never writes response data. The exiscan
5272 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5273 blindly assumes that writes won't block so it may never time out."
5275 PH/39 Allow G after quota size as well as K and M.
5277 PH/40 The value set for $authenticated_id in an authenticator may not contain
5278 binary zeroes or newlines because the value is written to log lines and
5279 to spool files. There was no check on this. Now the value is run through
5280 the string_printing() function so that such characters are converted to
5281 printable escape sequences.
5283 PH/41 $message_linecount is a new variable that contains the total number of
5284 lines in the message. Compare $body_linecount, which is the count for the
5287 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5288 bounce and delay warning messages, except in certain special cases, which
5291 (a) An SMTP error message from a remote host;
5292 (b) A message specified in a :fail: redirection;
5293 (c) A message specified in a "fail" command in a system filter;
5294 (d) A message specified in a FAIL return from the queryprogram router;
5295 (e) A message specified by the cannot_route_message router option.
5297 In these cases only, Exim does include the error details in bounce and
5298 warning messages. There are also a few cases where bland messages such
5299 as "unrouteable address" or "local delivery error" are given.
5301 PH/43 $value is now also set for the "else" part of a ${run expansion.
5303 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5304 being worked on, but at least Exim now implements the latest version to
5307 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5308 process to complete was treated as a delivery failure, a timeout while
5309 writing the message to the pipe was logged, but erroneously treated as a
5310 successful delivery. Such timeouts include transport filter timeouts. For
5311 consistency with the overall process timeout, these timeouts are now
5312 treated as errors, giving rise to delivery failures by default. However,
5313 there is now a new Boolean option for the pipe transport called
5314 timeout_defer, which, if set TRUE, converts the failures into defers for
5315 both kinds of timeout. A transport filter timeout is now identified in
5318 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5319 systems where "make" and "gmake" are different, calling "gmake" at top
5320 level broke things. I've arranged for the value of $(MAKE) to be passed
5321 from the Makefile to this script so that it can call the same version of
5325 A note about Exim versions 4.44 and 4.50
5326 ----------------------------------------
5328 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5329 changes of various kinds. As a consequence, a big documentation update was
5330 needed. This delayed the release for rather longer than seemed good, especially
5331 in the light of a couple of (minor) security issues. Therefore, the changes
5332 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5333 release. So 4.44 and 4.50 are in effect two different branches that both start
5336 I have left the 4.50 change log unchanged; it contains all the changes since
5337 4.43. The change log for 4.44 is below; many of its items are identical to
5338 those for 4.50. This seems to be the most sensible way to preserve the
5339 historical information.
5345 1. Minor wording change to the doc/README.SIEVE file.
5347 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5348 computation of the current number of files was incorrect.
5350 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5351 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5354 4. Give more explanation in the error message when the command for a transport
5355 filter fails to execute.
5357 5. There are several places where Exim runs a non-Exim command in a
5358 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5359 was being done only for the command run by the queryprogram router. It is
5360 now done for all such subprocesses. The other cases are: ${run, transport
5361 filters, and the commands run by the lmtp and pipe transports.
5363 6. Added CONFIGURE_GROUP build-time option.
5365 7. Some older OS have a limit of 256 on the maximum number of file
5366 descriptors. Exim was using setrlimit() to set 1000 as a large value
5367 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5368 systems. I've change it so that if it can't get 1000, it tries for 256.
5370 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5371 was an oversight, and furthermore, ever since the addition of extra
5372 controls (e.g. 4.43/32), the checks on when to allow different forms of
5373 "control" were broken. There should now be diagnostics for all cases when a
5374 control that does not make sense is encountered.
5376 9. Added the /retain_sender option to "control=submission".
5378 10. $recipients is now available in the predata ACL (oversight).
5380 11. Tidy the search cache before the fork to do a delivery from a message
5381 received from the command line. Otherwise the child will trigger a lookup
5382 failure and thereby defer the delivery if it tries to use (for example) a
5383 cached ldap connection that the parent has called unbind on.
5385 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5386 of $address_data from the recipient verification was clobbered by the
5387 sender verification.
5389 13. The value of address_data from a sender verification is now available in
5390 $sender_address_data in subsequent conditions in the ACL statement.
5392 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5394 15. Added a new option "connect=<time>" to callout options, to set a different
5397 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5398 was its contents. (It was OK if the option was not defined at all.)
5400 17. A "Completed" log line is now written for messages that are removed from
5401 the spool by the -Mrm option.
5403 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5404 information about exactly what failed.
5406 19. Added -dd to debug only the daemon process.
5408 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5409 handles timeouts, both on the server side and network timeouts. Renamed the
5410 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5412 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5413 It is now set to "smtps".
5415 22. $host_address is now set to the target address during the checking of
5416 ignore_target_hosts.
5418 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5419 being passed; this would have caused $sender_host_name to have been used if
5420 matching the list had actually called for a host name (not very likely,
5421 since this list is usually IP addresses). A host name is now passed as
5424 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5425 code that specifies a non-restarting handler (typically sigaction() in
5426 modern systems) in an attempt to fix a rare and obscure crash bug.
5428 25. Narrowed the window for a race in the daemon that could cause it to ignore
5429 SIGCHLD signals. This is not a major problem, because they are used only to
5430 wake it up if nothing else does.
5432 26. A malformed maildirsize file could cause Exim to calculate negative values
5433 for the mailbox size or file count. Odd effects could occur as a result.
5434 The maildirsize information is now recalculated if the size or filecount
5437 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5438 support for a long time. Removed HAVE_SYS_VFS_H.
5440 28. Installed the latest version of exipick from John Jetmore.
5442 29. In an address list, if the pattern was not a regular expression, an empty
5443 subject address (from a bounce message) matched only if the pattern was an
5444 empty string. Non-empty patterns were not even tested. This was the wrong
5445 because it is perfectly reasonable to use an empty address as part of a
5446 database query. An empty address is now tested by patterns that are
5447 lookups. However, all the other forms of pattern expect the subject to
5448 contain a local part and a domain, and therefore, for them, an empty
5449 address still always fails if the pattern is not itself empty.
5451 30. Exim went into a mad DNS loop when attempting to do a callout where the
5452 host was specified on an smtp transport, and looking it up yielded more
5453 than one IP address.
5455 31. Re-factored the code for checking spool and log partition space into a
5456 function that finds that data and another that does the check. The former
5457 is then used to implement four new variables: $spool_space, $log_space,
5458 $spool_inodes, and $log_inodes.
5460 32. The RFC2047 encoding function was originally intended for short strings
5461 such as real names; it was not keeping to the 75-character limit for
5462 encoded words that the RFC imposes. It now respects the limit, and
5463 generates multiple encoded words if necessary. To be on the safe side, I
5464 have increased the buffer size for the ${rfc2047: expansion operator from
5467 33. It is now permitted to omit both strings after an "if" condition; if the
5468 condition is true, the result is "true". As before, when the second string
5469 is omitted, a false condition yields an empty string. This makes it less
5470 cumbersome to write custom ACL and router conditions.
5472 34. Failure to deliver a bounce message always caused it to be frozen, even if
5473 there was an errors_to setting on the router. The errors_to setting is now
5476 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5477 canonical form (fully expanded) before being placed in
5478 $sender_host_address.
5480 36. The table in the code that translates DNS record types into text (T_A to
5481 "A" for instance) was missing entries for NS and CNAME. It is just possible
5482 that this could have caused confusion if both these types were looked up
5483 for the same domain, because the text type is used as part of Exim's
5484 per-process caching. But the chance of anyone hitting this buglet seems
5487 37. The dnsdb lookup has been extended in a number of ways.
5489 (1) There is a new type, "zns", which walks up the domain tree until it
5490 finds some nameserver records. It should be used with care.
5492 (2) There is a new type, "mxh", which is like "mx" except that it returns
5493 just the host names, not the priorities.
5495 (3) It is now possible to give a list of domains (or IP addresses) to be
5496 looked up. The behaviour when one of the lookups defers can be
5497 controlled by a keyword.
5499 (4) It is now possible to specify the separator character for use when
5500 multiple records are returned.
5502 38. The dnslists ACL condition has been extended: it is now possible to supply
5503 a list of IP addresses and/or domains to be looked up in a particular DNS
5506 39. Added log_selector=+queue_time_overall.
5508 40. When running the queue in the test harness, wait just a tad after forking a
5509 delivery process, to get repeatability of debugging output.
5511 41. Include certificate and key file names in error message when GnuTLS fails
5512 to set them up, because the GnuTLS error message doesn't include the name
5513 of the failing file when there is a problem reading it.
5515 42. Allow both -bf and -bF in the same test run.
5517 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5519 44. The "Exiscan patch" is now merged into the mainline Exim source.
5521 45. Sometimes the final signoff response after QUIT could fail to get
5522 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5523 < 0 before doing a fflush(). This bug looks as though it goes back to the
5524 introduction of TLS in release 3.20, but "sometimes" must have been rare
5525 because the tests only now provoked it.
5527 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5528 (this can affect the format of dates).
5530 47. exim_tidydb, when checking for the continued existence of a message for
5531 which it has found a message-specific retry record, was not finding
5532 messages that were in split spool directories. Consequently, it was
5533 deleting retry records that should have stayed in existence.
5535 48. Steve fixed some bugs in eximstats.
5537 49. The SPA authentication driver was not abandoning authentication and moving
5538 on to the next authenticator when an expansion was forced to fail,
5539 contradicting the general specification for all authenticators. Instead it
5540 was generating a temporary error. It now behaves as specified.
5542 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5543 (the order specifies the preference for clients). The order is now AES256,
5544 AES128, 3DES, ARCFOUR128.
5546 51. Small patch to Sieve code - explicitly set From: when generating an
5549 52. Exim crashed if a remote delivery caused a very long error message to be
5550 recorded - for instance if somebody sent an entire SpamAssassin report back
5551 as a large number of 550 error lines. This bug was coincidentally fixed by
5552 increasing the size of one of Exim's internal buffers (big_buffer) that
5553 happened as part of the Exiscan merge. However, to be on the safe side, I
5554 have made the code more robust (and fixed the comments that describe what
5557 53. Now that there can be additional text after "Completed" in log lines (if
5558 the queue_time_overall log selector is set), a one-byte patch to exigrep
5559 was needed to allow it to recognize "Completed" as not the last thing in
5562 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5563 patch that reportedly fixes this has been added. I am not expert enough to
5564 create a test for it. This is what the patch creator wrote:
5566 "I found a little strange behaviour of ldap code when working with
5567 Windows 2003 AD Domain, where users was placed in more than one
5568 Organization Units. When I tried to give exim partial DN, the exit code
5569 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5570 But simultaneously result of request was absolutely normal ldap result,
5571 so I produce this patch..."
5573 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5574 REFERENCE, so I have modified the code to exclude the patch when that macro
5577 55. Some experimental protocols are using DNS PTR records for new purposes. The
5578 keys for these records are domain names, not reversed IP addresses. The
5579 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5580 leaves it alone. Component reversal etc. now happens only for IP addresses.
5583 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5585 57. Double the size of the debug message buffer (to 2048) so that more of very
5586 long debug lines gets shown.
5588 58. The exicyclog utility now does better if the number of log files to keep
5589 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5591 59. Two changes related to the smtp_active_hostname option:
5593 (1) $smtp_active_hostname is now available as a variable.
5594 (2) The default for smtp_banner uses $smtp_active_hostname instead
5595 of $primary_hostname.
5597 60. The host_aton() function is supposed to be passed a string that is known
5598 to be a valid IP address. However, in the case of IPv6 addresses, it was
5599 not checking this. This is a hostage to fortune. Exim now panics and dies
5600 if the condition is not met. A case was found where this could be provoked
5601 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5602 components; fortuitously, this particular loophole had already been fixed
5603 by change 4.50/55 above.
