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
109 JH/01 Avoid costly startup code when not strictly needed. This reduces time
110 for some exim process initialisations. It does mean that the logging
111 of TLS configuration problems is only done for the daemon startup.
113 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
115 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
116 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
119 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
120 DKIM signing is being done. Previously a transport_filter would always
121 disable CHUNKING, falling back to traditional DATA.
123 JH/05 Regard command-line receipients as tainted.
125 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
127 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
128 PAM library frees one of the arguments given to it, despite the
129 documentation. Therefore a plain malloc must be used.
131 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
132 on-stack buffers were used, resulting in a taint trap when DSN information
133 copied from a received message was written into the buffer.
135 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
136 the ordering of its ARC headers. This caused a crash.
138 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
139 a new record was being constructed with information from the peer, a trap
142 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
143 installation would get error messages from DMARC verify, when it hit the
144 nonexistent file indicated by the default. Distros wanting DMARC enabled
145 should both provide the file and set the option.
146 Also enforce no DMARC verification for command-line sourced messages.
148 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
149 could, depending on the platform, hang at the STARTTLS response.
151 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
152 message on a connection. Previously if one message had ARC headers and
153 the following one did not, a crash could result when adding an
154 Authentication-Results: header.
156 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
157 Exim main code and Exim-related utities. The introduction of taint
158 tracking also did many adjustments to string handling. Since then, eximon
159 frequently terminated with an assert failure.
161 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
162 check for 452 responses. This slightly helps the inefficieny of doing
163 a large alias-expansion into a recipient-limited target. The max_rcpt
164 transport option still applies (and at the current default, will override
165 the new feature). The check is done for either cause of synch, and forces
166 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
167 connection. The new facility is not tunable at this time.
169 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
170 library live data was being used, so the results became garbage. Make
171 copies while it is still usable.
173 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
174 on delivery deferred (==) and failed (**) lines (if a delivery was
175 attemtped). Previously it was only on completion (=>) lines.
177 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
178 for the expansion of the server_scram_iter and server_scram_salt options.
180 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
181 are now specifically given a NO_DATA response without hitting the system
182 resolver. The library goes on to do the now-standard TXT lookup.
183 Use of dnsdb lookups is not affected.
185 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
186 only retrieve the errormessage once. Previously two calls to dlerror()
187 were used, and the second one (for mainlog/paniclog) retrieved null
190 JH/20 Taint checking: disallow use of tainted data for
191 - the appendfile transport file and directory options
192 - the pipe transport command
193 - the autoreply transport file, log and once options
194 - file names used by the redirect router (including filter files)
196 - paths used by single-key lookups
197 Previously this was permitted.
199 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
200 adjusted the size of a major service buffer; this failed because the
201 buffer was in use at the time. Change to a compile-time increase in the
202 buffer size, when this authenticator is compiled into exim.
204 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
205 previous fast-mode was untenable in the face of glibs using mmap to
206 support larger malloc requests.
208 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
209 New values supported, if defined on system where compiled:
210 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
211 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
213 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
214 running a limited number of proceses in parallel, a benefit is gained. The
215 amount varies with the platform hardware and load. The use of the option
216 queue_run_in_order means we cannot do this, as ordering becomes
219 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
220 had introduced a string-copy (for ensuring NUL-termination) which was not
221 appropriate for that case, which can include embedded NUL bytes in the
222 block of data. Investigation showed the copy to actually be needless, the
223 data being length-specified.
225 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
226 done during a receiving connection, and both used TLS, global info was
227 used rather than per-connection info for tracking the state of data
228 queued for transmission. This could result in a connection hang.
230 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
231 Previously, when delivering serveral messages down a single connection
232 only the first would provide a SIZE. This was due to the size information
233 not being properly tracked.
235 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
236 TAI (at 37 seconds currently), pretend to be in UTC for time-related
237 expansion and logging. Previously, spurious values such as a future
238 minute could be seen.
240 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
241 it could crash from a null-deref. This could also affect the
242 ${addresses: } operator and ${readsock } item.
244 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
245 message following a mime one, the variable was not reset.
247 JH/30 When an pipelined-connect fails at the first response, assume incorrect
248 cached capability (perhaps the peer reneged?) and immediately retry in
251 JH/31 Fix spurious detection of timeout while writing to transport filter.
253 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
254 an attempt to copy the string was made before checking it.
256 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
257 taint of the lookup key was maintained; we now regard the presence in the
258 filesystem as sufficient validation.
260 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
261 argument is supplied.
263 JH/35 The dsearch lookup now requires that the directory is an absolute path.
264 Previously this was not checked, and nonempty relative paths made an
265 access under Exim's current working directory.
267 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
268 Previously no event was raised.
270 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
271 parameter supplied by the sender MAIL FROM command. Previously it was
272 ignored, and only the check_spool_space option value for the required
275 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
276 the size of the signing public-key. Previously it was instead giving
277 the size of the signature hash.
279 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
280 the default. See the (new) dkim_verify_min_keysizes option.
282 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
283 and an ACL use a lookup for checking either the local_part or domain,
284 stale data could be accessed. Ensure that variable references are
285 dropped between messages.
287 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
288 by the client was not checked as pointing within response data before
289 being used. A malicious client could thus cause an out-of-bounds read and
290 possibly gain authentication. Fix by adding the check.
292 JH/42 Internationalisation: change the default for downconversion in the smtp
293 transport to be "if needed". Previously it was "as previously set" for
294 the message, which usually meant "if needed" for message-submission but
295 "no" for everything else. However, MTAs have been seen using SMTPUTF8
296 even when the envelope addresses did not need it, resulting in forwarding
297 failures to non-supporting MTAs. A downconvert in such cases will be
298 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
299 transport. The change does mean that addresses needing conversion will
300 be converted when previously a delivery failure would occur.
302 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
303 response was received it would be used unchecked in a fail-DSN, violating
304 standards on line-length limits. Truncate if needed.
306 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
307 doesn't work. (Additionally add a new main config option to configure the
314 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
315 side, to decode a TLS 1.3 packet capture.
317 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
318 Previously the default library behaviour applied, sending two, each in
321 JH/03 Debug output for ACL now gives the config file name and line number for
324 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
326 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
328 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
329 buffer overrun for (non-chunking) other transports.
331 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
332 TLS1.3, means that a server rejecting a client certificate is not visible
333 to the client until the first read of encrypted data (typically the
334 response to EHLO). Add detection for that case and treat it as a failed
335 TLS connection attempt, so that the normal retry-in-clear can work (if
336 suitably configured).
338 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
339 and/or domain. Found and fixed by Jason Betts.
341 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
342 configuration). If a CNAME target was not a wellformed name pattern, a
345 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
346 the OS reports them interleaved with other addresses.
348 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
349 used both for input and for a verify callout, both encrypted, SMTP
350 responses being sent by the server could be lost. This resulted in
351 dropped connections and sometimes bounces generated by a peer sending
354 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
355 string. Previously it was possible to cause undefined behaviour in a
356 library routine (usually a crash). Found by "zerons".
358 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
361 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
362 API was removed, so update to use the newer ones.
364 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
365 any timeout set, is taking a long time. Previously we would hang on to a
366 rotated logfile "forever" if the input was arriving with long gaps
367 (a previous attempt to fix addressed lack, for a long time, of initial
370 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
371 shared (NFS) environment. The length of the tempfile name is now
372 4 + 16 ("hdr.$message_exim_id") which might break on file
373 systems which restrict the file name length to lower values.
376 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
377 shared (NFS) environment.
379 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
380 did for all versions <4.90). Notably -M, -m, --invert, -I may be
383 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
384 on some platforms for bit 31.
386 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
387 to changes apparently associated with TLS1.3 handling some of the APIs
388 previously used were either nonfunctional or inappropriate. Strings
389 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
390 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
391 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
392 This affects log line X= elements, the $tls_{in,out}_cipher variables,
393 and the use of specific cipher names in the encrypted= ACL condition.
395 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
397 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
398 verification result was not updated unless hosts_require_ocsp applied.
400 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
401 queue_list_requires_admin set to false, non-admin users were denied the
404 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
405 directory-of-certs mode. Previously they were advertised despite the
408 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
409 A single TCP connection by a client will now hold a TLS connection open
410 for multiple message deliveries, by default. Previously the default was to
413 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
414 default. If built with the facility, DANE will be used. The facility
415 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
417 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
418 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
419 must be defined and you must still, unless you define DISABLE_TLS, manage
420 the the include-dir and library-file requirements that go with that
421 choice. Non-TLS builds are still supported.
423 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
426 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
427 default. If the platform supports and has the facility enabled, it will
428 be requested on all coneections.
430 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
431 controlled by the build-time option SUPPORT_PIPE_CONNECT.
433 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
435 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
436 success-DSN messages. Previously the From: header was always the default
437 one for these; the option was ignored.
439 JH/28 Fix the timeout on smtp response to apply to the whole response.
440 Previously it was reset for every read, so a teergrubing peer sending
441 single bytes within the time limit could extend the connection for a
442 long time. Credit to Qualsys Security Advisory Team for the discovery.
444 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
445 delivery address, which leaked information of the results of local
446 forwarding. Change to the original envelope recipient address, per
449 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
450 requested. Previously not bounce was generated and a log entry of
451 error ignored was made.
453 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
455 JH/32 Introduce a general tainting mechanism for values read from the input
456 channel, and values derived from them. Refuse to expand any tainted
457 values, to catch one form of exploit.
459 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
460 was unused and the unexpanded text used for the test. Found and
461 fixed by Ruben Jenster.
463 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
464 an attempt to use a TLS library read routine dereffed a nul pointer,
467 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
468 them in our smtp response.
470 JH/36 Have the general router option retry_use_local_part default to true when
471 any of the restrictive preconditions are set (to anything). Previously it
472 was only for check_local user. The change removes one item of manual
473 configuration which is required for proper retries when a remote router
474 handles a subset of addresses for a domain.
476 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
477 link count into consideration.
479 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
480 caused the extension of big_buffer, the following lines were ignored.
482 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
483 accordance with RFC 2308. Previously there was no expiry, so a longlived
484 receive process (eg. due to ACL delays) versus a short SOA value could
487 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
489 JH/39 Promote DMARC support to mainline.
491 JH/40 Bug 2452: Add a References: header to DSNs.
493 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
494 parameters. The relevant library call is documented as "Deprecated: This
495 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
496 3.6.0, DH parameters are negotiated following RFC7919."
498 HS/06 Change the default of dnssec_request_domains to "*"
500 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
501 carried on and emitted a BDAT command, even when PIPELINING was not
504 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
505 buffer was used for the filename, resulting in a trap when tainted
506 arguments (eg. $domain) were used.
508 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
509 recommended to avoid a possible server-load attack. The feature can be
510 re-enabled via the openssl_options main cofiguration option.
512 JH/45 local_scan API: documented the current smtp_printf() call. This changed
513 for version 4.90 - adding a "more data" boolean to the arguments.
514 Bumped the ABI version number also, this having been missed previously;
515 release versions 4.90 to 4.92.3 inclusive were effectively broken in
516 respect of usage of smtp_printf() by either local_scan code or libraries
517 accessed via the ${dlfunc } expansion item. Both will need coding
518 adjustment for any calls to smtp_printf() to match the new function
519 signature; a FALSE value for the new argument is always safe.
521 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
522 the file-offset (which the Linux syscall does, and exim expects); this
523 resulted in an indefinite loop.
525 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
526 to do ARC verification. The Authentication-Results: header line added
527 by the configuration then had no ARC item.
533 JH/01 Remove code calling the customisable local_scan function, unless a new
534 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
536 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
537 non-signal-safe functions being used.
539 JH/03 Bug 2269: When presented with a received message having a stupidly large
540 number of DKIM-Signature headers, disable DKIM verification to avoid
541 a resource-consumption attack. The limit is set at twenty.
543 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
544 report of oldest_pass in ${authres } in consequence, and separate out
545 some descriptions of reasons for verification fail.
547 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
548 files in the spool were present and unlocked. A queue-runner could spot
549 them, resulting in a duplicate delivery. Fix that by doing the unlock
550 after the unlink. Investigation by Tim Stewart. Take the opportunity to
551 add more error-checking on spoolfile handling while that code is being
554 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
555 No known attacks, no CVE, this is defensive hardening.
557 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
558 a queue-runner could start a delivery while other operations were ongoing.
559 Cutthrough delivery was a common victim, resulting in duplicate delivery.
560 Found and investigated by Tim Stewart. Fix by using the open message data
561 file handle rather than opening another, and not locally closing it (which
562 releases a lock) for that case, while creating the temporary .eml format
563 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
565 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
566 $sender_verify_failure/$recipient_verify_failure to "random".
568 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
571 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
572 Previously this would segfault.
574 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
577 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
578 like zero, since the resolver should be doing this for us, But we need one
579 as a CNAME but no MX presence gets the CNAME returned; we need to check
580 that doesn't point to an MX to declare it "no MX returned" rather than
581 "error, loop". A new main option is added so the older capability of
582 following some limited number of chain links is maintained.
584 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
586 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
587 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
588 now-deprecated earlier definitions used only specified the range up to TLS
589 1.2 (in the older-version library docs).
591 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
593 JH/15 Rework TLS client-side context management. Stop using a global, and
594 explicitly pass a context around. This enables future use of TLS for
595 connections to service-daemons (eg. malware scanning) while a client smtp
596 connection is using TLS; with cutthrough connections this is quite likely.
598 JH/16 Fix ARC verification to do AS checks in reverse order.
600 JH/17 Support a "tls" option on the ${readsocket } expansion item.
602 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
603 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
604 Previously the "utf8" would be re-prepended for every additional message.
606 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
607 Previously thery were accepted, resulting in issues when attempting to
608 forward messages to a non-supporting MTA.
610 PP/02 Let -n work with printing macros too, not just options.
612 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
613 one parent address was copied, and bogus data was used at delivery-logging
614 time. Either a crash (after delivery) or bogus log data could result.
615 Discovery and analysis by Tim Stewart.
617 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
618 Previously if the string ended mid-character, we did not insert the
619 promised '?' replacement.
621 PP/04 Documentation: current string operators work on bytes, not codepoints.
623 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
624 should pack well giving a smaller memory footprint so better caching and
625 therefore performance. Group the declarations where this can't be done so
626 that the byte-sized flag variables are not interspersed among pointer
627 variables, giving a better chance of good packing by the compiler.
629 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
630 non-null, to avoid issues with sites running BATV. Previously reports were
631 sent with an empty envelope sender so looked like bounces.
633 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
634 The ignore_error flag wasn't being returned from the filter subprocess so
635 was not set for later routers. Investigation and fix by Matthias Kurz.