5605 If there are any other similar loopholes, the new check in host_aton()
5606 itself should stop them being exploited. The report I received stated that
5607 data on the command line could provoke the exploit when Exim was running as
5608 exim, but did not say which command line option was involved. All I could
5609 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5610 running as the user.
5613 61. There was a buffer overflow vulnerability in the SPA authentication code
5614 (which came originally from the Samba project). I have added a test to the
5615 spa_base64_to_bits() function which I hope fixes it.
5618 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5619 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5621 63. The daemon start-up calls getloadavg() while still root for those OS that
5622 need the first call to be done as root, but it missed one case: when
5623 deliver_queue_load_max is set with deliver_drop_privilege. This is
5624 necessary for the benefit of the queue runner, because there is no re-exec
5625 when deliver_drop_privilege is set.
5627 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5628 This has been fixed.
5630 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5631 was in use, was not putting the data itself into the right store pool;
5632 consequently, it could be overwritten for a subsequent message in the same
5633 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5636 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5638 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5639 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5640 the function were treating the return as a boolean value, which happened to
5641 work because 0=false and not-0=true, but is not correct code.
5643 68. The host_aton() function was not handling scoped IPv6 addresses (those
5644 with, for example, "%eth0" on the end) correctly.
5646 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5647 negated items (that is, ~something) in unsigned ints. Some compilers
5648 apparently mutter when there is no cast.
5650 70. If an address verification called from an ACL failed, and did not produce a
5651 user-specific message (i.e. there was only a "system" message), nothing was
5652 put in $acl_verify_message. In this situation, it now puts the system
5655 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5656 session; change 4.31/43 added the unwanted input to the log line - except
5657 that it did not do this in the start of session case. It now does.
5659 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5660 This is wrong and can cause the other end to generate a synchronization
5661 error if it is another Exim or anything else that does the synchronization
5662 check. A QUIT command is no longer sent after a timeout.
5664 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5665 during host lookups.
5667 74. The defer_ok option of callout verification was not working if it was used
5668 when verifying addresses in header lines, that is, for this case:
5670 verify = header_sender/callout=defer_ok
5672 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5673 those file descriptors could be used for SMTP connections. If anything
5674 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5675 could be sent to the SMTP client, causing chaos. The daemon now opens
5676 stdin, stdout, and stderr to /dev/null when it puts itself into the
5679 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5680 log by default. The user can override this with suitable Perl magic.
5682 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5683 the log message when discard triggers, was not working for the DATA ACL or
5684 for the non-SMTP ACL.
5686 78. Error message wording change in sieve.c.
5688 79. If smtp_accept_max_per_host was set, the number of connections could be
5689 restricted to fewer than expected, because the daemon was trying to set up
5690 a new connection before checking whether the processes handling previous
5691 connections had finished. The check for completed processes is now done
5692 earlier. On busy systems, this bug wouldn't be noticed because something
5693 else would have woken the daemon, and it would have reaped the completed
5696 80. If a message was submitted locally by a user whose login name contained one
5697 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5698 It caused a spool format error. I have fixed the spool reading code. A
5699 related problem was that the "from" clause in the Received: line became
5700 illegal because of the space(s). It is now covered by ${quote_local_part.
5702 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5705 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5707 83. Patch from Sieve maintainer to fix unterminated string problem in
5708 "vacation" handling.
5710 84. Some minor changes to the Linux configuration files to help with other
5711 OS variants using glibc.
5713 85. One more patch for Sieve to update vacation handling to latest spec.
5716 ----------------------------------------------------
5717 See the note above about the 4.44 and 4.50 releases.
5718 ----------------------------------------------------
5724 1. Change 4.43/35 introduced a bug that caused file counts to be
5725 incorrectly computed when quota_filecount was set in an appendfile
5728 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5729 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5732 3. Give more explanation in the error message when the command for a transport
5733 filter fails to execute.
5735 4. There are several places where Exim runs a non-Exim command in a
5736 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5737 was being done only for the command run by the queryprogram router. It is
5738 now done for all such subprocesses. The other cases are: ${run, transport
5739 filters, and the commands run by the lmtp and pipe transports.
5741 5. Some older OS have a limit of 256 on the maximum number of file
5742 descriptors. Exim was using setrlimit() to set 1000 as a large value
5743 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5744 systems. I've change it so that if it can't get 1000, it tries for 256.
5746 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5747 was an oversight, and furthermore, ever since the addition of extra
5748 controls (e.g. 4.43/32), the checks on when to allow different forms of
5749 "control" were broken. There should now be diagnostics for all cases when a
5750 control that does not make sense is encountered.
5752 7. $recipients is now available in the predata ACL (oversight).
5754 8. Tidy the search cache before the fork to do a delivery from a message
5755 received from the command line. Otherwise the child will trigger a lookup
5756 failure and thereby defer the delivery if it tries to use (for example) a
5757 cached ldap connection that the parent has called unbind on.
5759 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5760 of $address_data from the recipient verification was clobbered by the
5761 sender verification.
5763 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5764 was its contents. (It was OK if the option was not defined at all.)
5766 11. A "Completed" log line is now written for messages that are removed from
5767 the spool by the -Mrm option.
5769 12. $host_address is now set to the target address during the checking of
5770 ignore_target_hosts.
5772 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5773 being passed; this would have caused $sender_host_name to have been used if
5774 matching the list had actually called for a host name (not very likely,
5775 since this list is usually IP addresses). A host name is now passed as
5778 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5779 code that specifies a non-restarting handler (typically sigaction() in
5780 modern systems) in an attempt to fix a rare and obscure crash bug.
5782 15. Narrowed the window for a race in the daemon that could cause it to ignore
5783 SIGCHLD signals. This is not a major problem, because they are used only to
5784 wake it up if nothing else does.
5786 16. A malformed maildirsize file could cause Exim to calculate negative values
5787 for the mailbox size or file count. Odd effects could occur as a result.
5788 The maildirsize information is now recalculated if the size or filecount
5791 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5792 support for a long time. Removed HAVE_SYS_VFS_H.
5794 18. Updated exipick to current release from John Jetmore.
5796 19. Allow an empty sender to be matched against a lookup in an address list.
5797 Previously the only cases considered were a regular expression, or an
5800 20. Exim went into a mad DNS lookup loop when doing a callout where the
5801 host was specified on the transport, if the DNS lookup yielded more than
5804 21. The RFC2047 encoding function was originally intended for short strings
5805 such as real names; it was not keeping to the 75-character limit for
5806 encoded words that the RFC imposes. It now respects the limit, and
5807 generates multiple encoded words if necessary. To be on the safe side, I
5808 have increased the buffer size for the ${rfc2047: expansion operator from
5811 22. Failure to deliver a bounce message always caused it to be frozen, even if
5812 there was an errors_to setting on the router. The errors_to setting is now
5815 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5816 canonical form (fully expanded) before being placed in
5817 $sender_host_address.
5819 24. Updated eximstats to version 1.33
5821 25. Include certificate and key file names in error message when GnuTLS fails
5822 to set them up, because the GnuTLS error message doesn't include the name
5823 of the failing file when there is a problem reading it.
5825 26. Expand error message when OpenSSL has problems setting up cert/key files.
5828 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5829 (this can affect the format of dates).
5831 28. exim_tidydb, when checking for the continued existence of a message for
5832 which it has found a message-specific retry record, was not finding
5833 messages that were in split spool directories. Consequently, it was
5834 deleting retry records that should have stayed in existence.
5836 29. eximstats updated to version 1.35
5837 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5838 1.35 - bugfix such that pie charts by volume are generated correctly
5840 30. The SPA authentication driver was not abandoning authentication and moving
5841 on to the next authenticator when an expansion was forced to fail,
5842 contradicting the general specification for all authenticators. Instead it
5843 was generating a temporary error. It now behaves as specified.
5845 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5846 (the order specifies the preference for clients). The order is now AES256,
5847 AES128, 3DES, ARCFOUR128.
5849 31. Small patch to Sieve code - explicitly set From: when generating an
5852 32. Exim crashed if a remote delivery caused a very long error message to be
5853 recorded - for instance if somebody sent an entire SpamAssassin report back
5854 as a large number of 550 error lines. This bug was coincidentally fixed by
5855 increasing the size of one of Exim's internal buffers (big_buffer) that
5856 happened as part of the Exiscan merge. However, to be on the safe side, I
5857 have made the code more robust (and fixed the comments that describe what
5860 33. Some experimental protocols are using DNS PTR records for new purposes. The
5861 keys for these records are domain names, not reversed IP addresses. The
5862 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5863 leaves it alone. Component reversal etc. now happens only for IP addresses.
5866 34. The host_aton() function is supposed to be passed a string that is known
5867 to be a valid IP address. However, in the case of IPv6 addresses, it was
5868 not checking this. This is a hostage to fortune. Exim now panics and dies
5869 if the condition is not met. A case was found where this could be provoked
5870 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5871 components; fortuitously, this particular loophole had already been fixed
5872 by change 4.50/55 or 4.44/33 above.
5874 If there are any other similar loopholes, the new check in host_aton()
5875 itself should stop them being exploited. The report I received stated that
5876 data on the command line could provoke the exploit when Exim was running as
5877 exim, but did not say which command line option was involved. All I could
5878 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5879 running as the user.
5882 35. There was a buffer overflow vulnerability in the SPA authentication code
5883 (which came originally from the Samba project). I have added a test to the
5884 spa_base64_to_bits() function which I hope fixes it.
5887 36. The daemon start-up calls getloadavg() while still root for those OS that
5888 need the first call to be done as root, but it missed one case: when
5889 deliver_queue_load_max is set with deliver_drop_privilege. This is
5890 necessary for the benefit of the queue runner, because there is no re-exec
5891 when deliver_drop_privilege is set.
5893 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5894 was in use, was not putting the data itself into the right store pool;
5895 consequently, it could be overwritten for a subsequent message in the same
5896 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5899 38. Sometimes the final signoff response after QUIT could fail to get
5900 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5901 < 0 before doing a fflush(). This bug looks as though it goes back to the
5902 introduction of TLS in release 3.20, but "sometimes" must have been rare
5903 because the tests only now provoked it.
5909 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5910 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5911 Now it returns an integer. A number of calls were still expecting a T/F
5912 return. Fortuitously, in all cases, the tests worked in OK situations,
5913 which is the norm. However, things would have gone wrong on any write
5914 failures on the smtp file descriptor. This function is used when sending
5915 messages over SMTP and also when doing verify callouts.
5917 2. When Exim is called to do synchronous delivery of a locally submitted
5918 message (the -odf or -odi options), it no longer closes stderr before doing
5921 3. Implemented the mua_wrapper option.
5923 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5925 5. Implemented the functions header_remove(), header_testname(),
5926 header_add_at_position(), and receive_remove_recipient(), and exported them
5929 6. If an ACL "warn" statement specified the addition of headers, Exim already
5930 inserted X-ACL-Warn: at the start if there was no header name. However, it
5931 was not making this test for the second and subsequent header lines if
5932 there were newlines in the string. This meant that an invalid header could
5933 be inserted if Exim was badly configured.
5935 7. Allow an ACL "warn" statement to add header lines at the start or after all
5936 the Received: headers, as well as at the end.
5938 8. Added the rcpt_4xx retry error code.
5940 9. Added postmaster_mailfrom=xxx to callout verification option.
5942 10. Added mailfrom=xxxx to the callout verification option, for verify=
5945 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5946 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5948 12. Inserted some casts to stop certain compilers warning when using pointer
5949 differences as field lengths or precisions in printf-type calls (mostly
5950 affecting debugging statements).
5952 13. Added optional readline() support for -be (dynamically loaded).
5954 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5955 same clock tick as a message's arrival, so that its received time was the
5956 same as the "first fail" time on the retry record, and that message
5957 remained on the queue past the ultimate address timeout, every queue runner
5958 would try a delivery (because it was past the ultimate address timeout) but
5959 after another failure, the ultimate address timeout, which should have then
5960 bounced the address, did not kick in. This was a "< instead of <=" error;
5961 in most cases the first failure would have been in the next clock tick
5962 after the received time, and all would be well.
5964 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5965 being recognized when the domain list was tested by the match_domain
5966 condition in an expansion string.
5968 16. Added the ${str2b64: operator.
5970 17. Exim was always calling setrlimit() to set a large limit for the number of
5971 processes, without checking whether the existing limit was already
5972 adequate. (It did check for the limit on file descriptors.) Furthermore,
5973 errors from getrlimit() and setrlimit() were being ignored. Now they are
5974 logged to the main and panic logs, but Exim does carry on, to try to do its
5975 job under whatever limits there are.
5977 18. Imported PCRE 5.0.
5979 19. Trivial typo in log message " temporarily refused connection" (the leading
5982 20. If the log selector return_path_on_delivery was set and an address was
5983 redirected to /dev/null, the delivery process crashed because it assumed
5984 that a return path would always be set for a "successful" delivery. In this
5985 case, the whole delivery is bypassed as an optimization, and therefore no
5988 21. Internal re-arrangement: the function for sending a challenge and reading
5989 a response while authentication was assuming a zero-terminated challenge
5990 string. It's now changed to take a pointer and a length, to allow for
5991 binary data in such strings.
5993 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5995 23. Exim was not respecting finduser_retries when seeking the login of the
5996 uid under which it was called; it was always trying 10 times. (The default
5997 setting of finduser_retries is zero.) Also, it was sleeping after the final
5998 failure, which is pointless.
6000 24. Implemented tls_on_connect_ports.
6002 25. Implemented acl_smtp_predata.
6004 26. If the domain in control=submission is set empty, Exim assumes that the
6005 authenticated id is a complete email address when it generates From: or
6006 Sender: header lines.
6008 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6009 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6010 chown and chgrp in /bin and hostname in /usr/bin.
6012 28. Exim was keeping the "process log" file open after each use, just as it
6013 does for the main log. This opens the possibility of it remaining open for
6014 long periods when the USR1 signal hits a daemon. Occasional processlog
6015 errors were reported, that could have been caused by this. Anyway, it seems
6016 much more sensible not to leave this file open at all, so that is what now
6019 29. The long-running daemon process does not normally write to the log once it
6020 has entered its main loop, and it closes the log before doing so. This is
6021 so that log files can straightforwardly be renamed and moved. However,
6022 there are a couple of unusual error situations where the daemon does write
6023 log entries, and I had neglected to close the log afterwards.
6025 30. The text of an SMTP error response that was received during a remote
6026 delivery was being truncated at 512 bytes. This is too short for some of
6027 the long messages that one sometimes sees. I've increased the limit to
6030 31. It is now possible to make retry rules that apply only when a message has a
6031 specific sender, in particular, an empty sender.
6033 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6034 it possible to be selective about when SMTP synchronization is enforced.
6036 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6038 32. Implemented hosts_connection_nolog.
6040 33. Added an ACL for QUIT.
6042 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6045 35. Added mailbox_size and mailbox_filecount to appendfile.
6047 36. Added control = no_multiline_responses to ACLs.
6049 37. There was a bug in the logic of the code that waits for the clock to tick
6050 in the case where the clock went backwards by a substantial amount such
6051 that the microsecond fraction of "now" was more than the microsecond
6052 fraction of "then" (but the whole seconds number was less).
6054 38. Added support for the libradius Radius client library this is found on
6055 FreeBSD (previously only the radiusclient library was supported).
6061 1. When certain lookups returned multiple values in the form name=value, the
6062 quoting of the values was not always being done properly. Specifically:
6063 (a) If the value started with a double quote, but contained no whitespace,
6065 (b) If the value contained whitespace other than a space character (i.e.
6066 tabs or newlines or carriage returns) it was not quoted.
6067 This fix has been applied to the mysql and pgsql lookups by writing a
6068 separate quoting function and calling it from the lookup code. The fix
6069 should probably also be applied to nisplus, ibase and oracle lookups, but
6070 since I cannot test any of those, I have not disturbed their existing code.
6072 2. A hit in the callout cache for a specific address caused a log line with no
6073 reason for rejecting RCPT. Now it says "Previous (cached) callout
6074 verification failure".
6076 3. There was an off-by-one bug in the queryprogram router. An over-long
6077 return line was truncated at 256 instead of 255 characters, thereby
6078 overflowing its buffer with the terminating zero. As well as fixing this, I
6079 have increased the buffer size to 1024 (and made a note to document this).
6081 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6082 when Exim is waiting for an SMTP response from a remote server, Exim
6083 restarts its select() call on the socket, thereby resetting its timeout.
6084 This is not a problem when such interrupts are rare. Somebody set up a cron
6085 job to run exiwhat every 2 minutes, which is less than the normal select()
6086 timeout (5 or 10 minutes). This meant that the select() timeout never
6087 kicked in because it was always reset. I have fixed this by comparing the
6088 time when an interrupt arrives with the time at the start of the first call
6089 to select(). If more time than the timeout has elapsed, the interrupt is
6090 treated as a timeout.
6092 5. Some internal re-factoring in preparation for the addition of Sieve
6093 extensions (by MH). In particular, the "personal" test is moved to a
6094 separate function, and given an option for scanning Cc: and Bcc: (which is
6095 not set for Exim filters).
6097 6. When Exim created an email address using the login of the caller as the
6098 local part (e.g. when creating a From: or Sender: header line), it was not
6099 quoting the local part when it contained special characters such as @.
6101 7. Installed new OpenBSD configuration files.
6103 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6104 try to make them clearer.
6106 9. Callout options, other than the timeout value, were being ignored when
6107 verifying sender addresses in header lines. For example, when using
6109 verify = header_sender/callout=no_cache
6111 the cache was (incorrectly) being used.
6113 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6114 only the Cygwin environment.
6116 11. When return_path_on_delivery was set as a log selector, if different remote
6117 addresses in the same message used different return paths and parallel
6118 remote delivery occurred, the wrong values would sometimes be logged.
6119 (Whenever a remote delivery process finished, the return path value from
6120 the most recently started remote delivery process was logged.)
6122 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6123 lines when AUTH and/or TLS are in use. This is the "received protocol"
6124 field. Exim used to use "asmtp" for authenticated SMTP, without any
6125 indication (in the protocol name) for TLS use. Now it follows the RFC and
6126 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6127 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6128 names appear in log lines as well as in Received: header lines.
6130 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6131 extensions, and comparison tests, and to fix some bugs.
6133 14. Changes to the "personal" filter test:
6135 (1) The test was buggy in that it was just doing the equivalent of
6136 "contains" tests on header lines. For example, if a user's address was
6137 anne@some.where, the "personal" test would incorrectly be true for
6139 To: susanne@some.where
6141 This test is now done by extracting each address from the header in turn,
6142 and checking the entire address. Other tests that are part of "personal"
6143 are now done using regular expressions (for example, to check local parts
6144 of addresses in From: header lines).
6146 (2) The list of non-personal local parts in From: addresses has been
6147 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6148 taken from the Sieve specification recommendations.
6150 (3) If the message contains any header line starting with "List-" it is
6151 treated as non-personal.
6153 (4) The test for "circular" in the Subject: header line has been removed
6154 because it now seems ill-conceived.
6156 15. Minor typos in src/EDITME comments corrected.
6158 16. Installed latest exipick from John Jetmore.
6160 17. If headers_add on a router specified a text string that was too long for
6161 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6162 of string_sprintf() is now avoided.
6164 18. $message_body_size was not set (it was always zero) when running the DATA
6165 ACL and the local_scan() function.
6167 19. For the "mail" command in an Exim filter, no default was being set for
6168 the once_repeat time, causing a random time value to be used if "once" was
6169 specified. (If the value happened to be <= 0, no repeat happened.) The
6170 default is now 0s, meaning "never repeat". The "vacation" command was OK
6171 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6172 (I found it when inspecting the code).
6174 20. There is now an overall timeout for performing a callout verification. It
6175 defaults to 4 times the callout timeout, which applies to individual SMTP
6176 commands during the callout. The overall timeout applies when there is more
6177 than one host that can be tried. The timeout is checked before trying the
6178 next host. This prevents very long delays if there are a large number of
6179 hosts and all are timing out (e.g. when the network connections are timing
6180 out). The value of the overall timeout can be changed by specifying an
6181 additional sub-option for "callout", called "maxwait". For example:
6183 verify = sender/callout=5s,maxwait=20s
6185 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6186 to the end before writing, but this should make it even safer).
6188 22. Exim was forgetting that it had advertised PIPELINING for the second and
6189 subsequent messages on an SMTP connection. It was also not resetting its
6190 memory on STARTTLS and an internal HELO.
6192 23. When Exim logs an SMTP synchronization error within a session, it now
6193 records whether PIPELINING has been advertised or not.
6195 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6196 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6197 rather than long int.
6199 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6201 26. Added the never_mail option to autoreply.
6207 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6208 crash if the getsockname() call failed; this can happen if a connection is
6209 closed very soon after it is established. The problem was simply in the
6210 order in which certain operations were done, causing Exim to try to write
6211 to the SMTP stream before it had set up the file descriptor. The bug has
6212 been fixed by making things happen in the correct order.
6218 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6219 before the connection was closed, thus losing the rejection response.
6221 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6222 some early Solaris releases, but causes trouble in current releases where
6223 socklen_t is defined.
6225 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6228 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6231 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6232 that releasing the top of it at the end releases what was used for sub-
6233 expansions (unless the block got too big). However, discard this block if
6234 the first thing is a variable or header, so that we can use its block when
6235 it is dynamic (useful for very large $message_headers, for example).
6237 6. Lookups now cache *every* query, not just the most recent. A new, separate
6238 store pool is used for this. It can be recovered when all lookup caches are
6239 flushed. Lookups now release memory at the end of their result strings.
6240 This has involved some general refactoring of the lookup sources.
6242 7. Some code has been added to the store_xxx() functions to reduce the amount
6243 of flapping under certain conditions.
6245 8. log_incoming_interface used to affect only the <= reception log lines. Now
6246 it causes the local interface and port to be added to several more SMTP log
6247 lines, for example "SMTP connection from", and rejection lines.
6249 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6251 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6253 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6254 could be overwritten at the end of the current message (or the start of a
6255 new message if it was set in a HELO ACL). The value is now preserved for
6256 the duration of the SMTP connection.
6258 12. If a transport had a headers_rewrite setting, and a matching header line
6259 contained an unqualified address, that address was qualified, even if it
6260 did not match any rewriting rules. The underlying bug was that the values
6261 of the flags that permit the existence of unqualified sender and recipient
6262 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6263 non-local messages, and by -bnq for local messages) were not being
6264 preserved with the message after it was received.
6266 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6267 "next input=" as part of the text comprising the host identity instead of
6268 the correct text. The code was using the same buffer for two different
6269 strings. However, depending on which order the printing function evaluated
6270 its arguments, the bug did not always show up. Under Linux, for example, my
6271 test suite worked just fine.
6273 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6274 doesn't work with some older versions of Perl. It has been changed to "my",
6275 which in any case is probably the better facility to use.