637 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
638 and a msg:complete for the whole, when a message is manually removed using
639 -Mrm. Developement by Matthias Kurz, hacked on by JH.
641 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
642 a "Gnu special" function, asprintf() in the DB utility binary builds; I
643 hope that is portable enough.
645 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
646 requiring a known-CA anchor certificate; make it now rely entirely on the
647 TLSA as an anchor. Checking the name on the leaf cert against the name
648 on the A-record for the host is still done for TA (but not for EE mode).
650 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
651 included in delivery lines for non-proxied connections, when compiled with
652 SUPPORT_SOCKS and running with proxy logging enabled.
654 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
655 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
656 move the existing event to fire before the normal logging of message
657 failure so that custom logging is bracketed by normal logging.
659 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
660 msg:fail:internal event. Developement by Matthias Kurz.
662 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
663 far too small for todays use of crypto signatures stored there. Go all
664 the way to the max DNS message size of 64kB, even though this might be
665 overmuch for IOT constrained device use.
667 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
668 copy a string over itself. The library routine is documented as not
669 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
671 JH/32 For main options check_spool_space and check_inode_space, where the
672 platform supports 64b integers, support more than the previous 2^31 kB
673 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
674 the previous G, M, k.
676 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
677 $authenticated_fail_id variable on authentication failure. Previously
680 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
681 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
682 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
683 GNUTLS_SEC_PARAM_MEDIUM.
685 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
686 side. Previously we would continue as if no SNI had been received.
688 JH/36 Harden the handling of string-lists. When a list consisted of a sole
689 "<" character, which should be a list-separator specification, we walked
690 off past the nul-terimation.
692 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
693 causes) even when the retry time is not yet met. Previously they were
694 not, meaning that when (say) an account was over-quota and temp-rejecting,
695 and multiple senders' messages were queued, only one sender would get
696 notified on each configured delay_warning cycle.
698 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
700 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
701 cipher-suites, an error can be left on the stack even for a succeeding
702 accept; this results in impossible error messages when a later operation
705 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
706 return error codes indicating retry. Under TLS1.3 this becomes required.
708 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
709 it only wrote the new authenticators, resulting in a lack of tracking of
710 peer changes of ESMTP extensions until the next cache flush.
712 JH/41 Fix the loop reading a message header line to check for integer overflow,
713 and more-often against header_maxsize. Previously a crafted message could
714 induce a crash of the recive process; now the message is cleanly rejected.
716 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
717 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
723 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
724 When redis_servers is set to a list of > 1 element, and the Redis servers
725 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
726 case of MOVED into a DEFER case instead, thus moving the query onto the
727 next server in the list. For a cluster of N elements, all N servers must
728 be defined in redis_servers.
730 GF/02 Catch and remove uninitialized value warning in exiqsumm
731 Check for existence of @ARGV before looking at $ARGV[0]
733 JH/01 Replace the store_release() internal interface with store_newblock(),
734 which internalises the check required to safely use the old one, plus
735 the allocate and data copy operations duplicated in both (!) of the
736 extant use locations.
738 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
739 modifier. This matches the restriction on the commandline.
741 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
742 Previously only the last row was returned.
744 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
745 we assumed that tags in the header were well-formed, and parsed the
746 element content after inspecting only the first char of the tag.
747 Assumptions at that stage could crash the receive process on malformed
750 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
751 While running the DKIM ACL we operate on the Permanent memory pool so that
752 variables created with "set" persist to the DATA ACL. Also (at any time)
753 DNS lookups that fail create cache records using the Permanent pool. But
754 expansions release any allocations made on the current pool - so a dnsdb
755 lookup expansion done in the DKIM ACL releases the memory used for the
756 DNS negative-cache, and bad things result. Solution is to switch to the
757 Main pool for expansions.
758 While we're in that code, add checks on the DNS cache during store_reset,
759 active in the testsuite.
760 Problem spotted, and debugging aided, by Wolfgang Breyha.
762 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
763 When none of the hosts presented to a transport match an already-open
764 connection, close it and proceed with the list. Previously we would
765 queue the message. Spotted by Lena with Yahoo, probably involving
768 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
769 Previously a spurious "250 OK id=" response was appended to the proper
772 JH/08 The "support for" informational output now, which built with Content
773 Scanning support, has a line for the malware scanner interfaces compiled
774 in. Interface can be individually included or not at build time.
776 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
777 by the template makefile "src/EDITME". The "STREAM" support for an older
778 ClamAV interface method is removed.
780 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
781 rows affected is given instead).
783 JH/11 The runtime Berkeley DB library version is now additionally output by
784 "exim -d -bV". Previously only the compile-time version was shown.
786 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
787 SMTP connection. Previously, when one had more receipients than the
788 first, an abortive onward connection was made. Move to full support for
789 multiple onward connections in sequence, handling cutthrough connection
790 for all multi-message initiating connections.
792 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
793 routers. Previously, a multi-recipient message would fail to match the
794 onward-connection opened for the first recipient, and cause its closure.
796 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
797 a timeout on read on a GnuTLS initiating connection, resulting in the
798 initiating connection being dropped. This mattered most when the callout
799 was marked defer_ok. Fix to keep the two timeout-detection methods
802 JH/15 Relax results from ACL control request to enable cutthrough, in
803 unsupported situations, from error to silently (except under debug)
804 ignoring. This covers use with PRDR, frozen messages, queue-only and
807 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
809 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
810 metadata, resulting in a crash in free().
812 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
813 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
814 Broken also in d185889f4, with init system revamp.
816 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
817 Previously we abruptly closed the connection after reading a malware-
818 found indication; now we go on to read the "scan ok" response line,
821 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
822 ACL. Previously, a crash would result.
824 JH/19 Speed up macro lookups during configuration file read, by skipping non-
825 macro text after a replacement (previously it was only once per line) and
826 by skipping builtin macros when searching for an uppercase lead character.
828 JH/20 DANE support moved from Experimental to mainline. The Makefile control
829 for the build is renamed.
831 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
832 was allocated for every new TLS startup, meaning one per message. Fix
833 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
835 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
836 reported the original. Fix to report (as far as possible) the ACL
837 result replacing the original.
839 JH/23 Fix memory leak during multi-message connections using STARTTLS under
840 OpenSSL. Certificate information is loaded for every new TLS startup,
841 and the resources needed to be freed.
843 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
845 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
848 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
849 DATA response info to the (existing) per-recipient response info for
850 the "C=" log element. It can have useful tracking info from the
851 destination system. Patch from Simon Arlott.
853 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
854 length value. Previously this would segfault.
856 HS/02 Support Avast multiline protoocol, this allows passing flags to
857 newer versions of the scanner.
859 JH/28 Ensure that variables possibly set during message acceptance are marked
860 dead before release of memory in the daemon loop. This stops complaints
861 about them when the debug_store option is enabled. Discovered specifically
862 for sender_rate_period, but applies to a whole set of variables.
863 Do the same for the queue-runner and queue-list loops, for variables set
864 from spool message files. Do the same for the SMTP per-message loop, for
865 certain variables indirectly set in ACL operations.
867 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
868 as a multi-recipient message from a mailinglist manager). The coding had
869 an arbitrary cutoff number of characters while checking for more input;
870 enforced by writing a NUL into the buffer. This corrupted long / fast
871 input. The problem was exposed more widely when more pipelineing of SMTP
872 responses was introduced, and one Exim system was feeding another.
873 The symptom is log complaints of SMTP syntax error (NUL chars) on the
874 receiving system, and refused recipients seen by the sending system
875 (propating to people being dropped from mailing lists).
876 Discovered and pinpointed by David Carter.
878 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
879 replaced by the ${authresults } expansion.
881 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
883 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
884 allows proper process termination in container environments.
886 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
887 Previously the "final dot" had a newline after it; ensure it is CR,LF.
889 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
890 and "err_perm", deprecated since 4.83 when the RFC-defined words
891 "temperror" and "permerror" were introduced.
893 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
894 transport-filters or DKIM-signing. The restriction was lost in the
895 consolidation of verify-callout and delivery SMTP handling.
896 Extend the restriction to also cover ARC-signing.
898 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
899 in defer=pass mode supply a 450 to the initiator. Previously the message
902 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
903 tls_require_ciphers is used as before.
905 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
906 "pass_unscanned". Only tmpfails from the scanner are written to
907 the paniclog, as they may require admin intervention (permission
908 denied, license issues). Other scanner errors (like decompression
909 bombs) do not cause a paniclog entry.
911 JH/36 Fix reinitialisation of DKIM logging variable between messages.
912 Previously it was possible to log spurious information in receive log
915 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
916 triggered odd behaviour from Outlook Express clients.
918 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
921 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
922 since the IETF WG has not yet settled on that versus the original
923 "bare" representation.
925 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
926 Previously the millisecond value corrupted the output.
927 Fix also for syslog_pid=no and log_selector +pid, for which the pid
928 corrupted the output.
934 JH/01 Rework error string handling in TLS interface so that the caller in
935 more cases is responsible for logging. This permits library-sourced
936 string to be attached to addresses during delivery, and collapses
937 pairs of long lines into single ones.
939 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
940 during configuration. Wildcards are allowed and expanded.
942 JH/02 Rework error string handling in DKIM to pass more info back to callers.
943 This permits better logging.
945 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
946 do not close it down and have the child transport start it up again on
947 the passed-on TCP connection. Instead, proxy the child (and any
948 subsequent ones) for TLS via a unix-domain socket channel. Logging is
949 affected: the continued delivery log lines do not have any DNSSEC, TLS
950 Certificate or OCSP information. TLS cipher information is still logged.
952 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
953 identical IP addresses on different listening ports. Will also affect
956 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
957 add noisy ifdef guards to special-case this sillyness.
958 Patch from Bernd Kuhls.
960 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
961 than 255 are no longer allowed.
963 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
964 Disable the session-cache too, which might reduce our load. Since we
965 currrectly use a new context for every connection, both as server and
966 client, there is no benefit for these.
967 GnuTLS appears to not support tickets server-side by default (we don't
968 call gnutls_session_ticket_enable_server()) but client side is enabled
969 by default on recent versions (3.1.3 +) unless the PFS priority string
972 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
973 <https://reproducible-builds.org/specs/source-date-epoch/>.
975 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
976 the check for any unsuccessful recipients did not notice the limit, and
977 erroneously found still-pending ones.
979 JH/08 Pipeline CHUNKING command and data together, on kernels that support
980 MSG_MORE. Only in-clear (not on TLS connections).
982 JH/09 Avoid using a temporary file during transport using dkim. Unless a
983 transport-filter is involved we can buffer the headers in memory for
984 creating the signature, and read the spool data file once for the
985 signature and again for transmission.
987 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
988 4.77 as the kernel support then wasn't solid, having issues in 64bit
989 mode. Now, it's been long enough. Add support for FreeBSD also.
991 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
992 case where the routing stage had gathered several addresses to send to
993 a host before calling the transport for the first, we previously failed
994 to close down TLS in the old transport process before passing the TCP
995 connection to the new process. The new one sent a STARTTLS command
996 which naturally failed, giving a failed delivery and bloating the retry
997 database. Investigation and fix prototype from Wolfgang Breyha.
999 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1000 false-negatives in the check that the sender had not preempted a response
1001 or prompt from Exim (running as a server), due to that code's lack of
1002 awareness of the SMTP input buffering.
1004 PP/04 Add commandline_checks_require_admin option.
1005 Exim drops privileges sanely, various checks such as -be aren't a
1006 security problem, as long as you trust local users with access to their
1007 own account. When invoked by services which pass untrusted data to
1008 Exim, this might be an issue. Set this option in main configuration
1009 AND make fixes to the calling application, such as using `--` to stop
1012 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1013 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1014 responses to those, into a single TLS record each way (this usually means
1015 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1018 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1019 affects you only if you're dancing at the edge of the param size limits.
1020 If you are, and this message makes sense to you, then: raise the
1021 configured limit or use OpenSSL 1.1. Nothing we can do for older
1024 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1025 cmdline element to get the documented default one. Previously it was
1028 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1029 are unsuable for non-message-specific reasons - eg. connection timeout,
1030 banner-time rejection.
1032 JH/16 Fix logging of delivery remote port, when specified by router, under
1035 PP/06 Repair manualroute's ability to take options in any order, even if one
1036 is the name of a transport.
1039 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1041 JH/17 Change the list-building routines interface to use the expanding-string
1042 triplet model, for better allocation and copying behaviour.
1044 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1045 Previously it was constructed the first time a possibly-matching string
1046 was met in the configuration file input during startup; now it is done
1049 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1050 compatible one, to avoid the (poorly documented) possibility of a config
1051 file in the working directory redirecting the DB files, possibly correpting
1052 some existing file. CVE-2017-10140 assigned for BDB.
1054 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1055 cache-hot. Previously, although the result was properly cached, the
1056 initial verify call returned a defer.
1058 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1059 the main verify for receipient in uncached-mode.
1061 JH/22 Retire historical build files to an "unsupported" subdir. These are
1062 defined as "ones for which we have no current evidence of testing".
1064 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1065 if present. Previously it was ignored.
1067 JH/24 Start using specified-initialisers in C structure init coding. This is
1068 a C99 feature (it's 2017, so now considered safe).
1070 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1071 if was a fixed-sized field and bitmask ops via macros; it is now more
1074 PP/07 GitHub PR 56: Apply MariaDB build fix.
1075 Patch provided by Jaroslav Škarvada.
1077 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1078 during Coverity cleanups [4.87 JH/47]
1079 Diagnosis and fix provided by Michael Fischer v. Mollard.
1081 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1082 the right size to place the terminating semicolon on its own folded
1083 line, the header hash was calculated to an incorrect value thanks to
1084 the (relaxed) space the fold became.
1086 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1087 and confused the parent.
1089 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1090 which could crash as a result. This could lead to undeliverable messages.
1092 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1095 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1096 matters on fast-turnover and PID-randomising systems, which were getting
1097 out-of-order delivery.
1099 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1100 a possibly-overlapping copy. The symptom was that "Remote host closed
1101 connection in response to HELO" was logged instead of the actual 4xx
1104 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1105 Previously only that bufferd was discarded, resulting in SYMTP command
1108 JH/32 DKIM: when a message has multiple signatures matching an identity given
1109 in dkim_verify_signers, run the dkim acl once for each. Previously only
1110 one run was done. Bug 2189.
1112 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1113 "panic the current process" to "deliberately defer". The panic log is
1114 still written with the problem list name; the mail and reject logs now
1115 get a temp-reject line for the message that was being handled, saying
1116 something like "domains check lookup or other defer". The SMTP 451
1117 message is still "Temporary local problem".
1119 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1120 A crafted sequence of BDAT commands could result in in-use memory beeing
1121 freed. CVE-2017-16943.
1123 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1124 from SMTP input. Previously it was always done; now only done for DATA
1125 and not BDAT commands. CVE-2017-16944.
1127 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1128 to the message (such as an overlong header line). Previously this was
1129 not done and we did not exit BDAT mode. Followon from the previous item
1130 though a different problem.
1136 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1137 than -2003 did; needs libidn2 in addition to libidn.
1139 JH/02 The path option on a pipe transport is now expanded before use.
1141 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1142 Patch provided by "Björn", documentation fix added too.
1144 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1145 missing a wire-to-host endian conversion.
1147 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1148 close after a BDAT command line could be taken as a following command,
1149 giving a synch failure. Fix by only checking for synch immediately
1150 before acknowledging the chunk.
1152 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1153 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1154 macro. Patches provided by Josh Soref.
1156 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1157 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1158 by IANA as a service extension. However, John Klensin suggests that we
1161 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1162 the dkim code may be unix-mode line endings rather than smtp wire-format
1163 CRLF, so prepend a CR to any bare LF.
1165 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1166 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1168 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1169 the first were themselves being wrongly included in the feed into dkim
1170 processing; with most chunk sizes in use this resulted in an incorrect
1171 body hash calculated value.
1173 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1174 DKIM signature block, for verification. Although advised against by
1175 standards it is specifically not ruled illegal.
1177 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1179 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1180 missing a body hash (the bh= tag).
1182 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1183 It seems that HAProxy sends the Proxy Protocol information in clear and
1184 only then does a TLS startup, so do the same.
1186 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1187 TCP connections (such as for Spamd) unless the daemon successfully set
1188 Fast Open mode on its listening sockets. This fixes breakage seen on
1189 too-old kernels or those not configured for Fast Open, at the cost of
1190 requiring both directions being enabled for TFO, and TFO never being used
1191 by non-daemon-related Exim processes.
1193 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1194 endings, at least on the first header line. Try to canonify any that get
1195 past that check, despite the cost.
1197 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1198 now limited to an arbitrary five deep, while parsing addresses with the
1199 strip_excess_angle_brackets option enabled.
1201 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1202 instead leave the unprompted TLS handshake in socket buffer for the
1203 TLS library to consume.
1205 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1207 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1209 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1210 region in one operation, for speed, and this leaves assigned pointers
1211 dangling. Add checks run only under the testsuite which checks all
1212 variables at a store-reset and panics on a dangling pointer; add code
1213 explicitly nulling out all the variables discovered. Fixes one known
1214 bug: a transport crash, where a dangling pointer for $sending_ip_address
1215 originally assigned in a verify callout, is re-used.
1217 PP/06 Drop '.' from @INC in various Perl scripts.
1219 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1221 PP/08 Reduce a number of compilation warnings under clang; building with
1222 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1223 should be warning-free.
1225 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1227 HS/01 Fix portability problems introduced by PP/08 for platforms where
1228 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1230 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1231 chunk. This allows us to accept broken chunked messages. We need a more
1232 general solution here.
1234 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1235 already-broken messages in the queue.
1237 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1239 JH/19 Fix reference counting bug in routing-generated-address tracking.
1245 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1246 supports it and a size is available (ie. the sending peer gave us one).
1248 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1249 years of being deprecated). The replacements are the ACLs
1250 acl_smtp_mime and acl_not_smtp_mime.
1252 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1253 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1254 or even an in-clear connection were permitted. Now, if the host lookup
1255 was dnssec and dane was requested then the host is only used if the
1256 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1257 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1258 if one fails this test.
1259 This means that a poorly-configured remote DNS will make it incommunicado;
1260 but it protects against a DNS-interception attack on it.
1262 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1263 non-noisy when a race steals the message being considered.
1265 JH/05 If main configuration option tls_certificate is unset, generate a
1266 self-signed certificate for inbound TLS connections.
1268 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1269 in rewrites and routers.
1271 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1272 and logged a warning sing 4.83; now they are a configuration file error.
1274 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1275 (lacking @domain). Apply the same qualification processing as RCPT.
1277 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1279 JH/10 Support ${sha256:} applied to a string (as well as the previous
1282 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1283 a cutthrough deliver is pending, as we always want to make a connection.
1284 This also avoids re-routing the message when later placing the cutthrough
1285 connection after a verify cache hit.
1286 Do not update it with the verify result either.
1288 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1289 when routing results in more than one destination address.
1291 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1292 signing (which inhibits the cutthrough capability). Previously only
1293 the presence of an option was tested; now an expansion evaluating as
1294 empty is permissible (obviously it should depend only on data available
1295 when the cutthrough connection is made).
1297 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1298 the relevant preceding SMTP command did not note the pipelining mode.
1300 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1301 Previously they were not counted.
1303 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1304 as one having no matching records. Previously we deferred the message
1305 that needed the lookup.
1307 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1308 distinguished as "(=".
1310 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1311 for missing MX records. Previously it only worked for missing A records.
1313 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1315 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1316 after the data-go-ahead and data-ack. Patch from Jason Betts.
1318 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1319 even for a "none" policy. Patch from Tony Meyer.
1321 JH/22 Fix continued use of a connection for further deliveries. If a port was
1322 specified by a router, it must also match for the delivery to be
1325 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1326 When one of the recipients of a message was unsuitable for the connection
1327 (has no matching addresses), we lost track of needing to mark it
1328 deferred. As a result mail would be lost.
1330 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1332 JH/25 Decoding ACL controls is now done using a binary search; the source code
1333 takes up less space and should be simpler to maintain. Merge the ACL
1334 condition decode tables also, with similar effect.
1336 JH/26 Fix problem with one_time used on a redirect router which returned the
1337 parent address unchanged. A retry would see the parent address marked as
1338 delivered, so not attempt the (identical) child. As a result mail would
1341 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1342 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1343 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1346 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1347 Main config options check_{log,spool}_{inodes,space} are now
1348 100 inodes, 10MB unless set otherwise in the configuration.
1350 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1351 Previously it only applied to the main-section connection policy
1354 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1356 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1357 by me. Added RFC7919 DH primes as an alternative.
1359 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1360 are not in the system include path.
1362 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1363 GnuTLS, when a session startup failed (eg because the client disconnected)
1364 Exim did stdio operations after fclose. This was exposed by a recent
1365 change which nulled out the file handle after the fclose.
1367 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1368 signed directly by the cert-signing cert, rather than an intermediate
1369 OCSP-signing cert. This is the model used by LetsEncrypt.
1371 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1373 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1374 an incoming connection.
1376 HS/02 Bug 1802: Do not half-close the connection after sending a request
1379 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1380 fallback to "prime256v1".
1382 JH/34 SECURITY: Use proper copy of DATA command in error message.
1383 Could leak key material. Remotely exploitable. CVE-2016-9963.
1389 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1390 and 3.4.4 - once the server is enabled to respond to an OCSP request
1391 it does even when not requested, resulting in a stapling non-aware
1392 client dropping the TLS connection.
1394 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1395 support variable-length bit vectors. No functional change.
1397 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1398 The I= interface field on outgoing lines is now after the H= remote
1399 host field, same as incoming lines. There is a separate
1400 outgoing_interface log selector which allows you to disable the
1403 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1404 If not running log_selector +smtp_connection the mainlog would be held
1405 open indefinitely after a "too many connections" event, including to a
1406 deleted file after a log rotate. Leave the per net connection logging
1407 leaving it open for efficiency as that will be quickly detected by the
1408 check on the next write.
1410 HS/01 Bug 1671: Fix post transport crash.
1411 Processing the wait-<transport> messages could crash the delivery
1412 process if the message IDs didn't exist for some reason. When
1413 using 'split_spool_directory=yes' the construction of the spool
1414 file name failed already, exposing the same netto behaviour.
1416 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1417 mime_regex ACL conditions.
1419 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1420 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1421 to HELO, local diagnostic string.
1423 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1424 log line to debug. Even when configured with a tls authenticator many
1425 client connections are expected to not authenticate in this way, so
1426 an authenticate fail is not an error.
1428 HS/02 Add the Exim version string to the process info. This way exiwhat
1429 gives some more detail about the running daemon.
1431 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1432 matter for fast-change records such as DNSBLs.
1434 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1435 retry record, even if constant. There may be multiple transports with
1436 different interface settings and the retry behaviour needs to be kept
1439 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1441 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1443 JH/10 Bug 840: fix log_defer_output option of pipe transport
1445 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1446 hosts_randomize. This matters a lot when combined with mua_wrapper.
1448 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1449 ${quote_pgsql:<string>} operator.
1451 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1453 JH/14 Change header file name for freeradius-client. Relevant if compiling
1454 with Radius support; from the Gentoo tree and checked under Fedora.
1456 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1458 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1459 option transport_filter, meaning no filtering.
1461 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1463 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1464 defaults to "*" (all hosts). The variable is now available when not built
1465 with TLS, default unset, mainly to enable keeping the testsuite sane.
1466 If a server certificate is not supplied (via tls_certificate) an error is
1467 logged, and clients will find TLS connections fail on startup. Presumably
1468 they will retry in-clear.
1469 Packagers of Exim are strongly encouraged to create a server certificate
1470 at installation time.
1472 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1473 with the $config_file variable.
1475 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1476 in transport context, after the attempt, and per-recipient. The latter type
1477 is per host attempted. The event data is the error message, and the errno
1478 information encodes the lookup type (A vs. MX) used for the (first) host,
1479 and the trailing two digits of the smtp 4xx response.
1481 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1482 to write to mainlog (or rejectlog, paniclog) in the window between file
1483 creation and permissions/ownership being changed. Particularly affects
1484 installations where exicyclog is run as root, rather than exim user;
1485 result is that the running daemon panics and dies.
1487 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1489 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1490 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1491 "pri" and "weight". Note that the previous implicit priority given by the
1492 list order is no longer honoured.
1494 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1495 for DKIM processing.
1497 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1498 by defining SUPPORT_SOCKS.
1500 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1501 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1502 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1503 variables are renamed to proxy_{local,external}_{address,port}.
1505 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1506 for a build by defining SUPPORT_I18N
1508 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1509 of the query string, and make ${quote_redis:} do that quoting.
1511 JH/29 Move Events support from Experimental to mainline, enabled by default
1512 and removable for a build by defining DISABLE_EVENT.
1514 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1516 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1517 cached by the daemon.
1519 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1520 by defining LOOKUP_REDIS. The libhiredis library is required.
1522 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1523 keys are given for lookup.
1525 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1526 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1527 only supported when built with TLS support. The PolarSSL SHA routines
1528 are still used when the TLS library is too old for convenient support.
1530 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1531 openssl_options), for security. OpenSSL forces this from version 1.1.0
1532 server-side so match that on older versions.
1534 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1535 allocation for $value could be released as the expansion processing
1536 concluded, but leaving the global pointer active for it.
1538 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1539 and to use the domains and local_parts ACL conditions.
1541 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1542 incorrectly not doubled on cutthrough transmission, hence seen as a
1543 body-termination at the receiving system - resulting in truncated mails.
1544 Commonly the sender saw a TCP-level error, and retransmitted the message
1545 via the normal store-and-forward channel. This could result in duplicates
1546 received - but deduplicating mailstores were liable to retain only the
1547 initial truncated version.
1549 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1551 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1553 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1554 we're in there, support oversigning also; bug 1309.
1556 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1558 HS/04 Add support for keep_environment and add_environment options.
1560 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1561 either intentional arithmetic overflow during PRNG, or testing config-
1564 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1565 delivery resulted in actual delivery. Cancel cutthrough before DATA
1568 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1569 rejects a recipient: pass the reject to the originator.
1571 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1572 Many were false-positives and ignorable, but it's worth fixing the
1575 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1576 for the new environment-manipulation done at startup. Move the routines
1577 from being local to tls.c to being global via the os.c file.
1579 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1580 an extract embedded as result-arg for a map, the first arg for extract
1581 is unavailable so we cannot tell if this is a numbered or keyed
1582 extraction. Accept either.
1588 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1591 JH/02 The smtp transport option "multi_domain" is now expanded.
1593 JH/03 The smtp transport now requests PRDR by default, if the server offers
1596 JH/04 Certificate name checking on server certificates, when exim is a client,
1597 is now done by default. The transport option tls_verify_cert_hostnames
1598 can be used to disable this per-host. The build option
1599 EXPERIMENTAL_CERTNAMES is withdrawn.
1601 JH/05 The value of the tls_verify_certificates smtp transport and main options
1602 default to the word "system" to access the system default CA bundle.
1603 For GnuTLS, only version 3.0.20 or later.
1605 JH/06 Verification of the server certificate for a TLS connection is now tried
1606 (but not required) by default. The verification status is now logged by
1607 default, for both outbound TLS and client-certificate supplying inbound
1610 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1613 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1614 Status Notification (bounce) messages are now MIME format per RFC 3464.
1615 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1616 under the control of the dsn_advertise_hosts option, and routers may
1617 have a dsn_lasthop option.
1619 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1620 default, modifiable by a malware= option. The list separator for
1621 the options can now be changed in the usual way. Bug 68.
1623 JH/10 The smtp_receive_timeout main option is now expanded before use.
1625 JH/11 The incoming_interface log option now also enables logging of the
1626 local interface on delivery outgoing connections.
1628 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1629 if the interface and destination host and port all match.
1631 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1634 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1635 Patch from Andrew Lewis.
1637 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1638 now supports optional time-restrictions, weighting, and priority
1639 modifiers per server. Patch originally by <rommer@active.by>.
1641 JH/16 The spamd_address main option now supports a mixed list of local
1642 and remote servers. Remote servers can be IPv6 addresses, and
1643 specify a port-range.
1645 JH/17 Bug 68: The spamd_address main option now supports an optional
1646 timeout value per server.
1648 JH/18 Bug 1581: Router and transport options headers_add/remove can
1649 now have the list separator specified.
1651 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1654 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1657 JH/21 Support for the A6 type of dns record is withdrawn.
1659 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1660 rather than the verbs used.
1662 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1663 from 255 to 1024 chars.
1665 JH/24 Verification callouts now attempt to use TLS by default.
1667 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1668 are generic router options now. The defaults didn't change.
1670 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1671 Original patch from Alexander Shikoff, worked over by JH.
1673 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1674 files if ZCAT_COMMAND is not executable.
1676 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1678 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1680 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1681 Normally benign, it bites when the pair was led to by a CNAME;
1682 modern usage is to not canonicalize the domain to a CNAME target
1683 (and we were inconsistent anyway for A-only vs AAAA+A).
1685 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1687 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1688 when evaluating $sender_host_dnssec.
1690 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1691 $sender_helo_dnssec variable.