6277 15. A really picky compiler found some instances of statements for creating
6278 error messages that either had too many or two few arguments for the format
6281 16. The size of the buffer for calls to the DNS resolver has been increased
6282 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6283 for addresses that have a lot of PTR records. This alleviates a problem; it
6284 does not fully solve it.
6286 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6287 buffer now truncates the list and logs the incident, which is the same
6288 action as happens when Exim is looking up a host name and its aliases.
6289 Previously in this situation something unpredictable would happen;
6290 sometimes it was "internal error: store_reset failed".
6292 18. If a server dropped the connection unexpectedly when an Exim client was
6293 using GnuTLS and trying to read a response, the client delivery process
6294 crashed while trying to generate an error log message.
6296 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6297 string, for example:
6299 warn message = H1: something\nH2: something
6301 the text was added as a single header line from Exim's point of view
6302 though it ended up OK in the delivered message. However, searching for the
6303 second and subsequent header lines using $h_h2: did not work. This has been
6304 fixed. Similarly, if a system filter added multiple headers in this way,
6305 the routers could not see them.
6307 20. Expanded the error message when iplsearch is called with an invalid key to
6308 suggest using net-iplsearch in a host list.
6310 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6311 ACLs are no longer actually imposed (and a message to that effect is
6314 22. If a "gecos" field in a passwd entry contained escaped characters, in
6315 particular, if it contained a \" sequence, Exim got it wrong when building
6316 a From: or a Sender: header from that name. A second bug also caused
6317 incorrect handling when an unquoted " was present following a character
6318 that needed quoting.
6320 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6321 was not being matched caselessly.
6323 24. Arranged for all hyphens in the exim.8 source to be escaped with
6326 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6327 properly. Recipient callout cache records were still being keyed to include
6328 the sender, even when use_sender was set false. This led to far more
6329 callouts that were necessary. The sender is no longer included in the key
6330 when use_sender is false.
6332 26. Added "control = submission" modifier to ACLs.
6334 27. Added the ${base62d: operator to decode base 62 numbers.
6336 28. dnsdb lookups can now access SRV records.
6338 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6339 the configuration file.
6341 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6342 (-v) mode. This makes the output for a verbose queue run more intelligible.
6344 31. Added a use_postmaster feature to recipient callouts.
6346 32. Added the $body_zerocount variable, containing the number of binary zero
6347 bytes in the message body.
6349 33. The time of last modification of the "new" subdirectory is now used as the
6350 "mailbox time last read" when there is a quota error for a maildir
6353 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6355 35. Added +ignore_unknown as a special item in host lists.
6357 36. Code for decoding IPv6 addresses in host lists is now included, even if
6358 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6359 address was recognized as an IP address, but was then not correctly decoded
6360 into binary, causing unexpected and incorrect effects when compared with
6367 1. Very minor rewording of debugging text in manualroute to say "list of
6368 hosts" instead of "hostlist".
6370 2. If verify=header_syntax was set, and a header line with an unqualified
6371 address (no domain) and a large number of spaces between the end of the
6372 name and the colon was received, the reception process suffered a buffer
6373 overflow, and (when I tested it) crashed. This was caused by some obsolete
6374 code that should have been removed. The fix is to remove it!
6376 3. When running in the test harness, delay a bit after writing a bounce
6377 message to get a bit more predictability in the log output.
6379 4. Added a call to search_tidyup() just before forking a reception process. In
6380 theory, someone could use a lookup in the expansion of smtp_accept_max_
6381 per_host which, without the tidyup, could leave open a database connection.
6383 5. Added the variables $recipient_data and $sender_data which get set from a
6384 lookup success in an ACL "recipients" or "senders" condition, or a router
6385 "senders" option, similar to $domain_data and $local_part_data.
6387 6. Moved the writing of debug_print from before to after the "senders" test
6390 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6391 problems for message scanning, either using a data ACL, or using
6392 local_scan() because the Received: header was not generated till after they
6393 were called (in order to set the time as the time of reception completion).
6394 I have revised the way this works. The header is now generated after the
6395 body is received, but before the ACL or local_scan() are called. After they
6396 are run, the timestamp in the header is updated.
6402 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6403 before starting a queue runner without re-exec. This happened only when
6404 deliver_drop_privilege was set or when the Exim user was set to root. The
6405 effect of the bug was that timeouts during subsequent deliveries caused
6406 crashes instead of being properly handled. The handler is now left at its
6407 default (and expected) setting.
6409 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6410 message, again when deliver_drop_privilege is set or Exim is run as root.
6411 The bug described in (1) was not present in this case, but the tidying up
6412 of the other signals was missing. I have made the two cases consistent.
6414 3. The ignore_target_hosts setting on a manualroute router was being ignored
6415 for hosts that were looked up using the /MX notation.
6417 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6420 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6421 operated on the sender address. After changing the $sender_address to <>
6422 for the sender address verify, Exim was re-instated it as the original
6423 (before rewriting) address, but remembering that it had rewritten it, so it
6424 wasn't rewriting it again. This bug also had the effect of breaking the
6425 sender address verification caching when the sender address was rewritten.
6427 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6428 This has been changed so that if the ip literal address matches
6429 ignore_target_hosts, the router declines.
6431 7. Added expansion conditions match_domain, match_address, and match_local_
6432 part (NOT match_host).
6434 8. The placeholder for the Received: header didn't have a length field set.
6436 9. Added code to Exim itself and to exim_lock to test for a specific race
6437 condition that could lead to file corruption when using MBX delivery. The
6438 issue is with the lockfile that is created in /tmp. If this file is removed
6439 after a process has opened it but before that process has acquired a lock,
6440 there is the potential for a second process to recreate the file and also
6441 acquire a lock. This could lead to two Exim processes writing to the file
6442 at the same time. The added code performs the same test as UW imapd; it
6443 checks after acquiring the lock that its file descriptor still refers to
6444 the same named file.
6446 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6447 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6450 11. Added the smtp_active_hostname option. If used, this will typically be made
6451 to depend on the incoming interface address. Because $interface_address is
6452 not set up until the daemon has forked a reception process, error responses
6453 that can happen earlier (such as "too many connections") no longer contain
6456 12. If an expansion in a condition on a "warn" statement fails because a lookup
6457 defers, the "warn" statement is abandoned, and the next ACL statement is
6458 processed. Previously this caused the whole ACL to be aborted.
6460 13. Added the iplsearch lookup type.
6462 14. Added ident_timeout as a log selector.
6464 15. Added tls_certificate_verified as a log selector.
6466 16. Added a global option tls_require_ciphers (compare the smtp transport
6467 option of the same name). This controls incoming TLS connections.
6469 17. I finally figured out how to make tls_require_ciphers do a similar thing
6470 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6471 before starting the TLS session.
6473 18. Tabs are now shown as \t in -bP output.
6475 19. If the log selector return_path_on_delivery was set, Exim crashed when
6476 bouncing a message because it had too many Received: header lines.
6478 20. If two routers both had headers_remove settings, and the first one included
6479 a superfluous trailing colon, the final name in the first list and the
6480 first name in the second list were incorrectly joined into one item (with a
6481 colon in the middle).
6487 1. Added -C and -D options to the exinext utility, mainly to make it easier
6488 to include in the automated testing, but these could be helpful when
6489 multiple configurations are in use.
6491 2. The exinext utility was not formatting the output nicely when there was
6492 an alternate port involved in the retry record key, nor when there was a
6493 message id as well (for retries that were specific to a specific message
6494 and a specific host). It was also confused by IPv6 addresses, because of
6495 the additional colons they contain. I have fixed the IPv4 problem, and
6496 patched it up to do a reasonable job for IPv6.
6498 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6499 delivery, the log line now contains "pipelined" if PIPELINING was used.
6501 4. An SMTP transport process used to panic and die if the bind() call to set
6502 an explicit outgoing interface failed. This has been changed; it is now
6503 treated in the same way as a connect() failure.
6505 5. A reference to $sender_host_name in the part of a conditional expansion
6506 that was being skipped was still causing a DNS lookup. This no longer
6509 6. The def: expansion condition was not recognizing references to header lines
6510 that used bh_ and bheader_.
6512 7. Added the _cache feature to named lists.
6514 8. The code for checking quota_filecount in the appendfile transport was
6515 allowing one more file than it should have been.
6517 9. For compatibility with Sendmail, the command line option
6525 and sets the incoming protocol and host name (for trusted callers). The
6526 host name and its colon can be omitted when only the protocol is to be set.
6527 Note the Exim already has two private options, -pd and -ps, that refer to
6528 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6529 "s", but I don't think that's a major issue.
6531 10. A number of refactoring changes to the code, none of which should affect
6534 (a) The number of logging options was getting close to filling up the
6535 32-bit word that was used as a bit map. I have split them into two classes:
6536 those that are passed in the argument to log_write(), and those that are
6537 only ever tested independently outside of that function. These are now in
6538 separate 32-bit words, so there is plenty of room for expansion again.
6539 There is no change in the user interface or the logging behaviour.
6541 (b) When building, for example, log lines, the code previously used a
6542 macro that called string_cat() twice, in order to add two strings. This is
6543 not really sufficiently general. Furthermore, there was one instance where
6544 it was actually wrong because one of the argument was used twice, and in
6545 one call a function was used. (As it happened, calling the function twice
6546 did not affect the overall behaviour.) The macro has been replaced by a
6547 function that can join an arbitrary number of extra strings onto a growing
6550 (c) The code for expansion conditions now uses a table and a binary chop
6551 instead of a serial search (which was left over from when there were very
6552 few conditions). Also, it now recognizes conditions like "pam" even when
6553 the relevant support is not compiled in: a suitably worded error message is
6554 given if an attempt is made to use such a condition.
6556 11. Added ${time_interval:xxxxx}.
6558 12. A bug was causing one of the ddress fields not to be passed back correctly
6559 from remote delivery subprocesses. The field in question was not being
6560 subsequently used, so this caused to problems in practice.
6562 13. Added new log selectors queue_time and deliver_time.
6564 14. Might have fixed a bug in maildirsizefile handling that threw up
6565 "unexpected character" debug warnings, and recalculated the data
6566 unnecessarily. In any case, I expanded the warning message to give more
6569 15. Added the message "Restricted characters in address" to the statements in
6570 the default ACL that block characters like @ and % in local parts.
6572 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6573 Three changes have been made:
6575 (a) There was a serious bug; a negative response to MAIL caused the whole
6576 recipient domain to be cached as invalid, thereby blocking all messages
6577 to all local parts at the same domain, from all senders. This bug has
6578 been fixed. The domain is no longer cached after a negative response to
6579 MAIL if the sender used is not empty.
6581 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6584 (c) A new callout option, "use_sender" has been added for people who want
6585 the modified behaviour.
6591 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6594 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6595 indeed breaks things for older releases.
6597 3. Added additional logging to the case where there is a problem reading data
6598 from a filter that is running in a subprocess using a pipe, in order to
6599 try to track down a specific problem.
6601 4. Testing facility fudge: when running in the test harness and attempting
6602 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6603 getting "No route to host". Convert this to a timeout.
6605 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6608 6. Some OS don't have socklen_t but use size_t instead. This affects the
6609 fifth argument of getsockopt() amongst other things. This is now
6610 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6611 can be set for individual OS. I have set it for SunOS5, OSF1, and
6612 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6613 some earlier ones do not.
6615 7. Change 4.30/15 was not doing the test caselessly.
6617 8. The standard form for an IPv6 address literal was being rejected by address
6618 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6619 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6620 this, as well as the form without the "IPv6" on the front (but only when
6621 address literals are enabled, of course).