1693 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1695 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1697 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1699 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1700 documented as working, but never had. Support all but $spam_report.
1702 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1703 added for tls authenticator.
1705 HS/03 Add perl_taintmode main config option
1711 TL/01 When running the test suite, the README says that variables such as
1712 no_msglog_check are global and can be placed anywhere in a specific
1713 test's script, however it was observed that placement needed to be near
1714 the beginning for it to behave that way. Changed the runtest perl
1715 script to read through the entire script once to detect and set these
1716 variables, reset to the beginning of the script, and then run through
1717 the script parsing/test process like normal.
1719 TL/02 The BSD's have an arc4random API. One of the functions to induce
1720 adding randomness was arc4random_stir(), but it has been removed in
1721 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1722 function when detected.
1724 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1725 cause callback expansion.
1727 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1728 syntax errors in an expansion can be treated as a string instead of
1729 logging or causing an error, due to the internal use of bool_lax
1730 instead of bool when processing it.
1732 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1733 server certificates when making smtp deliveries.
1735 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1737 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1739 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1741 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1742 Merged patch from Sebastian Wiedenroth.
1744 JH/05 Fix results-pipe from transport process. Several recipients, combined
1745 with certificate use, exposed issues where response data items split
1746 over buffer boundaries were not parsed properly. This eventually
1747 resulted in duplicates being sent. This issue only became common enough
1748 to notice due to the introduction of connection certificate information,
1749 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1751 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1752 size buffer was used, resulting in syntax errors when an expansion
1755 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1756 version 3.3.6 or later.
1758 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1759 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1760 both become "event_action", the variables become $event_name, $event_data
1761 and $event_defer_errno. There is a new variable $verify_mode, usable in
1762 routers, transports and related events. The tls:cert event is now also
1763 raised for inbound connections, if the main configuration event_action
1766 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1767 early OCSP support, but no stapling (appears to be less than 1.0.0).
1769 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1770 server certificate names available under the smtp transport option
1771 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1774 JH/10 Time-related extraction expansions from certificates now use the main
1775 option "timezone" setting for output formatting, and are consistent
1776 between OpenSSL and GnuTLS compilations. Bug 1541.
1778 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1779 encoded parameter in the incoming message. Bug 1558.
1781 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1782 include certificate info, eximon was claiming there were spoolfile
1785 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1787 JH/14 Log delivery-related information more consistently, using the sequence
1788 "H=<name> [<ip>]" wherever possible.
1790 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1791 are problematic for Debian distribution, omit them from the release
1794 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1796 JH/16 Fix string representation of time values on 64bit time_t architectures.
1799 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1800 output list separator was used.
1805 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1806 checkers that were complaining about end of non-void function with no
1809 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1810 This was a regression introduced in 4.83 by another bugfix.
1812 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1814 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1815 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1821 TF/01 Correctly close the server side of TLS when forking for delivery.
1823 When a message was received over SMTP with TLS, Exim failed to clear up
1824 the incoming connection properly after forking off the child process to
1825 deliver the message. In some situations the subsequent outgoing
1826 delivery connection happened to have the same fd number as the incoming
1827 connection previously had. Exim would try to use TLS and fail, logging
1828 a "Bad file descriptor" error.
1830 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1831 utilities have not been installed.
1833 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1834 temporary space as the ACL may create new global variables.
1836 TL/01 LDAP support uses per connection or global context settings, depending
1837 upon the detected version of the libraries at build time.
1839 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1840 to extract and use the src ip:port in logging and expansions as if it
1841 were a direct connection from the outside internet. PPv2 support was
1842 updated based on HAProxy spec change in May 2014.
1844 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1846 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1847 Properly escape header and check for NULL return.
1849 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1852 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1854 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1855 characters in header names, implemented as a verify condition.
1856 Contributed by Michael Fischer v. Mollard.
1858 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1859 results permerror and temperror. Previous values are deprecated but
1860 still accepted. In a future release, err_perm and err_temp will be
1861 completely removed, which will be a backward incompatibility if the
1862 ACL tests for either of these two old results. Patch contributed by
1863 user bes-internal on the mailing list.
1865 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1867 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1868 selectors, in both main and reject logs.
1870 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1873 JH/07 Add malware type "sock" for talking to simple daemon.
1875 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1877 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1878 routers/transports under cutthrough routing.
1880 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1881 numbers. Touch up "bool" conditional to keep the same definition.
1883 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1885 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1887 JH/12 Expand items in router/transport headers_add or headers_remove lists
1888 individually rather than the list as a whole. Bug 1452.
1890 Required for reasonable handling of multiple headers_ options when
1891 they may be empty; requires that headers_remove items with embedded
1892 colons must have them doubled (or the list-separator changed).
1894 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1895 view the policy declared in the DMARC record. Currently, $dmarc_status
1896 is a combined value of both the record presence and the result of the
1899 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1901 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1902 dnslookup router and the smtp transport (applying to the forward
1905 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1906 of ldap servers used for a specific lookup. Patch provided by Heiko
1909 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1910 New variable $lookup_dnssec_authenticated for observability.
1912 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1913 Patch submitted by Lars Timman.
1915 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1917 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1918 Requires trusted mode and valid format message id, aborts otherwise.
1919 Patch contributed by Heiko Schlichting.
1921 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1922 certextract with support for various fields. Bug 1358.
1924 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1925 is requested by default, modifiable by smtp transport option
1928 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1929 operate on certificate variables to give certificate fingerprints
1930 Also new ${sha256:cert_variable}.
1932 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1934 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1935 Christian Aistleitner.
1937 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1939 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1940 file. Patch from Wolfgang Breyha.
1942 JH/25 Expand the coverage of the delivery $host and $host_address to
1943 client authenticators run in verify callout. Bug 1476.
1945 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1946 align with daemon_smtp_ports. Bug 72.
1948 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1949 support and error reporting did not work properly.
1951 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1952 and is readable. Patch from Andrew Colin Kissa.
1954 TL/14 Enhance documentation of ${run expansion and how it parses the
1955 commandline after expansion, particularly in the case when an
1956 unquoted variable expansion results in an empty value.
1958 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1960 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1961 by renaming away from it. Addresses GH issue 15, reported by
1964 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1966 TL/15 SECURITY: prevent double expansion in math comparison functions
1967 (can expand unsanitized data). Not remotely exploitable.
1974 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1976 PP/02 Make -n do something, by making it not do something.
1977 When combined with -bP, the name of an option is not output.
1979 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1982 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1983 $sender_host_name and config options to manage this, and basic check
1986 PP/05 DSCP support for outbound connections and control modifier for inbound.
1988 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1989 (Only plugin which currently uses this is kerberos4, which nobody should
1990 be using, but we should make it available and other future plugins might
1991 conceivably use it, even though it would break NAT; stuff *should* be
1992 using channel bindings instead).
1994 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1995 name; added for Sendmail compatibility; requires admin caller.
1996 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1997 just ignore it); requires trusted caller.
1998 Also parse but ignore: -Ac -Am -X<logfile>
2001 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2003 TL/02 Add +smtp_confirmation as a default logging option.
2005 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2006 Patch by Magnus Holmgren from 2007-02-20.
2008 TL/04 Bugzilla 1281 - Spec typo.
2009 Bugzilla 1283 - Spec typo.
2010 Bugzilla 1290 - Spec grammar fixes.
2012 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2014 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2016 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2017 mailing list by by Dmitry Isaikin.
2019 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2021 JH/02 Support "G" suffix to numbers in ${if comparisons.
2023 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2025 NM/01 Bugzilla 1197 - Spec typo
2026 Bugzilla 1196 - Spec examples corrections
2028 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2030 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2031 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2034 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2035 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2037 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2038 "acl {{name}{arg}...}", and optional args on acl condition
2041 JH/05 Permit multiple router/transport headers_add/remove lines.
2043 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2045 JH/07 Avoid using a waiting database for a single-message-only transport.
2046 Performance patch from Paul Fisher. Bugzilla 1262.
2048 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2051 JH/09 Add $headers_added variable, with content from use of ACL modifier
2052 add_header (but not yet added to the message). Bugzilla 199.
2054 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2055 Pulled from Bugzilla 817 by Wolfgang Breyha.
2057 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2059 (nb: this is the same fix as in Exim 4.80.1)
2061 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2064 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2067 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2069 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2070 advertise SMTP AUTH mechanism to us, instead of a generic
2071 protocol violation error. Also, make Exim more robust to bad
2072 data from the Dovecot auth socket.
2074 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2076 When a queue runner is handling a message, Exim first routes the
2077 recipient addresses, during which it prunes them based on the retry
2078 hints database. After that it attempts to deliver the message to
2079 any remaining recipients. It then updates the hints database using
2082 So if a recipient address works intermittently, it can get repeatedly
2083 deferred at routing time. The retry hints record remains fresh so the
2084 address never reaches the final cutoff time.
2086 This is a fairly common occurrence when a user is bumping up against
2087 their storage quota. Exim had some logic in its local delivery code
2088 to deal with this. However it did not apply to per-recipient defers
2089 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2091 This change adds a proper retry rule check during routing so that the
2092 final cutoff time is checked against the message's age. We only do
2093 this check if there is an address retry record and there is not a
2094 domain retry record; this implies that previous attempts to handle
2095 the address had the retry_use_local_parts option turned on. We use
2096 this as an approximation for the destination being like a local
2097 delivery, as in LMTP.
2099 I suspect this new check makes the old local delivery cutoff check
2100 redundant, but I have not verified this so I left the code in place.
2102 TF/02 Correct gecos expansion when From: is a prefix of the username.
2104 Test 0254 submits a message to Exim with the header
2108 When I ran the test suite under the user fanf2, Exim expanded
2109 the header to contain my full name, whereas it should have added
2110 a Resent-Sender: header. It erroneously treats any prefix of the
2111 username as equal to the username.
2113 This change corrects that bug.
2115 GF/01 DCC debug and logging tidyup
2116 Error conditions log to paniclog rather than rejectlog.
2117 Debug lines prefixed by "DCC: " to remove any ambiguity.
2119 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2121 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2122 Bug spotted by Jeremy Harris; was flawed since initial commit.
2123 Would have resulted in OCSP responses post-SNI triggering an Exim
2124 NULL dereference and crash.
2126 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2128 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2129 Bug detection, analysis and fix by Samuel Thibault.
2130 Bugzilla 1331, Debian bug #698092.
2132 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2134 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2135 Server implementation by Todd Lyons, client by JH.
2136 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2137 config variable "prdr_enable" controls whether the server
2138 advertises the facility. If the client requests PRDR a new
2139 acl_data_smtp_prdr ACL is called once for each recipient, after
2140 the body content is received and before the acl_smtp_data ACL.
2141 The client is controlled by both of: a hosts_try_prdr option
2142 on the smtp transport, and the server advertisement.
2143 Default client logging of deliveries and rejections involving
2144 PRDR are flagged with the string "PRDR".
2146 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2147 fclose(). Diagnosis by Todd Lyons.
2149 PP/17 Update configure.default to handle IPv6 localhost better.
2150 Patch by Alain Williams (plus minor tweaks).
2153 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2154 This is now consistent with GnuTLS, and is now documented: the
2155 previous undocumented portable approach to treating the option as
2156 unset was to force an expansion failure. That still works, and
2157 an empty string is now equivalent.
2159 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2160 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2161 not performing validation itself.
2163 PP/20 Added force_command boolean option to pipe transport.
2164 Patch from Nick Koston, of cPanel Inc.
2166 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2169 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2171 PP/21 Fix eximon continuous updating with timestamped log-files.
2172 Broken in a format-string cleanup in 4.80, missed when I repaired the
2173 other false fix of the same issue.
2174 Report and fix from Heiko Schlichting.
2177 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2178 Report from Prashanth Katuri.
2180 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2181 It's SecureTransport, so affects any MacOS clients which use the
2182 system-integrated TLS libraries, including email clients.
2184 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2185 using a MIME ACL for non-SMTP local injection.
2186 Report and assistance in diagnosis by Warren Baker.
2188 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2190 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2192 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2193 last id that failed so it may be referenced in subsequent ACL's.
2195 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2198 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2199 ldap library initialization, allowing self-signed CA's to be
2200 used. Also properly sets require_cert option later in code by
2201 using NULL (global ldap config) instead of ldap handle (per
2202 session). Bug diagnosis and testing by alxgomz.
2204 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2205 the src/util/ subdirectory.
2207 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2208 renamed to Transport Post Delivery Action by Jeremy Harris, as
2211 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2212 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2213 redis_servers = needs to be configured which will be used by the redis
2214 lookup. Patch from Warren Baker, of The Packet Hub.
2216 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2218 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2219 hostname or reverse DNS when processing a host list. Used suggestions
2220 from multiple comments on this bug.
2222 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2224 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2225 Missed a few lines, added it to make the runtest require no keyboard
2228 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2229 contains upper case chars. Make router use caseful_local_part.
2231 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2232 support when GnuTLS has been built with p11-kit.
2238 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2240 This, or similar/improved, will also be change PP/11 of 4.82.
2246 PP/01 Handle short writes when writing local log-files.
2247 In practice, only affects FreeBSD (8 onwards).
2248 Bugzilla 1053, with thanks to Dmitry Isaikin.
2250 NM/01 Bugzilla 949 - Documentation tweak
2252 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2255 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2257 PP/02 Implemented gsasl authenticator.
2259 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2261 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2262 `pkg-config foo` for cflags/libs.
2264 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2265 with rest of GSASL and with heimdal_gssapi.
2267 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2268 `pkg-config foo` for cflags/libs for the TLS implementation.
2270 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2271 properties get this fed in as external SSF. A number of robustness
2272 and debugging improvements to the cyrus_sasl authenticator.
2274 PP/08 cyrus_sasl server now expands the server_realm option.
2276 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2277 Patch by Jeremy Harris.
2279 PP/10 Added dbmjz lookup type.
2281 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2283 PP/12 MAIL args handles TAB as well as SP, for better interop with
2284 non-compliant senders.
2285 Analysis and variant patch by Todd Lyons.
2287 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2288 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2289 Patch from Dirk Mueller <dmueller@suse.com>
2291 PP/13 tls_peerdn now print-escaped for spool files.
2292 Observed some $tls_peerdn in wild which contained \n, which resulted
2293 in spool file corruption.
2295 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2296 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2297 or write after TLS renegotiation, which otherwise led to messages
2300 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2301 as a tracking header (ie: a signed header comes before the signature).
2302 Patch from Wolfgang Breyha.
2304 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2305 comma-sep list; embedded commas doubled.
2307 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2309 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2311 Report and patch from Dmitry Banschikov.
2313 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2314 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2315 failures. We appear to now support TLS1.1+ with Exim.
2317 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2318 lets Exim select keys and certificates based upon TLS SNI from client.
2319 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2320 before an outbound SMTP session. New log_selector, +tls_sni.
2322 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2323 NULL dereference. Report and patch from Alun Jones.
2325 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2326 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2327 Not seeing resolver debug output on NetBSD, but suspect this is a
2328 resolver implementation change.
2330 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2331 Left warnings. Added "eximon gdb" invocation mode.
2333 PP/21 Defaulting "accept_8bitmime" to true, not false.
2335 PP/22 Added -bw for inetd wait mode support.
2337 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2338 locate the relevant includes and libraries. Made this the default.
2340 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2341 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2343 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2344 This may cause build issues on older platforms.
2346 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2347 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2348 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2349 Added SNI support via GnuTLS too.
2350 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2352 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2354 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2355 Applied second patch from Janne, implementing suggestion to default
2356 multiple-strings-in-record handling to match SPF spec.
2358 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2360 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2361 read-only, out of scope).
2362 Patch from Wolfgang Breyha, report from Stuart Northfield.
2364 PP/29 Fix three issues highlighted by clang analyser static analysis.
2365 Only crash-plausible issue would require the Cambridge-specific
2366 iplookup router and a misconfiguration.
2367 Report from Marcin Mirosław.
2369 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2371 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2372 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2373 As part of this, removing so much warning spew let me fix some minor
2374 real issues in debug logging.
2376 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2377 assignment on my part. Fixed.
2379 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2380 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2381 Janne Snabb (who went above and beyond: thank you).
2383 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2384 string otherwise requires a connection and a bunch more work and it's
2385 relatively easy to get wrong. Should also expose TLS library linkage
2388 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2389 64-bit ${eval} (JH/03).
2391 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2392 GNU libc to support some of the 64-bit stuff, should not lead to
2393 conflicts. Defined before os.h is pulled in, so if a given platform
2394 needs to override this, it can.
2396 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2397 protection layer was required, which is not implemented.
2398 Bugzilla 1254, patch from Wolfgang Breyha.
2400 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2401 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2402 tls_dhparam take prime identifiers. Also unbreak combination of
2403 OpenSSL+DH_params+TLSSNI.
2405 PP/39 Disable SSLv2 by default in OpenSSL support.
2411 PP/01 Solaris build fix for Oracle's LDAP libraries.
2412 Bugzilla 1109, patch from Stephen Usher.
2414 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2416 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2419 TF/02 Fix a couple more cases where we did not log the error message
2420 when unlink() failed. See also change 4.74-TF/03.
2422 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2423 lock up or crash if it happened to be inside a call to libc when it
2424 got a SIGUSR1 from exiwhat.
2426 The SIGUSR1 handler appends the current process status to the process
2427 log which is later printed by exiwhat. It used to use the general
2428 purpose logging code to do this, but several functions it calls are
2429 not safe for signals.
2431 The new output code in the SIGUSR1 handler is specific to the process
2432 log, and simple enough that it's easy to inspect for signal safety.
2433 Removing some special cases also simplifies the general logging code.
2434 Removing the spurious timestamps from the process log simplifies
2437 TF/04 Improved ratelimit ACL condition.
2439 The /noupdate option has been deprecated in favour of /readonly which
2440 has clearer semantics. The /leaky, /strict, and /readonly update modes
2441 are mutually exclusive. The update mode is no longer included in the
2442 database key; it just determines when the database is updated. (This
2443 means that when you upgrade Exim will forget old rate measurements.)
2445 Exim now checks that the per_* options are used with an update mode that
2446 makes sense for the current ACL. For example, when Exim is processing a
2447 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2448 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2449 must specify per_mail/readonly. If you omit the update mode it defaults to
2450 /leaky where that makes sense (as before) or /readonly where required.
2452 The /noupdate option is now undocumented but still supported for
2453 backwards compatibility. It is equivalent to /readonly except that in
2454 ACLs where /readonly is required you may specify /leaky/noupdate or
2455 /strict/noupdate which are treated the same as /readonly.
2457 A useful new feature is the /count= option. This is a generalization
2458 of the per_byte option, so that you can measure the throughput of other
2459 aggregate values. For example, the per_byte option is now equivalent
2460 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2462 The per_rcpt option has been generalized using the /count= mechanism
2463 (though it's more complicated than the per_byte equivalence). When it is
2464 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2465 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2466 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2467 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2468 Note that using per_rcpt with a non-readonly update mode in more than
2469 one ACL will cause the recipients to be double-counted. (The per_mail
2470 and per_byte options don't have this problem.)
2472 The handling of very low rates has changed slightly. If the computed rate
2473 is less than the event's count (usually one) then this event is the first
2474 after a long gap. In this case the rate is set to the same as this event's
2475 count, so that the first message of a spam run is counted properly.
2477 The major new feature is a mechanism for counting the rate of unique
2478 events. The new per_addr option counts the number of different
2479 recipients that someone has sent messages to in the last time period. It
2480 behaves like per_rcpt if all the recipient addresses are different, but
2481 duplicate recipient addresses do not increase the measured rate. Like
2482 the /count= option this is a general mechanism, so the per_addr option
2483 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2484 example, measure the rate that a client uses different sender addresses
2485 with the options per_mail/unique=$sender_address. There are further
2486 details in the main documentation.
2488 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2490 TF/06 Removed a few PCRE remnants.
2492 TF/07 Automatically extract Exim's version number from tags in the git
2493 repository when doing development or release builds.
2495 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2496 Bugzilla 879. Patch from Paul Fisher.
2498 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2499 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2502 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2504 PP/05 Variable $av_failed, true if the AV scanner deferred.
2505 Bugzilla 1078. Patch from John Horne.
2507 PP/06 Stop make process more reliably on build failure.
2508 Bugzilla 1087. Patch from Heiko Schlittermann.
2510 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2511 Bugzilla 1089. Patch from Heiko Schlittermann.
2513 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2514 Bugzilla 1131. Patch from Holger Weiß.
2516 PP/09 Handle IPv6 addresses with SPF.
2517 Bugzilla 860. Patch from Wolfgang Breyha.
2519 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2521 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2524 PP/11 match_* no longer expand right-hand-side by default.
2525 New compile-time build option, EXPAND_LISTMATCH_RHS.
2526 New expansion conditions, "inlist", "inlisti".
2528 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2530 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2532 PP/14 fix log_write() format string regression from TF/03.
2533 Bugzilla 1152. Patch from Dmitry Isaikin.
2539 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2541 PP/02 Harmonised TLS library version reporting; only show if debugging.
2542 Layout now matches that introduced for other libraries in 4.74 PP/03.
2544 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2546 PP/04 New "dns_use_edns0" global option.
2548 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2551 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2552 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2554 TK/01 Updated PolarSSL code to 0.14.2.
2555 Bugzilla 1097. Patch from Andreas Metzler.
2557 PP/07 Catch divide-by-zero in ${eval:...}.
2558 Fixes bugzilla 1102.
2560 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2563 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2564 format-string attack -- SECURITY: remote arbitrary code execution.
2566 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2567 time unintentionally subject to list matching rules, letting the header
2568 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2569 arbitrary string expansion). This allowed for information disclosure.
2571 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2572 INT_MIN/-1 -- value coerced to INT_MAX.
2578 NM/01 Workaround for PCRE version dependency in version reporting
2581 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2582 This fixes portability to compilers other than gcc, notably
2583 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2585 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2586 makefiles for portability to HP-UX and POSIX correctness.
2588 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2589 Also via indented variable definition in the Makefile.
2590 (Debugging by Oliver Heesakkers).
2592 PP/02 Restore caching of spamd results with expanded spamd_address.
2593 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2595 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2596 Improves build reliability. Fix from: Frank Elsner
2598 NM/02 Fix wide character breakage in the rfc2047 coding
2599 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2601 NM/03 Allow underscore in dnslist lookups
2602 Fixes bug 1026. Patch from Graeme Fowler
2604 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2605 Code patches from Adam Ciarcinski of NetBSD.
2607 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2610 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2611 is logged, to avoid truncation. Patch from John Horne.
2613 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2614 Patch from Jakob Hirsch.
2616 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2617 SQL string expansion failure details.
2618 Patch from Andrey Oktyabrski.
2620 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2621 Patch from Simon Arlott.
2623 PP/09 New lookups functionality failed to compile on old gcc which rejects
2624 extern declarations in function scope.
2625 Patch from Oliver Fleischmann
2627 PP/10 Use sig_atomic_t for flags set from signal handlers.
2628 Check getgroups() return and improve debugging.
2629 Fixed developed for diagnosis in bug 927 (which turned out to be
2632 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2633 Patch from Mark Zealey.
2635 PP/12 Bugzilla 1056: Improved spamd server selection.
2636 Patch from Mark Zealey.
2638 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2639 Based on patch from Heiko Schlittermann.
2641 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2642 Patch from Uwe Doering, sign-off by Michael Haardt.
2644 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2645 variable declaration deep within a block. Bug and patch from
2648 PP/15 lookups-Makefile IRIX compatibility coercion.
2650 PP/16 Make DISABLE_DKIM build knob functional.
2652 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2653 Patch by Simon Arlott
2655 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2656 variable argument macros. Our copy now differs from upstream.
2662 TF/01 Failure to get a lock on a hints database can have serious
2663 consequences so log it to the panic log.
2665 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2666 controlled using the smtp_confirmation log selector.
2668 TF/03 Include the error message when we fail to unlink a spool file.
2670 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2671 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2672 for maintaining out-of-tree patches for some time.
2674 PP/01 Bugzilla 139: Documentation and portability issues.
2675 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2676 Handle per-OS dynamic-module compilation flags.
2678 PP/02 Let /dev/null have normal permissions.
2679 The 4.73 fixes were a little too stringent and complained about the
2680 permissions on /dev/null. Exempt it from some checks.
2681 Reported by Andreas M. Kirchwitz.
2683 PP/03 Report version information for many libraries, including
2684 Exim version information for dynamically loaded libraries. Created
2685 version.h, now support a version extension string for distributors
2686 who patch heavily. Dynamic module ABI change.
2688 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2689 privilege escalation vulnerability whereby the Exim run-time user
2690 can cause root to append content of the attacker's choosing to
2693 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2696 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2697 If dropping privileges for untrusted macros, we disabled normal logging
2698 on the basis that it would fail; for the Exim run-time user, this is not
2699 the case, and it resulted in successful deliveries going unlogged.
2700 Fixed. Reported by Andreas Metzler.
2706 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2707 only prepend for the Resent-* case. Fixes regression introduced in
2708 Exim 4.70 by NM/22 for Bugzilla 607.
2710 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2711 increasing numbers of administrators be bitten by this.
2713 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2715 PP/03 Bugzilla 994: added openssl_options main configuration option.
2717 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2719 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2721 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2723 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2724 without a peer certificate, leading to a segfault because of an
2725 assumption that peers always have certificates. Be a little more
2726 paranoid. Problem reported by Martin Tscholak.
2728 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2729 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2730 NB: ClamAV planning to remove STREAM in "middle of 2010".
2731 CL also introduces -bmalware, various -d+acl logging additions and
2732 more caution in buffer sizes.
2734 PP/09 Implemented reverse_ip expansion operator.
2736 PP/10 Bugzilla 937: provide a "debug" ACL control.
2738 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2740 PP/12 Bugzilla 973: Implement --version.
2742 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2744 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2746 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2748 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2749 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2750 ignore trailing whitespace.
2752 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2754 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2757 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2758 Notification from Dr Andrew Aitchison.
2760 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2761 ExtendedDetectionInfo response format.
2762 Notification from John Horne.
2764 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2767 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2768 XSL and documented dependency on system catalogs, with examples of how
2771 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2774 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2775 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2776 configuration file which is writeable by the Exim user or group.
2778 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2779 of configuration files to cover files specified with the -C option if
2780 they are going to be used with root privileges, not just the default
2783 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2784 option (effectively making it always true).
2786 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2787 files to be used while preserving root privileges.
2789 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2790 that rogue child processes cannot use them.
2792 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2793 run-time user, instead of root.
2795 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2796 Exim run-time user without dropping privileges.
2798 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2799 result string, instead of calling string_vformat() twice with the same
2802 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2803 for other users. Others should always drop root privileges if they use
2804 -C on the command line, even for a whitelisted configure file.
2806 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2808 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2814 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2815 $data_path, and $header_path variables; fixed documentation bugs and
2818 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2819 exipick to access non-standard spools, including the "frozen" queue
2822 NM/01 Bugzilla 965: Support mysql stored procedures.
2823 Patch from Alain Williams
2825 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2827 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2828 Patch from Andreas Metzler
2830 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2831 Patch from Kirill Miazine
2833 NM/05 Bugzilla 671: Added umask to procmail example.
2835 JJ/03 installed exipick 20100323.0, fixing doc bug
2837 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2838 directory. Notification and patch from Dan Rosenberg.
2840 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2842 TK/02 Improve log output when DKIM signing operation fails.
2844 MH/01 Treat the transport option dkim_domain as a colon separated
2845 list, not as a single string, and sign the message with each element,
2846 omitting multiple occurences of the same signer.
2848 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2849 Bugzilla 985, 986. Patch by Simon Arlott
2851 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2852 Patch by Simon Arlott
2854 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2855 MBX locking. Notification from Dan Rosenberg.
2861 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2863 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2865 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2867 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2869 NM/04 Bugzilla 915: Flag broken perl installation during build.
2875 TK/01 Added patch by Johannes Berg that expands the main option
2876 "spamd_address" if it starts with a dollar sign.
2878 TK/02 Write list of recipients to X-Envelope-Sender header when building
2879 the mbox-format spool file for content scanning (suggested by Jakob
2882 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2883 (http://www.dcc-servers.net/) support via dccifd. Activated by
2884 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2886 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2887 by Mark Daniel Reidel <mr@df.eu>.
2889 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2890 When building exim an external PCRE library is now needed -
2891 PCRE is a system library on the majority of modern systems.
2892 See entry on PCRE_LIBS in EDITME file.
2894 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2895 conversation. Added nologin parameter to request.
2896 Patch contributed by Kirill Miazine.
2898 TF/01 Do not log submission mode rewrites if they do not change the address.
2900 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2902 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2903 log files in place. Contributed by Roberto Lima.
2905 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2907 TF/03 Bugzilla 615: When checking the local_parts router precondition
2908 after a local_part_suffix or local_part_prefix option, Exim now
2909 does not use the address's named list lookup cache, since this
2910 contains cached lookups for the whole local part.
2912 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2913 Robert Millan. Documentation is in experimental-spec.txt.