6623 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6625 10. Exim crashed if a message with an empty sender address specified by -f
6626 encountered a router with an errors_to setting. This could be provoked only
6627 by a command such as
6631 where an empty string was supplied; "<>" did not hit this bug.
6633 11. Installed PCRE release 4.5.
6635 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6636 remained set. It is now erased.
6638 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6639 times from message ids (which are base 36 rather than the normal 62).
6641 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6642 were being counted as actual protocol errors, and logged if the log
6643 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6644 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6645 and DATA following a set of rejected RCPTs do not count as protocol errors.
6646 In other words, Exim assumes they were pipelined, though this may not
6647 actually be the case. Of course, in all cases the client gets an
6648 appropriate error code.
6650 15. If a lookup fails in an ACL condition, a message about the failure may
6651 be available; it is used if testing the ACL cannot continue, because most
6652 such messages specify what the cause of the deferral is. However, some
6653 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6654 that caused an old message to be retained and used if a later statement
6655 caused a defer, replacing the real cause of the deferral.
6657 16. If an IP address had so many PTR records that the DNS lookup buffer
6658 was not large enough to hold them, Exim could crash while trying to process
6659 the truncated data. It now detects and logs this case.
6661 17. Further to 4.21/58, another change has been made: if (and only if) the
6662 first line of a message (the first header line) ends with CRLF, a bare LF
6663 in a subsequent header line has a space inserted after it, so as not to
6664 terminate the header.
6666 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6667 unnecessarily, used atoi() instead of strtol(), and didn't check the
6668 termination when getting file sizes from file names by regex.
6670 19. Completely re-implemented the support for maildirsize files, in the light
6671 of a number of problems with the previous contributed implementation
6672 (4.30/29). In particular:
6674 . If the quota is zero, the maildirsize file is maintained, but no quota is
6677 . If the maildir directory does not exist, it is created before any attempt
6678 to write a maildirsize file.
6680 . The quota value in the file is just a cache; if the quota is changed in
6681 the transport, the new value overrides.
6683 . A regular expression is available for excluding directories from the
6686 20. The autoreply transport checks the characters in options that define the
6687 message's headers; it allows continued headers, but it was checking with
6688 isspace() after an embedded newline instead of explicitly looking for a
6691 21. If all the "regular" hosts to which an address was routed had passed their
6692 expiry times, and had not reached their retry times, the address was
6693 bounced, even if fallback hosts were defined. Now Exim should go on to try
6696 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6697 equivalent code in the SMTP transport. Some hosts send humungous responses
6698 to HELO/EHLO, more than 1024 it seems.
6700 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6701 gives compiler warnings in some environments. I've now done it "properly",
6704 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6705 (because of problems with the built-in one) was declared to return uschar *
6706 instead of char *, causing compiler failure.
6708 25. Fixed a file descriptor leak when processing alias/forward files.
6710 26. Fixed a minor format string issue in dbfn.c.
6712 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6714 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6715 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6718 29. When a sender address is verified, it is cached, to save repeating the test
6719 when there is more than one recipient in a message. However, when the
6720 verification involves a callout, it is possible for different callout
6721 options to be set for different recipients. It is too complicated to keep
6722 track of this in the cache, so now Exim always runs a verification when a
6723 callout is required, relying on the callout cache for the optimization.
6724 The overhead is duplication of the address routing, but this should not be
6727 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6728 to be verified with callout=postmaster, and the main callout worked but the
6729 postmaster check failed, the verification correctly failed. However, if a
6730 subsequent RCPT command asked for sender verification *without* the
6731 postmaster check, incorrect caching caused this verification also to fail,
6734 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6735 it was not caching the DNS options (qualify_single, search_parents) that
6736 were used when the lookup failed. A subsequent lookup with different
6737 options therefore always gave the same answer, though there were cases
6738 where it should not have. (Example: a "domains = !$mx_any" option on a
6739 dnslookup router: the "domains" option is always processed without any
6740 widening, but the router might have qualify_single set.) Now Exim uses the
6741 cached value only when the same options are set.
6743 32. Added John Jetmore's "exipick" utility to the distribution.
6745 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6746 than a timeout (e.g. a certificate is required, and is not provided), an
6747 Exim server now closes the connection immediately. Previously it waited for
6748 the client to close - but if the client is SSL, it seems that they each
6749 wait for each other, leading to a delay before one of them times out.
6751 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6752 maintained 0.8.x compatibility because I don't think many are using it, and
6753 it is clearly obsolete.
6755 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6758 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6759 client certificate was expired. A simple patch fixes this, though I don't
6760 understand the full logic of why the verify callback is called multiple
6763 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6764 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6765 which causes problems with some clients (such as the Certicom SSL Plus
6766 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6767 disables the coutermeasure allowing Eudora to connect."
6769 38. Exim was not checking that a write() to a log file succeeded. This could
6770 lead to Bad Things if a log got too big, in particular if it hit a file
6771 size limit. Exim now panics and dies if it cannot write to a log file, just
6772 as it does if it cannot open a log file.
6774 39. Modified OS/Makefile-Linux so that it now contains
6776 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6778 The two -D definitions ensure that Exim is compiled with large file
6779 support, which makes it possible to handle log files that are bigger than
6782 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6783 instance) a domain was checked against a named list that involved a lookup,
6784 causing $domain_data to be set, then another domain was checked against the
6785 same list, then the first domain was re-checked, the value of $domain_data
6786 after the final check could be wrong. In particular, if the second check
6787 failed, it could be set empty. This bug probably also applied to
6790 41. The strip_trailing_dot option was not being applied to the address given
6791 with the -f command-line option.
6793 42. The code for reading a message's header from the spool was incrementing
6794 $received_count, but never initializing it. This meant that the value was
6795 incorrect (doubled) while delivering a message in the same process in which
6796 it was received. In the most common configuration of Exim, this never
6797 happens - a fresh exec is done - but it can happen when
6798 deliver_drop_privilege is set.
6800 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6801 it now includes up to 150 characters of the unexpected data in the log
6804 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6805 and building data strings. The size of both of these buffers was 10 000
6806 bytes - far larger than anybody would *ever* want, thought I. Needless to
6807 say, somebody hit the limit. I have increased the maximum line length to
6808 20 000 and the maximum data length of concatenated lines to 100 000. I have
6809 also fixed two bugs, because there was no checking on these buffers. Tsk,
6810 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6811 buffer is too small.
6813 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6814 lsearch lookups. Now it does.
6816 46. When parsing a route_list item in a manualroute router, a fixed-length
6817 buffer was used for the list of hosts. I made this 1024 bytes long,
6818 thinking that nobody would ever have a list of hosts that long. Wrong.
6819 Somebody had a whole pile of complicated expansion conditions, and the
6820 string was silently truncated, leading to an expansion error. It turns out
6821 that it is easier to change to an unlimited length (owing to other changes
6822 that have happened since this code was originally written) than to build
6823 structure for giving a limitation error. The length of the item that
6824 expands into the list of hosts is now unlimited.
6826 47. The lsearch lookup could not handle data where the length of text line was
6827 more than 4095 characters. Such lines were truncated, leading to shortened
6828 data being returned. It should now handle lines of any length.
6830 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6831 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6834 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6835 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6836 no longer happens in comment lines. A list of replacements is now placed
6837 at the head of all of the source files, except those whose only change is
6838 to replace PERL_COMMAND in the very first #! line.
6840 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6841 messages on the queue, with a bottom-up merge sort, using code contributed
6842 by Michael Haardt. This should make operations like -bp somewhat faster on
6843 large queues. It won't affect queue runners, except when queue_run_in_order
6846 51. Installed eximstats 1.31 in the distribution.
6848 52. Added support for SRV lookups to the dnslookup router.
6850 53. If an ACL referred to $message_body or $message_body_end, the value was not
6851 reset for any messages that followed in the same SMTP session.
6853 54. The store-handling optimization for building very long strings was not
6854 differentiating between the different store pools. I don't think this
6855 actually made any difference in practice, but I've tidied it.
6857 55. While running the routers to verify a sender address, $sender_address
6858 was still set to the sender address. This is wrong, because when routing to
6859 send a bounce to the sender, it would be empty. Therefore, I have changed
6860 it so that, while verifying a sender address, $sender_address is set to <>.
6861 (There is no change to what happens when verifying a recipient address.)
6863 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6864 target A or AAAA records (if not already returned) without resetting the
6865 qualify_single or search_parents options of the DNS resolver. These are
6866 inappropriate in this case because the targets of MX and SRV records must
6867 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6868 target that, when qualified, matched something in the local domain. These
6869 two options are now turned off when doing these lookups.
6871 57. It seems that at least some releases of Reiserfs (which does not have the
6872 concept of a fixed number of inodes) returns zero and not -1 for the
6873 number of available inodes. This interacted badly with check_spool_inodes,
6874 which assumed that -1 was the "no such thing" setting. What I have done is
6875 to check that the total number of inodes is greater than zero before doing
6876 the test of how many are available.
6878 58. When a "warn" ACL statement has a log_message modifier, the message is
6879 remembered, and not repeated. This is to avoid a lot of repetition when a
6880 message has many recipients that cause the same warning to be written.
6881 However, Exim was preserving the list of already written lines for an
6882 entire SMTP session, which doesn't seem right. The memory is now reset if a
6883 new message is started.
6885 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6886 debugging output unless log_rewrite was also set.
6888 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6889 of (int)(handle) when we know that handle contains (void *)(-1).
6891 61. The Exim daemon panic-logs an error return when it closes the incoming
6892 connection. However "connection reset by peer" seems to be common, and
6893 isn't really an error worthy of noting specially, so that particular error
6896 62. When Exim is trying to find all the local interfaces, it used to panic and
6897 die if the ioctl to get the interface flags failed. However, it seems that
6898 on at least one OS (Solaris 9) it is possible to have an interface that is
6899 included in the list of interfaces, but for which you get a failure error
6900 for this call. This happens when the interface is not "plumbed" into a
6901 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6902 failure of the "get flags" call assumes that the interface is down.
6904 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6905 makes life easier for people who are doing arithmetic on fields extracted
6906 from dates, where you often get leading zeros that should not be
6907 interpreted as octal.
6909 64. Added qualify_domain to the redirect router, to override the global
6912 65. If a pathologically long header line contained very many addresses (the
6913 report of this problem mentioned 10 000) and each of them was rewritten,
6914 Exim could use up a very large amount of memory. (It kept on making new
6915 copies of the header line as it rewrote, and never released the old ones.)
6916 At the expense of a bit more processing, the header rewriting function has
6917 been changed so that it no longer eats memory in this way.
6919 66. The generation of the Received: header has been moved from the time that a
6920 message starts to be received, to the time that it finishes. The timestamp
6921 in the Received: header should now be very close to that of the <= log
6922 line. There are two side-effects of this change:
6924 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6925 logged header lines no longer include the local Received: line, because
6926 it has not yet been created. The same applies to a copy of the message
6927 that is returned to a non-SMTP sender when a message is rejected.
6929 (b) When a filter file is tested using -bf, no additional Received: header
6930 is added to the test message. After some thought, I decided that this
6933 This change does not affect the value of $received_for. It is still set
6934 after address rewriting, but before local_scan() is called.
6936 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6938 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6939 gave an unhelpful panic error message, and a defer error. I have managed to
6940 change this behaviour so that it now rejects any supplied certificate,
6941 which seems right, as the list of acceptable certificates is empty.
6943 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6944 gave an unhelpful defer error. I have not managed to make this reject any
6945 supplied certificates, but the error message it gives is "no certificate
6946 supplied", which is not helpful.