2915 TF/04 Bugzilla 668: Fix parallel build (make -j).
2917 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2919 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2920 Patch provided by Jan Srzednicki.
2922 TF/05 Leading white space used to be stripped from $spam_report which
2923 wrecked the formatting. Now it is preserved.
2925 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2926 that they are available at delivery time.
2928 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2930 TF/08 TLS error reporting now respects the incoming_interface and
2931 incoming_port log selectors.
2933 TF/09 Produce a more useful error message if an SMTP transport's hosts
2934 setting expands to an empty string.
2936 NM/06 Bugzilla 744: EXPN did not work under TLS.
2937 Patch provided by Phil Pennock.
2939 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2940 Patch provided by Richard Godbee.
2942 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2943 acl_smtp_notquit, added index entry.
2945 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2946 Patch provided by Eugene Bujak.
2948 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2949 accept(). Patch provided by Maxim Dounin.
2951 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2952 Patch provided by Phil Pennock.
2954 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2956 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2957 Patch provided by Brad "anomie" Jorsch.
2959 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2960 Patch provided by Dean Brooks.
2962 TK/05 Add native DKIM support (does not depend on external libraries).
2964 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2965 Patch provided by Graeme Fowler.
2967 NM/16 Bugzilla 851: Documentation example syntax fix.
2969 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2971 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2974 NM/19 Bugzilla 745: TLS version reporting.
2975 Patch provided by Phil Pennock.
2977 NM/20 Bugzilla 167: bool: condition support.
2978 Patch provided by Phil Pennock.
2980 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2981 clients. Patch provided by Phil Pennock.
2983 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2984 Patch provided by Brad "anomie" Jorsch.
2986 NM/23 Bugzilla 687: Fix misparses in eximstats.
2987 Patch provided by Heiko Schlittermann.
2989 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2990 Patch provided by Heiko Schlittermann.
2992 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2993 plus update to original patch.
2995 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2997 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2998 Patch provided by David Brownlee.
3000 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3002 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3004 NM/30 Bugzilla 888: TLS documentation bugfixes.
3006 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3008 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3009 Unlike the original bugzilla I have changed all shell scripts in src tree.
3011 NM/33 Bugzilla 898: Transport filter timeout fix.
3012 Patch by Todd Rinaldo.
3014 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3015 Patch by Serge Demonchaux.
3017 NM/35 Bugzilla 39: Base64 decode bug fixes.
3018 Patch by Jakob Hirsch.
3020 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3022 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3024 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3026 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3032 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3033 ALT-N's libdkim that I have put here:
3034 http://duncanthrax.net/exim-experimental/
3036 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3037 were called 'true' and it seems that C99 defines that as a reserved
3038 keyword to be used with 'bool' variable types. That means you could
3039 not include C99-style headers which use bools without triggering
3040 build errors in sieve.c.
3042 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3043 as mailq or other aliases. Changed the --help handling significantly
3044 to do whats expected. exim_usage() emits usage/help information.
3046 SC/01 Added the -bylocaldomain option to eximstats.
3048 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3050 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3052 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3058 PH/01 Another patch from the Sieve maintainer.
3060 PH/02 When an IPv6 address is converted to a string for single-key lookup
3061 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3062 dots are used instead of colons so that keys in lsearch files need not
3063 contain colons. This was done some time before quoting was made available
3064 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3065 (notated using the quote facility) so as to distinguish them from IPv4
3066 keys. This meant that lookups for IP addresses in host lists did not work
3067 for iplsearch lookups.
3069 This has been fixed by arranging for IPv6 addresses to be expressed with
3070 colons if the lookup type is iplsearch. This is not incompatible, because
3071 previously such lookups could never work.
3073 The situation is now rather anomalous, since one *can* have colons in
3074 ordinary lsearch keys. However, making the change in all cases is
3075 incompatible and would probably break a number of configurations.
3077 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3080 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3081 conversion specification without a maximum field width, thereby enabling
3082 a rogue spamd server to cause a buffer overflow. While nobody in their
3083 right mind would setup Exim to query an untrusted spamd server, an
3084 attacker that gains access to a server running spamd could potentially
3085 exploit this vulnerability to run arbitrary code as the Exim user.
3087 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3088 $primary_hostname instead of what libspf2 thinks the hosts name is.
3090 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3091 a directory entry by the name of the lookup key. Previously, if a
3092 symlink pointed to a non-existing file or a file in a directory that
3093 Exim lacked permissions to read, a lookup for a key matching that
3094 symlink would fail. Now it is enough that a matching directory entry
3095 exists, symlink or not. (Bugzilla 503.)
3097 PH/03 The body_linecount and body_zerocount variables are now exported in the
3100 PH/04 Added the $dnslist_matched variable.
3102 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3103 This means they are set thereafter only if the connection becomes
3106 PH/06 Added the client_condition to authenticators so that some can be skipped
3107 by clients under certain conditions.
3109 PH/07 The error message for a badly-placed control=no_multiline_responses left
3110 "_responses" off the end of the name.
3112 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3114 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3115 (without spaces) instead of just copying the configuration text.
3117 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3119 PH/11 Added $max_received_linelength.
3121 PH/12 Added +ignore_defer and +include_defer to host lists.
3123 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3124 way in which PCRE > 7.0 is built.
3126 PH/14 Implemented queue_only_load_latch.
3128 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3129 MAIL command. The effect was to mangle the value on 64-bit systems.
3131 PH/16 Another patch from the Sieve maintainer.
3133 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3135 PH/18 If a system quota error occurred while trying to create the file for
3136 a maildir delivery, the message "Mailbox is full" was not appended to the
3137 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3138 only to a quota excession during the actual writing of the file.
3140 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3141 characters?) which causes problems in log lines. The DN values are now
3142 passed through string_printing() before being added to log lines.
3144 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3145 and InterBase are left for another time.)
3147 PH/21 Added message_body_newlines option.
3149 PH/22 Guard against possible overflow in moan_check_errorcopy().
3151 PH/23 POSIX allows open() to be a macro; guard against that.
3153 PH/24 If the recipient of an error message contained an @ in the local part
3154 (suitably quoted, of course), incorrect values were put in $domain and
3155 $local_part during the evaluation of errors_copy.
3161 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3162 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3165 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3166 issue a MAIL command.
3168 PH/02 In an ACL statement such as
3170 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3172 if a client was not listed at all, or was listed with a value other than
3173 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3174 the condition was not true (as it should be), so access was not denied.
3175 The bug was that the ! inversion was incorrectly passed on to the second
3176 item. This has been fixed.
3178 PH/03 Added additional dnslists conditions == and =& which are different from
3179 = and & when the dns lookup returns more than one IP address.
3181 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3182 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3184 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3185 FSYNC, which compiles an option called disable_fsync that allows for
3186 bypassing fsync(). The documentation is heavily laced with warnings.
3188 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3190 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3191 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3192 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3193 including adding "make clean"; (3) Added -fPIC when compiling the test
3194 dynamically loaded module, to get rid of a warning.
3196 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3197 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3198 = 0s. The bug is otherwise harmless.
3200 PH/07 There was a bug in the dovecot authenticator such that the value of
3201 $auth1 could be overwritten, and so not correctly preserved, after a
3202 successful authentication. This usually meant that the value preserved by
3203 the server_setid option was incorrect.
3205 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3207 PH/09 Installed PCRE release 7.0.
3209 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3210 run for batched SMTP input. It is now run at the start of every message
3211 in the batch. While fixing this I discovered that the process information
3212 (output by running exiwhat) was not always getting set for -bs and -bS
3213 input. This is fixed, and it now also says "batched" for BSMTP.
3215 PH/11 Added control=no_pipelining.
3217 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3218 patch, slightly modified), and move the expansion of helo_data till after
3219 the connection is made in the smtp transport (so it can use these
3222 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3224 PH/14 Added log_selector = +pid.
3226 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3228 PH/16 Add ${if forany and ${if forall.
3230 PH/17 Added dsn_from option to vary the From: line in DSNs.
3232 PH/18 Flush SMTP output before performing a callout, unless control =
3233 no_callout_flush is set.
3235 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3236 was true (the default) a successful delivery failed to delete the retry
3237 item, thus causing premature timeout of the address. The bug is now
3240 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3242 PH/21 Long custom messages for fakedefer and fakereject are now split up
3243 into multiline responses in the same way that messages for "deny" and
3244 other ACL rejections are.
3246 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3247 with slight modification.
3249 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3250 draft, changing the syntax and factoring some duplicate code".
3252 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3253 for deliveries of the second and subsequent messages over the same SMTP
3256 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3257 ${reduce, with only minor "tidies".
3259 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3261 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3262 expansion side effects.
3264 PH/27 When a message times out after an over-quota error from an Exim-imposed
3265 quota, the bounce message says "mailbox is full". This message was not
3266 being given when it was a system quota that was exceeded. It now should
3269 MH/03 Made $recipients available in local_scan(). local_scan() already has
3270 better access to the recipient list through recipients_list[], but
3271 $recipients can be useful in postmaster-provided expansion strings.
3273 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3274 in the case of a MAIL command with additional options following the
3275 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3276 were accidentally chopped off.
3278 PH/29 SMTP synchronization checks are implemented when a command is read -
3279 there is a check that no more input is waiting when there shouldn't be
3280 any. However, for some commands, a delay in an ACL can mean that it is
3281 some time before the response is written. In this time, more input might
3282 arrive, invalidly. So now there are extra checks after an ACL has run for
3283 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3284 pipelining has not been advertised.
3286 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3288 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3289 returned in the SMTP response when smtp_return_error_details was false.
3290 This has been fixed.
3292 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3293 instead of the C I/O that was originally supplied, because problems were
3294 reported on Solaris.
3296 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3297 Exim which did not show up earlier: it was assuming that a call to
3298 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3299 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3300 was a macro that became an assignment, so it seemed to work. This has
3301 changed to a proper function call with a void return, hence the compile
3302 error. Exim's code has been fixed.
3304 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3307 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3309 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3311 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3312 because it turns out to be too big or there is a timeout) there is no
3313 "Completed" line in the log. When some messages of this type were
3314 selected by exigrep, they were listed as "not completed". Others were
3315 picked up by some special patterns. I have improved the selection
3316 criteria to be more general.
3318 PH/38 The host_find_failed option in the manualroute router can now be set
3319 to "ignore", to completely ignore a host whose IP address cannot be
3320 found. If all hosts are ignored, the behaviour is controlled by the new
3321 host_all_ignored option.
3323 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3324 homing or because of multiple MX records with /mx) generated more than
3325 one IP address, and the following item turned out to be the local host,
3326 all the secondary addresses of the first item were incorrectly removed
3327 from the list, along with the local host and any following hosts (which
3328 is what is supposed to happen).
3330 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3331 whoever called Exim into the -H file. In the case of the daemon it was
3332 behaving confusingly. When first started, it used values for whoever
3333 started the daemon, but after a SIGHUP it used the Exim user (because it
3334 calls itself on a restart). I have changed the code so that it now always
3337 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3338 message are rejected with the same error (e.g. no authentication or bad
3339 sender address), and a DATA command is nevertheless sent (as can happen
3340 with PIPELINING or a stupid MUA), the error message that was given to the
3341 RCPT commands is included in the rejection of the DATA command. This is
3342 intended to be helpful for MUAs that show only the final error to their
3345 PH/42 Another patch from the Sieve maintainer.
3347 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3348 Eximstats - Fixed some broken HTML links and added missing column headers
3350 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3351 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3353 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3355 PH/43 Yet another patch from the Sieve maintainer.
3357 PH/44 I found a way to check for a TCP/IP connection going away before sending
3358 the response to the final '.' that terminates a message, but only in the
3359 case where the client has not sent further data following the '.'
3360 (unfortunately, this is allowed). However, in many cases there won't be
3361 any further data because there won't be any more messages to send. A call
3362 to select() can be used: if it shows that the input is "ready", there is
3363 either input waiting, or the socket has been closed. An attempt to read
3364 the next input character can distinguish the two cases. Previously, Exim
3365 would have sent an OK response which the client would never have see.
3366 This could lead to message repetition. This fix should cure that, at
3367 least in a lot of common cases.
3369 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3370 advertised in response to EHLO.
3376 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3377 fixed by 4.65/MH/01 (is this a record?) are fixed:
3379 (i) An empty string was always treated as zero by the numeric comparison
3380 operators. This behaviour has been restored.
3382 (ii) It is documented that the numeric comparison operators always treat
3383 their arguments as decimal numbers. This was broken in that numbers
3384 starting with 0 were being interpreted as octal.
3386 While fixing these problems I realized that there was another issue that
3387 hadn't been noticed. Values of message_size_limit (both the global option
3388 and the transport option) were treated as octal if they started with 0.
3389 The documentation was vague. These values are now always treated as
3390 decimal, and I will make that clear in the documentation.
3396 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3397 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3400 MH/01 Don't check that the operands of numeric comparison operators are
3401 integers when their expansion is in "skipping" mode (fixes bug
3402 introduced by 4.64-PH/07).
3404 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3405 child addresses, Exim now panics and dies. Previously, because the count
3406 is held in a short int, deliveries were likely to be lost. As such a
3407 large number of recipients for a single message is ridiculous
3408 (performance will be very, very poor), I have chosen to impose a limit
3409 rather than extend the field.
3415 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3416 leftover -K file (the existence of which was triggered by #402).
3417 While we were at it, introduced process PID as part of the -K
3418 filename. This should rule out race conditions when creating
3421 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3422 processing considerably. Previous code took too long for large mails,
3423 triggering a timeout which in turn triggers #401.
3425 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3426 in the DK code in transports.c. sendfile() is not really portable,
3427 hence the _LINUX specificness.
3429 TF/01 In the add_headers option to the mail command in an Exim filter,
3430 there was a bug that Exim would claim a syntax error in any
3431 header after the first one which had an odd number of characters
3434 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3435 callout verification, Exim cached a "reject" for the entire domain. This
3436 is correct for most verifications, but it is not correct for a recipient
3437 verification with use_sender or use_postmaster set, because in that case
3438 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3439 case of MAIL FROM:<> rejection from other early rejections (e.g.
3440 rejection of HELO). When verifying a recipient using a non-null MAIL
3441 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3442 Whatever the result of the callout, the value of the domain cache is
3443 left unchanged (for any other kind of callout, getting as far as trying
3444 RCPT means that the domain itself is ok).
3446 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3449 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3450 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3451 session, the daemon ignores ECONNECT errors and logs others; it now
3452 ignores EPIPE as well.
3454 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3455 (quoted-printable decoding).
3457 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3458 later the small subsequent patch to fix an introduced bug.
3460 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3462 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3464 PH/08 An error is now given if message_size_limit is specified negative.
3466 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3467 to be given (somewhat) arbitrary names.