6948 70. exigrep's output now also includes lines that are not associated with any
6949 message, but which match the given pattern. Implemented by a patch from
6950 Martin Sluka, which also tidied up the Perl a bit.
6952 71. Recipient callout verification, like sender verification, was using <> in
6953 the MAIL FROM command. This isn't really the right thing, since the actual
6954 sender may affect whether the remote host accepts the recipient or not. I
6955 have changed it to use the actual sender in the callout; this means that
6956 the cache record is now keyed on a recipient/sender pair, not just the
6957 recipient address. There doesn't seem to be a real danger of callout loops,
6958 since a callout by the remote host to check the sender would use <>.
6959 [SEE ABOVE: changed after hitting problems.]
6961 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6962 temporary errors. However, in the case of such a code being given after
6963 the end of a data transmission (i.e. after ".") Exim was failing to write
6964 a retry record for the message. (Yes, there was some broken host that was
6965 actually sending 8xx at this point.)
6967 73. An unknown lookup type in a host list could cause Exim to panic-die when
6968 the list was checked. (An example that provoked this was putting <; in the
6969 middle of a list instead of at the start.) If this happened during a DATA
6970 ACL check, a -D file could be left lying around. This kind of configuration
6971 error no longer causes Exim to die; instead it causes a defer error. The
6972 incident is still logged to the main and panic logs.
6974 74. Buglet left over from Exim 3 conversion. The message "too many messages
6975 in one connection" was written to the rejectlog but not the mainlog, except
6976 when address rewriting (yes!) was being logged.
6978 75. Added write_rejectlog option.
6980 76. When a system filter was run not as root (that is, when system_filter_user
6981 was set), the values of the $n variables were not being returned to the
6982 main process; thus, they were not subsequently available in the $sn
6985 77. Added +return_path_on_delivery log selector.
6987 78. A connection timeout was being treated differently from recipients deferred
6988 when testing hosts_max_try with a message that was older than the host's
6989 retry timeout. (The host should not be counted, thus allowing all hosts to
6990 be tried at least once before bouncing.) This may have been the cause of an
6991 occasionally reported bug whereby a message would remain on the queue
6992 longer than the retry timeout, but would be bounced if a delivery was
6993 forced. I say "may" because I never totally pinned down the problem;
6994 setting up timeout/retry tests is difficult. See also the next item.
6996 79. The ultimate address timeout was not being applied to errors that involved
6997 a combination of host plus message (for example, a timeout on a MAIL
6998 command). When an address resolved to a number of possible hosts, and they
6999 were not all tried for each delivery (e.g. because of hosts_max_try), a
7000 message could remain on the queue longer than the retry timeout.
7002 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7005 81. Fixed an obscure SMTP outgoing bug which required at least the following
7006 conditions: (a) there was another message waiting for the same server;
7007 (b) the server returned 5xx to all RCPT commands in the first message so
7008 that the message was not completed; (c) the server dropped the connection
7009 or gave a negative response to the RSET that Exim sends to abort the
7010 transaction. The observed case was a dropped connection after DATA that had
7011 been sent in pipelining mode. That is, the server had advertised PIPELINING
7012 but was not implementing it correctly. The effect of the bug was incorrect
7013 behaviour, such as trying another host, and this could lead to a crash.
7019 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7020 and daemon.c were passed as pointers to ints; they should have been
7021 pointers to socklen_t variables (which are typically unsigned ints).
7023 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7026 3. Fixed a really odd bug that affected only the testing scheme; patching a
7027 certain fixed string in the binary changed the value of another string that
7028 happened to be identical to the end of the original first string.
7030 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7031 name", it returns that address as the IP address. On some operating
7032 systems (e.g. Solaris), it also passes back the IP address string as the
7033 "host name". However, on others (e.g. Linux), it passes back an empty
7034 string. Exim wasn't checking for this, and was changing the host name to an
7035 empty string, assuming it had been canonicalized.
7037 5. Although rare, it is permitted to have more than one PTR record for a given
7038 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7039 all the names associated with an address, because they do in Solaris.
7040 However, it seems that they do not in Linux for data that comes from the
7041 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7042 I found this out when I moved to a new Linux workstation and tried to run
7043 the Exim test suite.
7045 To get round this problem I have changed the code so that it now does its
7046 own call to the DNS to look up PTR records when searching for a host name.
7047 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7048 addresses that are only in /etc/hosts are still found.
7050 This behaviour is, however, controlled by an option called host_lookup_
7051 order, which defaults to "bydns:byaddr". If people want to use the other
7052 order, or indeed, just use one or the other means of lookup, they can
7053 specify it in this variable.
7055 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7056 some operating systems, this comes back from gethostbyaddr() as an empty
7057 string, and this is what Exim used to test for. However, it seems that in
7058 other systems, "." is yielded. Exim now tests for this case too.
7060 7. The values of check_spool_space and check_log_space are now held internally
7061 as a number of kilobytes instead of an absolute number of bytes. If a
7062 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7063 kilobyte. This means that much larger values can be stored.
7065 8. Exim monitor: an attempt to get the action menu when not actually pointing
7066 at a message produces an empty menu entitled "No message selected". This
7067 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7068 no entries in it ("Shell widget menu has zero width and/or height"). So I
7069 have added a single, blank menu entry in this case.
7071 9. Added ${quote_local_part.
7073 10. MIME decoding is now applied to the contents of Subject: header lines when
7076 11. Now that a reference to $sender_host_address automatically causes a reverse
7077 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7078 host lookup before query-style lookups in lists that might use this
7079 variable. This has therefore been abolished, and the "net-" prefix is no
7080 longer necessary for query-style lookups.
7082 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7083 have been a typo for LFLAGS, so it has been changed.
7085 13. The install script calls Exim with "-C /dev/null" in order to find the
7086 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7087 to be output. However, since Exim outputs its version number before the
7088 error, it didn't break the script. It just looked ugly. I fixed this by
7089 always allowing "-C /dev/null" if the caller is root.
7091 14. Ignore overlarge ACL variable number when reading spool file - insurance
7092 against a later release with more variables having written the file.
7094 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7095 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7096 this, as well as the form without the "IPv6" on the front.
7098 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7099 OS/Makefile-Darwin file.
7101 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7102 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7104 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7105 variable. However, some OS use ints for inodes. I've added cast to long int
7106 to get rid of the compiler warning.
7108 19. I had forgotten to lock out "/../" in configuration file names when
7109 ALT_CONFIG_PREFIX was set.
7111 20. Routers used for verification do not need to specify transports. However,
7112 if such a router generated a host list, and callout was configured, Exim
7113 crashed, because it could not find a port number from the (non-existent)
7114 transport. It now assumes port 25 in this circumstance.
7116 21. Added the -t option to exigrep.
7118 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7119 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7120 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7121 them is set, LOOKUP_LSEARCH is forced.
7123 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7125 24. Added sender and host information to the "rejected by local_scan()" log
7126 line; previously there was no indication of these.
7128 25. Added .include_if_exists.
7130 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7131 Linux. It turns out that not all file systems support this. Apparently some
7132 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7133 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7134 not supported on the file descriptor, is now ignored when Exim is trying to
7135 sync a directory. This applies only to Linux.
7137 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7139 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7140 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7143 29. Added support for maildirsize files from supplied patch (modified a bit).
7145 30. The use of :fail: followed by an empty string could lead Exim to respond to
7146 sender verification failures with (e.g.):
7148 550 Verification failed for <xxx>
7149 550 Sender verify failed
7151 where the first response line was missing the '-' that indicates it is not
7152 the final line of the response.
7154 31. The loop for finding the name of the user that called Exim had a hardwired
7155 limit of 10; it now uses the value of finduser_retries, which is used for
7156 all other user lookups.
7158 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7161 33. Exim was neglecting to zero errno before one call of strtol() when
7162 expanding a string and expecting an integer value. On some systems this
7163 resulted in spurious "integer overflow" errors. Also, it was casting the
7164 result into an int without checking.
7166 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7167 did not work. The code looks as if it has *never* worked, though it appears
7168 to have been documented since at least release 1.62. I have made it work.
7170 35. The "timeout_DNS" error in retry rules, also documented since at least
7171 1.62, also never worked. As it isn't clear exactly what this means, and
7172 clearly it isn't a major issue, I have abolished the feature by treating it
7173 as "timeout", and writing a warning to the main and panic logs.
7175 36. The display of retry rules for -brt wasn't always showing the error code
7178 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7179 timeout_connect_A, timeout_connect_MX.
7181 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7182 to the empty sender.
7184 39. The daemon was not analysing the content of -oX till after it had closed
7185 stderr and disconnected from the controlling terminal. This meant that any
7186 syntax errors were only noted on the panic log, and the return code from
7187 the command was 0. By re-arranging the code a little, I've made the
7188 decoding happen first, so such errors now appear on stderr, and the return
7189 code is 1. However, the actual setting up of the sockets still happens in
7190 the disconnected process, so errors there are still only recorded on the
7193 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7194 connections (as happens on some IP stacks) was logged at start up time as
7195 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7196 it from "IPv6 and IPv4", which means that two separate sockets are being
7199 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7200 says whether AF_INET or AF_INET6 was passed as an argument.
7202 42. Exiwhat output was messed up when time zones were included in log
7205 43. Exiwhat now gives more information about the daemon's listening ports,
7206 and whether -tls-on-connect was used.
7208 44. The "port" option of the smtp transport is now expanded.
7210 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7211 silently ignored. Now an error message is written to the main and panic
7214 46. There's a new ACL modifier called "logwrite" which writes to a log file
7215 as soon as it is encountered.
7217 47. Added $local_user_uid and $local_user_gid at routing time.
7219 48. Exim crashed when trying to verify a sender address that was being
7222 49. Exim was recognizing only a space character after ".include". It now also
7223 recognizes a tab character.
7225 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7226 extracting the relevant information from the specification. The man page no
7227 longer contains scrambled data for the -d option, and I've added a section
7228 at the front about calling Exim under different names.
7230 51. Added "extra_headers" argument to the "mail" command in filter files.
7232 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7235 53. Installed eximstats 1.29.
7237 54. Added transport_filter_timeout as a generic transport option.
7239 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7240 Cc: header lines. This was required by RFC 822, but it not required by RFC
7243 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7244 incoming messages that did not have them. Now it does so only if the
7245 message originates locally, that is, if there is no associated remote host
7246 address. When Resent- header lines are present, this applies to the Resent-
7247 lines rather than the non-Resent- lines.
7249 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7250 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7252 58. Messages for configuration errors now include the name of the main
7253 configuration file - useful now that there may be more than one file in a
7254 list (.included file names were always shown).
7256 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7257 for those rare installations that do not start the daemon as root or run it
7258 setuid root. I've cut out the call to initgroups() if the daemon is not
7261 60. The Exim user and group can now be bound into the binary as text strings
7262 that are looked up at the start of Exim's processing.
7264 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7266 62. Added $mailstore_basename variable.
7268 63. Installed patch to sieve.c from Michael Haardt.
7270 64. When Exim failed to open the panic log after failing to open the main log,
7271 the original message it was trying to log was written to stderr and debug
7272 output, but if they were not available (the usual case in production), it
7273 was lost. Now it is written to syslog before the two lines that record the
7274 failures to open the logs.
7276 65. Users' Exim filters run in subprocesses under the user's uid. It is
7277 possible for a "deliver" command or an alias in a "personal" command to
7278 provoke an address rewrite. If logging of address rewriting is configured,
7279 this fails because the process is not running as root or exim. There may be
7280 a better way of dealing with this, but for the moment (because 4.30 needs
7281 to be released), I have disabled address rewrite logging when running a
7282 filter in a non-root, non-exim process.