3469 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3472 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3473 miscellaneous code fixes
3475 PH/10 Added the log_reject_target ACL modifier to specify where to log
3478 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3479 hostname. This is wrong, because it relates to the incoming message (and
3480 probably the interface on which it is arriving) and not to the outgoing
3481 callout (which could be using a different interface). This has been
3482 changed to use the value of the helo_data option from the smtp transport
3483 instead - this is what is used when a message is actually being sent. If
3484 there is no remote transport (possible with a router that sets up host
3485 addresses), $smtp_active_hostname is used.
3487 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3488 tweaks were necessary in order to get it to work (see also 21 below):
3489 (a) The code assumed that strncpy() returns a negative number on buffer
3490 overflow, which isn't the case. Replaced with Exim's string_format()
3492 (b) There were several signed/unsigned issues. I just did the minimum
3493 hacking in of casts. There is scope for a larger refactoring.
3494 (c) The code used strcasecmp() which is not a standard C function.
3495 Replaced with Exim's strcmpic() function.
3496 (d) The code set only $1; it now sets $auth1 as well.
3497 (e) A simple test gave the error "authentication client didn't specify
3498 service in request". It would seem that Dovecot has changed its
3499 interface. Fortunately there's a specification; I followed it and
3500 changed what the client sends and it appears to be working now.
3502 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3505 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3506 address A is aliased to B and C, where B exists and C does not. Without
3507 -v the output is "A verified" because verification stops after a
3508 successful redirection if more than one address is generated. However,
3509 with -v the child addresses are also verified. Exim was outputting "A
3510 failed to verify" and then showing the successful verification for C,
3511 with its parentage. It now outputs "B failed to verify", showing B's
3512 parentage before showing the successful verification of C.
3514 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3515 look up a TXT record in a specific list after matching in a combined
3518 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3519 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3520 they consult the DNS. I had assumed they would set it the way they
3521 wanted; and indeed my experiments on Linux seem to show that in some
3522 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3523 To be on the safe side, however, I have now made the interface to
3524 host_find_byname() similar to host_find_bydns(), with an argument
3525 containing the DNS resolver options. The host_find_byname() function now
3526 sets these options at its start, just as host_find_bydns() does. The smtp
3527 transport options dns_qualify_single and dns_search_parents are passed to
3528 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3529 of host_find_byname() use the default settings of RES_DEFNAMES
3530 (qualify_single) but not RES_DNSRCH (search_parents).
3532 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3533 spool_read_header() do less string testing, by means of a preliminary
3534 switch on the second character of optional "-foo" lines. (This is
3535 overdue, caused by the large number of possibilities that now exist.
3536 Originally there were few.) While I was there, I also converted the
3537 str(n)cmp tests so they don't re-test the leading "-" and the first
3538 character, in the hope this might squeeze out yet more improvement.
3540 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3541 flag allowing group syntax was set by the header_syntax check but not
3542 turned off, possible causing trouble later; (2) The flag was not being
3543 set at all for the header_verify test, causing "group"-style headers to
3544 be rejected. I have now set it in this case, and also caused header_
3545 verify to ignore an empty address taken from a group. While doing this, I
3546 came across some other cases where the code for allowing group syntax
3547 while scanning a header line wasn't quite right (mostly, not resetting
3548 the flag correctly in the right place). These bugs could have caused
3549 trouble for malformed header lines. I hope it is now all correct.
3551 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3552 with the "reply" argument non-NULL. The code, however (which originally
3553 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3554 but it didn't always do it. This confused somebody who was copying the
3555 code for some other use. I have removed all the tests.
3557 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3558 feature that was used to support insecure browsers during the U.S. crypto
3559 embargo. It requires special client support, and Exim is probably the
3560 only MTA that supported it -- and would never use it because real RSA is
3561 always available. This code has been removed, because it had the bad
3562 effect of slowing Exim down by computing (never used) parameters for the
3563 RSA_EXPORT functionality.
3565 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3566 authenticator to fail if there's a tab character in the incoming data
3567 (there should never be unless someone is messing about, as it's supposed
3568 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3569 if the connection is using TLS or if the remote IP is the same as the
3570 local IP, and the "valid-client-cert option" if a client certificate has
3573 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3574 authenticators. This can be used for authorization after authentication
3575 succeeds. (In the case of plaintext, it servers for both authentication
3578 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3579 if any retry times were supplied.
3581 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3582 connection, where there is no client IP address to check. In this
3583 situation, the verify now always succeeds.
3585 PH/25 Applied John Jetmore's -Mset patch.
3587 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3589 PH/27 In a string expansion for a processed (not raw) header when multiple
3590 headers of the same name were present, leading whitespace was being
3591 removed from all of them, but trailing whitespace was being removed only
3592 from the last one. Now trailing whitespace is removed from each header
3593 before concatenation. Completely empty headers in a concatenation (as
3594 before) are ignored.
3596 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3597 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3599 PH/29 [Removed. This was a change that I later backed out, and forgot to
3600 correct the ChangeLog entry (that I had efficiently created) before
3601 committing the later change.]
3603 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3604 address errors (4xx response to RCPT) over the same connection as other
3605 messages routed to the same hosts. Such deliveries are always "forced",
3606 so retry times are not inspected. This resulted in far too many retries
3607 for the affected addresses. The effect occurred only when there were more
3608 hosts than the hosts_max_try setting in the smtp transport when it had
3609 the 4xx errors. Those hosts that it had tried were not added to the list
3610 of hosts for which the message was waiting, so if all were tried, there
3611 was no problem. Two fixes have been applied:
3613 (i) If there are any address or message errors in an SMTP delivery, none
3614 of the hosts (tried or untried) are now added to the list of hosts
3615 for which the message is waiting, so the message should not be a
3616 candidate for sending over the same connection that was used for a
3617 successful delivery of some other message. This seems entirely
3618 reasonable: after all the message is NOT "waiting for some host".
3619 This is so "obvious" that I'm not sure why it wasn't done
3620 previously. Hope I haven't missed anything, but it can't do any
3621 harm, as the worst effect is to miss an optimization.
3623 (ii) If, despite (i), such a delivery is accidentally attempted, the
3624 routing retry time is respected, so at least it doesn't keep
3625 hammering the server.
3627 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3628 in ${readsocket because some servers need this prod.
3630 PH/32 Added some extra debug output when updating a wait-xxx database.
3632 PH/33 The hint "could be header name not terminated by colon", which has been
3633 given for certain expansion errors for a long time, was not being given
3634 for the ${if def:h_colon_omitted{... case.
3636 PH/34 The spec says: "With one important exception, whenever a domain list is
3637 being scanned, $domain contains the subject domain." There was at least
3638 one case where this was not true.
3640 PH/35 The error "getsockname() failed: connection reset by peer" was being
3641 written to the panic log as well as the main log, but it isn't really
3642 panic-worthy as it just means the connection died rather early on. I have
3643 removed the panic log writing for the ECONNRESET error when getsockname()
3646 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3647 runs only) independently of the message's sender address. This meant
3648 that, if the 4xx error was in fact related to the sender, a different
3649 message to the same recipient with a different sender could confuse
3650 things. In particular, this can happen when sending to a greylisting
3651 server, but other circumstances could also provoke similar problems.
3652 I have changed the default so that the retry time for these errors is now
3653 based a combination of the sender and recipient addresses. This change
3654 can be overridden by setting address_retry_include_sender=false in the
3657 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3658 remote server are returned as part of bounce messages. This was not
3659 happening for LMTP over a pipe (the lmtp transport), but now it is the
3660 same for both kinds of LMTP.
3662 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3663 in header lines that were in fact CNAMEs. This is no longer the case.
3665 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3666 and queue runs started by the daemon processed all messages. This has
3667 been fixed so that -R and -S can now usefully be given with -q<time>.
3669 PH/40 Import PCRE release 6.7 (fixes some bugs).
3671 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3673 PH/42 Give an error if -q is specified more than once.
3675 PH/43 Renamed the variables $interface_address and $interface_port as
3676 $received_ip_address and $received_port, to make it clear that these
3677 values apply to message reception, and not to the outgoing interface when
3678 a message is delivered. (The old names remain recognized, of course.)
3680 PH/44 There was no timeout on the connect() call when using a Unix domain
3681 socket in the ${readsocket expansion. There now is.
3683 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3684 be meaningful with "accept".
3686 SC/01 Eximstats V1.43
3687 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3689 SC/02 Eximstats V1.44
3690 Use a glob alias rather than an array ref in the generated
3691 parser. This improves both readability and performance.
3693 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3694 Collect SpamAssassin and rejection statistics.
3695 Don't display local sender or destination tables unless
3696 there is data to show.
3697 Added average volumes into the top table text output.
3699 SC/04 Eximstats V1.46
3700 Collect data on the number of addresses (recipients)
3701 as well as the number of messages.
3703 SC/05 Eximstats V1.47
3704 Added 'Message too big' to the list of mail rejection
3705 reasons (thanks to Marco Gaiarin).
3707 SC/06 Eximstats V1.48
3708 Mainlog lines which have GMT offsets and are too short to
3709 have a flag are now skipped.
3711 SC/07 Eximstats V1.49 (Alain Williams)
3712 Added the -emptyok flag.
3714 SC/08 Eximstats V1.50
3715 Fixes for obtaining the IP address from reject messages.
3717 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3718 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3719 whitespace changes from 4.64-PH/27
3721 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3724 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3725 are found, allow negative numbers in numeric criteria)
3727 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3729 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3730 to match changes made in 4.64-PH/43
3732 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3734 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3735 "domain" is actually an IP address, and force a failure. This locks out
3736 those revolvers/nameservers that support "A-for-A" lookups, in
3737 contravention of the specifications.
3739 PH/48 When a host name was looked up from an IP address, and the subsequent
3740 forward lookup of the name timed out, the host name was left in
3741 $sender_host_name, contrary to the specification.
3743 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3744 restricted to single-key lookups, Exim was not diagnosing an error if
3745 * or *@ was used with a query-style lookup.
3747 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3749 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3750 long ago, but noone interested enough thought of it. Let's just say that
3751 the "1.1" means that there are some new functions that weren't there at
3752 some point in the past.
3754 PH/51 Error processing for expansion failure of helo_data from an smtp
3755 transport during callout processing was broken.
3757 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3758 tested/used via the -bh/-bhc/-bs options.
3760 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3761 bug, fixed in subsequent PCRE releases).
3763 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3764 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3766 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3772 SC/01 Use a glob alias rather than an array ref in eximstats generated
3773 parser. This improves both readability and performance.
3775 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3776 Don't display local sender or destination tables in eximstats unless
3777 there is data to show.
3778 Added average volumes into the eximstats top table text output.
3780 SC/03 Collect data on the number of addresses (recipients) as well
3781 as the number of messages in eximstats.
3783 TF/01 Correct an error in the documentation for the redirect router. Exim
3784 does (usually) call initgroups() when daemonizing.
3786 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3787 with consistent privilege compared to when running as a daemon.
3789 TF/03 Note in the spec that $authenticated_id is not set for local
3790 submissions from trusted users.
3792 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3793 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3795 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3796 by adding some example configuration directives to the default
3797 configuration file. A little bit of work is required to uncomment the
3798 directives and define how usernames and passwords are checked, but
3799 there is now a framework to start from.
3801 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3802 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3803 without this. I don't know how relevant this is to other LDAP libraries.
3805 PH/02 Add the verb name to the "unknown ACL verb" error.
3807 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3809 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3811 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3812 directory not expanded when it should be if an expanded home directory
3813 was set for the address (which is overridden by the transport).
3815 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3818 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3819 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3820 because it is too late at that time, and has no effect.
3822 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3823 security issue with \' (bugzilla #107). I could not use the
3824 PQescapeStringConn() function, because it needs a PGconn value as one of
3827 PH/08 When testing addresses using -bt, indicate those final addresses that
3828 are duplicates that would not cause an additional delivery. At least one
3829 person was confused, thinking that -bt output corresponded to deliveries.
3830 (Suppressing duplicates isn't a good idea as you lose the information
3831 about possibly different redirections that led to the duplicates.)
3833 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3834 systems where poll() doesn't work, in particular OS X.
3836 PH/10 Added more information to debugging output for retry time not reached.
3838 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3839 operations in malware.c.
3841 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3844 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3845 syslog_duplication set false, log lines that would normally be written
3846 both the the main log and to the reject log were not written to syslog at
3849 PH/14 In the default configuration, change the use of "message" in ACL warn
3850 statements to "add_header".
3852 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3853 not followed by a command (e.g. "seen endif").
3855 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3856 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3859 PH/17 Added extra conditions to the default value of delay_warning_condition
3863 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3864 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3865 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3868 The Auto-Submitted: and various List- headers are standardised, whereas I
3869 don't think Precedence: ever was.
3871 PH/18 Refactored debugging code in route_finduser() to show more information,
3872 in particular, the error code if getpwnam() issues one.
3874 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3875 This is apparently needed in addition to the PH/07 change above to avoid
3876 any possible encoding problems.
3878 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3879 but not after initializing Perl.
3881 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3882 output them only if debugging. By default they are written stderr,
3883 apparently, which is not desirable.
3885 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3888 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3891 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3893 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3894 authenticated or an ident call has been made. Suppress the default
3895 values for $authenticated_id and $authenticated_sender (but permit -oMai
3896 and -oMas) when testing with -bh.
3898 PH/24 Re-jigged the order of the tests in the default configuration so that the
3899 tests for valid domains and recipients precede the DNS black list and CSA
3900 tests, on the grounds that those ones are more expensive.
3902 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3903 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3904 This bug exists in every version of Exim that I still have, right back to
3907 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3908 However, an attempt to turn on case-sensitivity in a regex key by
3909 including (?-i) didn't work because the subject string was already
3910 lowercased, and the effects were non-intuitive. It turns out that a
3911 one-line patch can be used to allow (?-i) to work as expected.
3917 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3918 other effects) broke the use of negated acl sub-conditions.
3920 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3923 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3924 "Deny" causes Exim to reject the incoming connection with a 554 error.
3925 Unfortunately, if there is a major crisis, such as a disk failure,
3926 tcp-wrappers gives "deny", whereas what one would like would be some
3927 kind of temporary error. A kludge has been added to help with this.
3928 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3929 554 error is used if errno is still zero or contains ENOENT (which occurs
3930 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3933 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3935 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3936 errors. Otherwise a message that provokes a temporary error (when other
3937 messages do not) can cause a whole host to time out.
3939 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3940 addresses were routed directly to files or pipes from a redirect router.
3941 File deliveries just didn't batch; pipe deliveries might have suffered
3944 PH/06 A failure to get a lock for a hints database would erroneously always say
3945 "Failed to get write lock", even when it was really a read lock.