7288 1. The buildconfig auxiliary program wasn't quoting the value set for
7289 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7290 not defined. This bug was masked in 4.22 by the effect that was fixed in
7293 2. Some messages that were rejected after a message id was allocated were
7294 shown as "incomplete" by exigrep. It no longer does this for messages that
7295 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7297 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7298 have allow_domain_literals set, the ID did not get logged in the <= line.
7299 Domain literals are now always recognized in Message-ID: header lines.
7301 4. The first argument for a ${extract expansion item is the key name or field
7302 number. Leading and trailing spaces in this item were not being ignored,
7303 causing some misleading effects.
7305 5. When deliver_drop_privilege was set, single queue runner processes started
7306 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7307 same command in the process it spins off) were not dropping privilege.
7309 6. When the daemon running as "exim" started a queue runner, it always
7310 re-executed Exim in the spun-off process. This is a waste of effort when
7311 deliver_drop_privilege is set. The new process now just calls the
7312 queue-runner function directly.
7318 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7321 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7322 ignored. Though the use of -oP was forcing the writing of a pid file, it
7323 was always written to the default place.
7325 3. If the message "no IP address found for host xxxx" is generated during
7326 incoming verification, it is now followed by identification of the incoming
7327 connection (so you can more easily find what provoked it).
7329 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7331 5. Added some features to "harden" Exim a bit more against certain attacks:
7333 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7334 be put in Local/Makefile. This is like the never_users runtime option,
7335 but it cannot be overridden. The default setting is "root".
7337 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7338 prefix string with which any file named in a -C command line option
7341 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7342 is retained for -C and -D only if the caller of Exim is root. Without
7343 it, the exim user may also use -C and -D and retain privilege.
7345 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7346 command line option is disabled.
7348 6. Macro names set by the -D option must start with an upper case letter, just
7349 like macro names defined in the configuration file.
7351 7. Added "dereference=" facility to LDAP.
7353 8. Two instances of the typo "uknown" in the source files are fixed.
7355 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7356 the Configure-Makefile script screwed up while processing it.
7358 10. Incorporated PCRE 4.4.
7360 11. The SMTP synchronization check was not operating right at the start of an
7361 SMTP session. For example, it could not catch a HELO sent before the client
7362 waited for the greeting. There is now a check for outstanding input at the
7363 point when the greeting is written. Because of the duplex, asynchronous
7364 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7365 way, but not yet received, when the check is performed.
7367 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7368 on TCP/IP sockets, because this apparently causes some broken clients to
7371 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7372 unchanged) from the Cygwin maintainer.
7374 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7375 was supported instead of testing for mbx. Effectively a typo.
7377 15. The spa authenticator server code was not checking that the input it
7378 received was valid base64.
7380 16. The debug output line for the "set" modifier in ACLs was not showing the
7381 name of the variable that was being set.
7383 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7385 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7386 lookup on its own. Something else was needed to trigger the lookup. For
7387 example, a match in host_lookup or the need for a host name in a host list.
7388 Now, if $sender_host_name is referenced and the host name has not yet been
7389 looked up, a lookup is performed. If the lookup fails, the variable remains
7390 empty, and $host_lookup_failed is set to "1".
7392 19. Added "eqi" as a case-independent comparison operator.
7394 20. The saslauthd authentication condition could segfault if neither service
7395 nor realm was specified.
7397 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7398 error message that was logged was misleading, and incoming SMTP
7399 connections were dropped. The message is now more accurate, and temporary
7400 errors are given to SMTP connections.
7402 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7403 (except RSET) with a 421 error, until QUIT is received. However, it was
7404 failing to send a response to QUIT.
7406 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7407 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7408 ACL is now run after the helo_try_verify_hosts code.
7410 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7411 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7412 case-independent, so other case variants are also recognized). Apparently
7413 some systems use these upper case variants.
7415 25. If more than two messages were waiting for the same host, and a transport
7416 filter was specified for the transport, Exim sent two messages over the
7417 same TCP/IP connection, and then failed with "socket operation on non-
7418 socket" when it tried to send the third.
7420 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7422 27. The extern definition of crypt16() in expand.c was not being excluded when
7423 the OS had its own crypt16() function.
7425 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7426 as a preferred synonym for return_size_limit, both as an option and as an
7429 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7431 30. Changed the default configuration ACL to relax the local part checking rule
7432 for addresses that are not in any local domains. For these addresses,
7433 slashes and pipe symbols are allowed within local parts, but the sequence
7434 /../ is explicitly forbidden.
7436 31. SPA server authentication was not clearing the challenge buffer before
7439 32. log_message in a "warn" ACL statement was writing to the reject log as
7440 well as to the main log, which contradicts the documentation and doesn't
7441 seem right (because no rejection is happening). So I have stopped it.
7443 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7444 However, I am unable to do any testing of this.
7446 34. Fixed an infelicity in the appendfile transport. When checking directories
7447 for a mailbox, to see if any needed to be created, it was accidentally
7448 using path names with one or more superfluous leading slashes; tracing
7449 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7451 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7452 contents are added to the log line that is written for every discarded
7453 recipient. (Previously a log_message setting was ignored.)
7455 36. The ${quote: operator now quotes the string if it is empty.
7457 37. The install script runs exim in order to find its version number. If for
7458 some reason other than non-existence or emptiness, which it checks, it
7459 could not run './exim', it was installing it with an empty version number,
7460 i.e. as "exim-". This error state is now caught, and the installation is
7463 38. An argument was missing from the function that creates an error message
7464 when Exim fails to connect to the socket for saslauthd authentication.
7465 This could cause Exim to crash, or give a corrupted message.
7467 39. Added isip, isip4, and isip6 to ${if conditions.
7469 40. The ACL variables $acl_xx are now saved with the message, and can be
7470 accessed later in routers, transports, and filters.
7472 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7473 strings in the file are not string-expanded.
7475 42. If a MAIL command specified a SIZE value that was too large to fit into an
7476 int variable, the check against message_size_limit failed. Such values are
7477 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7478 day this will have to be increased, but I don't think I want to be around
7479 when emails are that large.
7486 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7487 iconv() is not standard in FreeBSD.
7489 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7490 IPv6 enabled. The observed symptom was a segmentation fault on return from
7491 the function os_common_find_running_interfaces() in src/os.c.
7493 3. In the check_special_case() function in daemon.c I had used "errno" as an
7494 argument name, which causes warnings on some systems. This was basically a
7495 typo, since it was named "eno" in the comments!
7497 4. The code that waits for the clock to tick (at a resolution of some fraction
7498 of a second) so as to ensure message-id uniqueness was always waiting for
7499 at least one whole tick, when it could have waited for less. [This is
7500 almost certainly not relevant at current processor speeds, where it is
7501 unlikely to ever wait at all. But we try to future-proof.]
7503 5. The function that sleeps for a time interval that includes fractions of a
7504 second contained a race. It did not block SIGALRM between setting the
7505 timer, and suspending (a couple of lines later). If the interval was short
7506 and the sigsuspend() was delayed until after it had expired, the suspension
7507 never ended. On busy systems this could lead to processes getting stuck for
7510 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7511 process, before it forks any delivery processes. The open lookup caching
7512 mechanism meant that the open file or database connection was passed into
7513 the delivery process. The problem was that delivery processes always tidy
7514 up cached lookup data. This could cause a problem for the next delivery
7515 process started by the queue runner, because the external queue runner
7516 process does not know about the closure. So the next delivery process
7517 still has data in the lookup cache. In the case of a file lookup, there was
7518 no problem because closing a file descriptor in a subprocess doesn't affect
7519 the parent. However, if the lookup was caching a connection to a database,
7520 the connection was closed, and the second delivery process was likely to
7521 see errors such as "PGSQL: query failed: server closed the connection
7522 unexpectedly". The problem has been fixed by closing all cached lookups
7523 in a queue runner before running a delivery process.
7525 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7526 seem to have the "const" qualifier which it has on other OS. I've
7529 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7530 *of the same type* (client or server) with the same public name that an
7531 error should be diagnosed.
7533 9. When Exim looked up a host name for an IP address, but failed to find the
7534 original IP address when looking up the host name (a safety check), it
7535 output the message "<ip address> does not match any IP for NULL", which was
7536 confusing, to say the least. The bug was that the host name should have
7537 appeared instead of "NULL".
7539 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7540 user that is built into the binary, and the -C or -D options is used, root
7541 privilege is dropped before the configuration file is read. In addition,
7542 logging is switched to stderr instead of the normal log files. If the
7543 configuration then re-defines the Exim user, the unprivileged environment
7544 is probably not what is expected, so Exim logs a panic warning message (but
7547 However, if deliver_drop_privilege is set, the unprivileged state may well
7548 be exactly what is intended, so the warning has been cut out in that case,
7549 and Exim is allowed to try to write to its normal log files.
7555 1. smtp_return_error_details was not giving details for temporary sender
7556 or receiver verification errors.
7558 2. Diagnose a configuration error if two authenticators have the same public
7561 3. Exim used not to create the message log file for a message until the first
7562 delivery attempt. This could be confusing when incoming messages were held
7563 for policy or load reasons. The message log file is now created at the time
7564 the message is received, and an initial "Received" line is written to it.
7566 4. The automatically generated man page for command line options had a minor
7567 bug that caused no ill effects; however, a more serious problem was that
7568 the procedure for building the man page automatically didn't always
7569 operate. Consequently, release 4.20 contains an out-of-date version. This
7570 shouldn't happen again.
7572 5. When building Exim with embedded Perl support, the script that builds the
7573 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7574 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7576 6. The freeze_tell option was not being used for messages that were frozen on
7577 arrival, either by an ACL or by local_scan().
7579 7. Added the smtp_incomplete_transaction log selector.
7581 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7582 was accepting AUTH without a new EHLO.
7584 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7585 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7588 10. Logging of TCP/IP connections (when configured) now happens in the main
7589 daemon process instead of the child process, so that the TCP/IP connection
7590 count is more accurate (but it can never be perfect).
7592 11. The use of "drop" in a nested ACL was not being handled correctly in the
7593 outer ACL. Now, if condition failure induced by the nested "drop" causes
7594 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7595 or "require"), the connection is dropped.
7597 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7598 that yield "discard" can now be used with an "accept" or a "discard" verb,
7599 but an error is generated for any others (because I can't see a useful way
7600 to define what should happen).
7602 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7603 are now processed in the same way as lines in the Exim configuration file.
7604 In particular, continuation lines are supported.
7606 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7608 15. Added -ti meaning -t -i.
7610 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7611 domains, and warn by logging if others are found.
7613 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7614 returned from GIFCONF when Exim is trying to find the list of interfaces on
7615 a host. The code in os.c has been modified to copy each ifreq to an aligned
7616 structure in all cases.
7618 Also, in some cases, the returned ifreq's were being copied to a 'struct
7619 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7620 means the last couple of bytes of an IPv6 address could be chopped if the
7621 ifreq contained only a normal sockaddr (14 bytes storage).
7623 18. Named domain lists were not supported in the hosts_treat_as_local option.
7624 An entry such as +xxxx was not recognized, and was treated as a literal
7627 19. Ensure that header lines added by a DATA ACL are included in the reject log
7628 if the ACL subsequently rejects the message.
7630 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7631 MD5 (which is deprecated).
7633 21. When testing a filter file using -bf, Exim was writing a message when it
7634 took the sender from a "From " line in the message, but it was not doing so
7635 when it took $return_path from a Return-Path: header line. It now does.
7637 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7638 with a valid header line field name (a series of printing characters
7639 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7641 23. Changed "disc" in the source to "disk" to conform to the documentation and
7642 the book and for uniformity.
7644 24. Ignore Sendmail's -Ooption=value command line item.