3947 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3948 of 0600. This has been changed to use the value of the lockfile_mode
3949 option (which defaults to 0600).
3951 PH/08 Applied small patch from the Sieve maintainer.
3953 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3954 folder from quota calculations, a direct delivery into this folder messed
3955 up the contents of the maildirsize file. This was because the regex was
3956 used only to exclude .Trash (or whatever) when the size of the mailbox
3957 was calculated. There was no check that a delivery was happening into an
3958 excluded directory. This bug has been fixed by ignoring all quota
3959 processing for deliveries into excluded directories.
3961 PH/10 Added the maildirfolder_create_regex option to appendfile.
3967 PH/01 The code for finding all the local interface addresses on a FreeBSD
3968 system running IPv6 was broken. This may well have applied to all BSD
3969 systems, as well as to others that have similar system calls. The broken
3970 code found IPv4 interfaces correctly, but gave incorrect values for the
3971 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3972 that it would not match correctly against @[] and not recognize the IPv6
3975 PH/02 The ipliteral router was not recognizing addresses of the form user@
3976 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3978 PH/03 Added disable_ipv6.
3980 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3981 decoded form, because it is most often used to construct To: headers
3982 lines in autoreplies, and the decoded form may well be syntactically
3983 invalid. However, $reply_address has leading white space removed, and all
3984 newlines turned into spaces so that the autoreply transport does not
3987 PH/05 If group was specified without a user on a router, and no group or user
3988 was specified on a transport, the group from the router was ignored.
3990 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3991 for visible compile-time settings that can be used to change these
3992 numbers, for those that want even more. Backwards compatibility with old
3993 spool files has been maintained. However, going back to a previous Exim
3994 release will lost any variables that are in spool files.
3996 PH/07 Two small changes when running in the test harness: increase delay when
3997 passing a TCP/IP connection to a new process, in case the original
3998 process has to generate a bounce, and remove special handling of
3999 127.0.0.2 (sic), which is no longer necessary.
4001 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4002 be the same on different OS.
4004 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4007 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4008 whether --show-vars was specified or not
4010 JJ/02 exipick: Added support for new ACL variable spool format introduced
4013 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4014 syntactically invalid From: or Reply-to: line, and a filter used this to
4015 generate an autoreply, and therefore failed to obtain an address for the
4016 autoreply, Exim could try to deliver to a non-existent relative file
4017 name, causing unrelated and misleading errors. What now happens is that
4018 it logs this as a hard delivery error, but does not attempt to create a
4021 PH/11 The exinext utility has a -C option for testing purposes, but although
4022 the given file was scanned by exinext itself; it wasn't being passed on
4023 when Exim was called.
4025 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4026 an end-of-file indication when reading a command response.
4028 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4029 compiled. In many other places in Exim, IPv6 addresses are always
4030 recognized, so I have changed this. It also means that IPv4 domain
4031 literals of the form [IPV4:n.n.n.n] are now always recognized.
4033 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4034 used if the router is not running as root, for example, when verifying at
4035 ACL time, or when using -bh. The debugging output from this situation was
4036 non-existent - all you got was a failure to exec. I have made two
4039 (a) Failures to set uid/gid, the current directory, or a process leader
4040 in a subprocess such as that created by queryprogram now generate
4041 suitable debugging output when -d is set.
4043 (b) The queryprogram router detects when it is not running as root,
4044 outputs suitable debugging information if -d is set, and then runs
4045 the subprocess without attempting to change uid/gid.
4047 PH/15 Minor change to Makefile for building test_host (undocumented testing
4050 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4051 additional section of a DNS packet that returns MX or SRV records.
4052 Instead, it always explicitly searches for A/AAAA records. This avoids
4053 major problems that occur when a DNS server includes only records of one
4054 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4055 fixed another bug: if SRV records were looked up and the corresponding
4056 address records were *not* found in the additional section, the port
4057 values from the SRV records were lost.
4059 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4060 using the correct key (the original address) when searching the retry
4061 rules in order to find which one to use for generating the retry hint.
4063 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4064 adding the default one. Similarly, if it contains a Reply-To: header, the
4065 errors_reply_to option, if set, is not used.
4067 PH/19 When calculating a retry time, Exim used to measure the "time since
4068 failure" by looking at the "first failed" field in the retry record. Now
4069 it does not use this if it is later than than the arrival time of the
4070 message. Instead it uses the arrival time. This makes for better
4071 behaviour in cases where some deliveries succeed, thus re-setting the
4072 "first failed" field. An example is a quota failure for a huge message
4073 when small messages continue to be delivered. Without this change, the
4074 "time since failure" will always be short, possible causing more frequent
4075 delivery attempts for the huge message than are intended.
4076 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4078 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4079 $1, $2, $3) because the numerical variables can be reset during some
4080 expansion items (e.g. "match"), thereby losing the authentication data.
4082 PH/21 Make -bV show the size of off_t variables so that the test suite can
4083 decide whether to run tests for quotas > 2G.
4085 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4086 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4087 filecount value is greater than 2G or if a quota value is greater than 2G
4088 on a system where the size of off_t is not greater than 4, a panic error
4091 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4092 never match. The debug and -bh output now contains an explicit error
4093 message indicating a malformed IPv4 address or mask.
4095 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4096 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4097 PH/23 above applies.
4099 PH/25 Do not write to syslog when running in the test harness. The only
4100 occasion when this arises is a failure to open the main or panic logs
4101 (for which there is an explicit test).
4103 PH/26 Added the /no_tell option to "control=freeze".
4105 PH/27 If a host name lookup failed very early in a connection, for example, if
4106 the IP address matched host_lookup and the reverse lookup yielded a name
4107 that did not have a forward lookup, an error message of the form "no IP
4108 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4109 could be logged. Now it outputs the IP address instead of "NULL".
4111 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4112 allows the sender and the authenticated sender to be set when
4113 submitting a message from within Exim. Since child_open_exim() is
4114 documented for local_scan(), the new function should be too.
4116 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4117 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4118 results in an empty string is now treated as unset.
4120 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4122 PH/31 Added sender_verify_fail logging option.
4124 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4125 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4126 tidied the source and removed it altogether.
4128 PH/33 When a queue run was abandoned because the load average was too high, a
4129 log line was always written; now it is written only if the queue_run log
4130 selector is set. In addition, the log line for abandonment now contains
4131 information about the queue run such as the pid. This is always present
4132 in "start" and "stop" lines but was omitted from the "abandon" line.
4134 PH/34 Omit spaces between a header name and the colon in the error message that
4135 is given when verify = headers_syntax fails (if there are lots of them,
4136 the message gets confusing).
4138 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4139 names, as there are now some PTR records that contain slashes. This check
4140 is only to protect against broken name servers that fall over on strange
4141 characters, so the fact that it applies to all lookups doesn't matter.
4143 PH/36 Now that the new test suite is complete, we can remove some of the
4144 special code in Exim that was needed for the old test suite. For example,
4145 sorting DNS records because real resolvers return them in an arbitrary
4146 order. The new test suite's fake resolver always returns records in the
4149 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4150 bounces) except when queue_only is set, to avoid logging races between
4151 the different processes.
4153 PH/38 Panic-die if .include specifies a non-absolute path.
4155 PH/39 A tweak to the "H" retry rule from its user.
4157 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4158 a label. They prevented compilation on older perls.
4160 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4161 a warning to be raised on newish perls.
4163 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4164 on queue. Changes to match documented behaviour of showing count of
4165 messages matching specified criteria.
4167 PH/40 Changed the default ident timeout from 30s to 5s.
4169 PH/41 Added support for the use of login_cap features, on those BSD systems
4170 that have them, for controlling the resources used by pipe deliveries.
4172 PH/42 The content-scanning code uses fopen() to create files in which to put
4173 message data. Previously it was not paying any attention to the mode of
4174 the files. Exim runs with umask(0) because the rest of the code creates
4175 files with open(), and sets the required mode explicitly. Thus, these
4176 files were ending up world-writeable. This was not a big issue, because,
4177 being within the spool directory, they were not world-accessible. I have
4178 created a function called modefopen, which takes an additional mode
4179 argument. It sets umask(777), creates the file, chmods it to the required
4180 mode, then resets the umask. All the relevant calls to fopen() in the
4181 content scanning code have been changed to use this function.
4183 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4184 to 24 hours. This avoids potential overflow problems when processing G
4185 and H retry rules. I suspect nobody ever tinkers with this value.
4187 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4189 PH/45 When the plaintext authenticator is running as a client, the server's
4190 challenges are checked to ensure they are valid base64 strings. By
4191 default, the authentication attempt is cancelled if an invalid string is
4192 received. Setting client_ignore_invalid_base64 true ignores these errors.
4193 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4194 they are received. Thus, the responses can be made to depend on the
4195 challenges. If an invalid string is ignored, an empty string is placed in
4198 PH/46 Messages that are created by the autoreply transport now contains a
4199 References: header, in accordance with RFCs 2822 and 3834.
4201 PH/47 Added authenticated_sender_force to the smtp transport.
4203 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4205 PH/49 Installed latest patch from the Sieve maintainer.
4207 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4208 was also set, the appendfile transport was unnecessarily scanning a
4209 directory of message files (e.g. for maildir delivery) to find the count
4210 of files (along with the size), even though it did not need this
4211 information. It now does the scan only if it needs to find either the
4212 size of the count of files.
4214 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4216 PH/52 Two bugs concerned with error handling when the smtp transport is
4219 (i) Exim was not creating retry information for temporary errors given
4220 for individual recipients after the DATA command when the smtp transport
4221 was used in LMTP mode. This meant that they could be retried too
4222 frequently, and not timed out correctly.
4224 (ii) Exim was setting the flag that allows error details to be returned
4225 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4226 recipients that were returned after the DATA command.
4228 PH/53 This is related to PH/52, but is more general: for any failing address,
4229 when detailed error information was permitted to be returned to the
4230 sender, but the error was temporary, then after the final timeout, only
4231 "retry timeout exceeded" was returned. Now it returns the full error as
4232 well as "retry timeout exceeded".
4234 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4235 do this, and (what is worse) MTAs that accept it.
4237 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4238 will now be deprecated.
4240 PH/56 New os.c-cygwin from the Cygwin maintainer.
4242 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4243 formats (previously only available in exim formats via -bpr, -bpru,
4244 and -bpra. Now also available in native and exiqgrep formats)
4246 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4247 with very large, slow to parse queues
4249 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4251 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4253 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4254 responses to authentication challenges, though it was showing the
4255 challenges; (ii) I've removed the CR characters from the debug output for
4258 PH/58 Allow for the insertion of a newline as well as a space when a string
4259 is turned into more than one encoded-word during RFC 2047 encoding. The
4260 Sieve code now uses this.
4262 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4263 data_4xx, lost_connection, tls_required.
4265 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4266 message was being sent as an SMTP response.
4268 PH/61 Add -l and -k options to exicyclog.
4270 PH/62 When verifying, if an address was redirected to one new address, so that
4271 verification continued, and the new address failed or deferred after
4272 having set something in $address_data, the value of $address_data was not
4273 passed back to the ACL. This was different to the case when no
4274 redirection occurred. The value is now passed back in both cases.
4276 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4277 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4278 use login_cap.h, so on its own it isn't the distinguishing feature. The
4279 new name refers directly to the setclassresources() function.
4281 PH/65 Added configuration files for NetBSD3.
4283 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4285 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4286 is preferred over IPv4.
4288 PH/68 The bounce_return_message and bounce_return_body options were not being
4289 honoured for bounces generated during the reception of non-SMTP messages.
4290 In particular, this applied to messages rejected by the ACL. This bug has
4291 been fixed. However, if bounce_return_message is true and bounce_return_
4292 body is false, the headers that are returned for a non-SMTP message
4293 include only those that have been read before the error was detected.
4294 (In the case of an ACL rejection, they have all been read.)
4296 PH/69 The HTML version of the specification is now built in a directory called
4297 spec_html instead of spec.html, because the latter looks like a path with
4298 a MIME-type, and this confuses some software.
4300 PH/70 Catch two compiler warnings in sieve.c.
4302 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4303 function verify_get_ident() calls ip_connect() to connect a socket, but
4304 if the "connect()" function timed out, ip_connect() used to close the
4305 socket. However, verify_get_ident() also closes the socket later, and in
4306 between Exim writes to the log, which may get opened at this point. When
4307 the socket was closed in ip_connect(), the log could get the same file
4308 descriptor number as the socket. This naturally causes chaos. The fix is
4309 not to close the socket in ip_connect(); the socket should be closed by
4310 the function that creates it. There was only one place in the code where
4311 this was missing, in the iplookup router, which I don't think anybody now
4312 uses, but I've fixed it anyway.
4314 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4315 well as to direct DNS lookups. Otherwise the handling of names in host
4316 lists is inconsistent and therefore confusing.
4322 PH/01 Two changes to the default runtime configuration:
4324 (1) Move the checks for relay_from_hosts and authenticated clients from
4325 after to before the (commented out) DNS black list checks.
4327 (2) Add control=submission to the relay_from_hosts and authenticated
4328 clients checks, on the grounds that messages accepted by these
4329 statements are most likely to be submissions.
4331 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4333 (1) Generate an error if the third argument for the ${prvs expansion is
4336 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4339 (3) Reset the variables that are obtained from the first argument of
4340 ${prvscheck and used in the second argument before leaving the code,
4341 because their memory is reclaimed, so using them afterwards may do
4344 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4345 one (it's much easier than Tom thought :-).
4347 (5) Because of (4), we can now allow for the use of $prvscheck_result
4348 inside the third argument.
4350 PH/03 For some reason, the default setting of PATH when running a command from
4351 a pipe transport was just "/usr/bin". I have changed it to
4354 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4355 anything to be listed in the output from -bV.
4357 PH/05 When a filter generated an autoreply, the entire To: header line was
4358 quoted in the delivery log line, like this:
4360 => >A.N.Other <ano@some.domain> <original@ddress> ...
4362 This has been changed so that it extracts the operative address. There
4363 may be more than one such address. If so, they are comma-separated, like
4366 => >ano@some.domain,ona@other.domain <original@ddress> ...
4368 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4369 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4370 not being looked up in the rDNS to get a host name, Exim was showing the
4371 IP address twice in Received: lines, even though the IP addresses were
4372 identical. For example:
4374 Received: from [1.2.3.4] (helo=[1.2.3.4])
4376 However, if the real host name was known, it was omitting the HELO data
4377 if it matched the actual IP address. This has been tidied up so that it
4378 doesn't show the same IP address twice.
4380 PH/07 When both +timestamp and +memory debugging was on, the value given by
4381 