7646 25. When execve() failed while trying to run a command in a pipe transport,
7647 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7648 could be confused with a return value of 69 from the command itself. This
7649 has been changed to 127, the value the shell returns if it is asked to run
7650 a non-existent command. The wording for the related log line suggests a
7651 non-existent command as the problem.
7653 26. If received_header_text expands to an empty string, do not add a Received:
7654 header line to the message. (Well, it adds a token one on the spool, but
7655 marks it "old" so that it doesn't get used or transmitted.)
7657 27. Installed eximstats 1.28 (addition of -nt option).
7659 28. There was no check for failure on the call to getsockname() in the daemon
7660 code. This can fail if there is a shortage of resources on the system, with
7661 ENOMEM, for example. A temporary error is now given on failure.
7663 29. Contrary to the C standard, it seems that in some environments, the
7664 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7665 program. Exim now does this explicitly; it affects the formatting of
7666 timestamps using strftime().
7668 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7669 complaints. I've now initialized all the variables, to avoid this.
7671 32. Header lines added by a system filter were not being "seen" during
7672 transport-time rewrites.
7674 33. The info_callback() function passed to OpenSSL is set up with type void
7675 (*)(SSL *, int, int), as described somewhere. However, when calling the
7676 function (actually a macro) that sets it up, the type void(*)() is
7677 expected. I've put in a cast to prevent warnings from picky compilers.
7679 34. If a DNS black list lookup found a CNAME record, but there were no A
7680 records associated with the domain it pointed at, Exim crashed.
7682 35. If a DNS black list lookup returned more than one A record, Exim ignored
7683 all but the first. It now scans all returned addresses if a particular IP
7684 value is being sought. In this situation, the contents of the
7685 $dnslist_value variable are a list of all the addresses, separated by a
7688 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7689 to accept a host name and all its aliases if the forward lookup for any of
7690 them yielded the IP address of the incoming connection. Now it accepts only
7691 those names whose forward lookup yields the correct IP address. Any other
7692 names are discarded. This closes a loophole whereby a rogue DNS
7693 administrator could create reverse DNS records to break through a
7694 wildcarded host restriction in an ACL.
7696 37. If a user filter or a system filter that ran in a subprocess used any of
7697 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7698 the wrong values were passed to the pipe command ($thisaddress had the
7699 value of $0, $0 had the value of $1, etc). This bug was introduced by
7700 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7702 38. Improved the line breaking for long SMTP error messages from ACLs.
7703 Previously, if there was no break point between 40 and 75 characters, Exim
7704 left the rest of the message alone. Two changes have been made: (a) I've
7705 reduced the minimum length to 35 characters; (b) if it can't find a break
7706 point between 35 and 75 characters, it looks ahead and uses the first one
7707 that it finds. This may give the occasional overlong line, but at least the
7708 remaining text gets split now.
7710 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7711 file descriptors might be low, and that setting 1000 would always raise it.
7712 It turns out that in some environments, the limit is already over 1000 and
7713 that lowering it causes trouble. So now Exim takes care not to decrease it.
7715 40. When delivering a message, the value of $return_path is set to $sender_
7716 address at the start of routing (routers may change the value). By an
7717 oversight, this default was not being set up when an address was tested by
7718 -bt or -bv, which affected the outcome if any router or filter referred to
7721 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7722 log only when "message" or "log_message" are set. However, if one of the
7723 conditions was an address verification, or a call to a nested ACL, the
7724 messages generated by the underlying test were being passed through. This
7725 no longer happens. The underlying message is available in $acl_verify_
7726 message for both "message" and "log_message" expansions, so it can be
7727 passed through if needed.
7729 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7730 new expansion $bh_ to give the encoded byte string without charset
7731 translation. Translation happens only if iconv() is available; HAVE_ICONV
7732 indicates this at build time. HEADERS_CHARSET gives the charset to
7733 translate to; headers_charset can change it in the configuration, and
7734 "headers charset" can change it in an individual filter file.
7736 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7737 that creates RFC 2047 encoded "words" labels them as that charset instead
7738 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7739 expansion operator; (ii) when Exim creates a From: line for a local
7740 message; (iii) when a header line is rewritten to include a "phrase" part.
7742 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7743 buggy, causing it to skip the first lines of messages whose message ID
7744 ended in 'D'. This would not have bitten before Exim release 4.14, because
7745 message IDs were unlikely to end in 'D' before then. The effect was to have
7746 incorrect size information for certain domains.
7748 45. #include "config.h" was missing at the start of the crypt16.c module. This
7749 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7752 46. If there was a timeout during a "random" callout check, Exim treated it as
7753 a failure of the random address, and carried on sending RSET and the real
7754 address. If the delay was just some slowness somewhere, the response to the
7755 original RCPT would be taken as a response to RSET and so on, causing
7756 mayhem of various kinds.
7758 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7759 when I implemented it. It didn't allow for the fact that some option values
7760 may legitimately be negative (e.g. size_addition), and it didn't even do
7761 the right test for positive values.
7763 48. Domain names in DNS records are case-independent. Exim always looks them up
7764 in lower case. Some resolvers return domain names in exactly the case they
7765 appear in the zone file, that is, they may contain uppercase letters. Not
7766 all resolvers do this - some return always lower case. Exim was treating a
7767 change of case by a resolver as a change of domain, similar to a widening
7768 of a domain abbreviation. This triggered its re-routing code and so it was
7769 trying to route what was effectively the same domain again. This normally
7770 caused routing to fail (because the router wouldn't handle the domain
7771 twice). Now Exim checks for this case specially, and just changes the
7772 casing of the domain that it ultimately uses when it transmits the message
7775 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7778 50. If a filter generated a file delivery with a non-absolute name (possible if
7779 no home directory exists for the router), the forbid_file option was not
7782 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7783 the existing equality matching.
7785 52. Exim was using ints instead of ino_t variables in some places where it was
7786 dealing with inode numbers.
7788 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7789 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7790 called TMPDIR, and if it finds it is different, it changes its value.
7792 54. The smtp_printf() function is now made available to local_scan() so
7793 additional output lines can be written before returning. There is also an
7794 smtp_fflush() function to enable the detection of a dropped connection.
7795 The variables smtp_input and smtp_batched_input are exported to
7798 55. Changed the default runtime configuration: the message "Unknown user"
7799 has been removed from the ACL, and instead placed on the localuser router,
7800 using the cannot_route_message feature. This means that any verification
7801 failures that generate their own messages won't get overridden. Similarly,
7802 the "Unrouteable address" message that was in the ACL for unverifiable
7803 relay addresses has also been removed.
7805 56. Added hosts_avoid_esmtp to the smtp transport.
7807 57. The exicyclog script was not checking for the esoteric option
7808 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7809 will work only if exicyclog is run under the appropriate euid.
7811 58. Following a discussion on the list, the rules by which Exim recognises line
7812 endings on incoming messages have been changed. The -dropcr and drop_cr
7813 options are now no-ops, retained only for backwards compatibility. The
7814 following line terminators are recognized: LF CRLF CR. However, special
7815 processing applies to CR:
7817 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7818 nor a local message in the state where . is a terminator.
7820 (ii) If a bare CR is encountered in a header line, an extra space is added
7821 after the line terminator so as not to end the header. The reasoning
7822 behind this is that bare CRs in header lines are most likely either
7823 to be mistakes, or people trying to play silly games.
7825 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7826 was being incorrectly given as 18 bytes larger than it should have been.
7827 This is a VOB (very old bug).
7829 60. This may never have affected anything current, but just in case it has:
7830 When the local host is found other than at the start of a list of hosts,
7831 the local host, those with the same MX, and any that follow, are discarded.
7832 When the list in question was part of a longer list of hosts, the following
7833 hosts (not currently being processed) were also being discarded. This no
7834 longer happens. I'm not sure if this situation could ever has previously
7837 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7840 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7841 header. This is something that is recommended in a new Internet Draft, and
7842 is something that is documented as being done by Sendmail. There are two
7843 possible values. For messages generated by the autoreply transport, Exim
7846 Auto-Submitted: auto-replied
7848 whereas for all other generated messages (e.g. bounces) it adds
7850 Auto-Submitted: auto-generated
7852 63. The "personal" condition in filters now includes a test for the
7853 Auto-Submitted: header. If it contains the string "auto-" the message it
7854 not considered personal.
7856 64. Added rcpt_include_affixes as a generic transport option.
7858 65. Added queue_only_override (default true).
7860 66. Added the syslog_duplication option.
7862 67. If what should have been the first header line of a message consisted of
7863 a space followed by a colon, Exim was mis-interpreting it as a header line.
7864 It isn't of course - it is syntactically invalid and should therefore be
7865 treated as the start of the message body. The misbehaviour could have
7866 caused a number of strange effects, including loss of data in subsequent
7867 header lines, and spool format errors.
7869 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7870 client host had authenticated. This control can now be exercised by an ACL
7871 for more flexibility.
7873 69. By default, callouts do not happen when testing with -bh. There is now a
7874 variant, -bhc, which does actually run the callout code, including
7875 consulting and updating the callout cache.
7877 70. Added support for saslauthd authentication, courtesy of Alexander
7880 71. If statvfs() failed on the spool or log directories while checking their
7881 size for availability, Exim confusingly gave the error "space shortage".
7882 Furthermore, in debugging mode it crashed with a floating point exception.
7883 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7884 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7885 serious problem, Exim now writes to the main and panic logs when this
7886 happens, with details of the failure. It then refuses to accept the
7887 incoming message, giving the message "spool directory problem" or "log
7888 directory problem" with a 421 code for SMTP messages.
7890 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7891 0, 1, and 2 exist, because some OS complain for execs without them (see
7892 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7893 descriptors. However, the code omitted to check that the open succeeded,
7894 causing mysterious errors if for some reason the permissions on /dev/null
7895 got screwed. Now Exim writes a message to the main and panic logs, and
7896 bombs out if it can't open /dev/null.
7898 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7899 interact so that it is all more flexible. It is supposed to remain
7900 backwards compatible. Also added extra_local_interfaces.
7902 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7903 to bomb out with an assertion failure - to the client this appears as a
7904 connection drop. This problem occurs in the part of the code that was taken
7905 from the Samba project. Fortunately, the assertion is in a very simple
7906 function, so I have fixed this by reproducing the function inline in the
7907 one place where it is called, and arranging for authentication to fail
7908 instead of killing the process with assert().
7910 75. The SPA client code was not working when the server requested OEM rather
7911 than Unicode encoding.
7913 76. Added code to make require_files with a specific uid setting more usable in
7914 the case where statting the file as root fails - usually a non-root-mounted
7915 NFS file system. When this happens and the failure is EACCES, Exim now
7916 forks a subprocess and does the per-uid checking as the relevant uid.
7918 77. Added process_log_path.
7920 78. If log_file_path was not explicitly set, a setting of check_log_space or
7921 check_log_inodes was ignored.
7923 79. If a space check for the spool or log partitions fails, the incident is now
7924 logged. Of course, in the latter case the data may get lost...
7926 80. Added the %p formatting code to string_format() so that it can be used to
7927 print addresses in debug_print(). Adjusted all the address printing in the
7928 debugging in store.c to use %p rather than %d.
7930 81. There was a concern that a line of code in smtp_in.c could overflow a
7931 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7932 initially expressed, the concern was not well-founded, because trailing
7933 spaces are removed early. However, if the trailing spaces were followed by
7934 a NULL, they did not get removed, so the overflow was possible. Two fixes
7937 (a) I re-wrote the offending code in a cleaner fashion.
7938 (b) If an incoming SMTP command contains a NULL character, it is rejected
7941 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7942 runs initgroups(), so that if the Exim user is in any additional groups,
7943 they will be used during message reception.
7949 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob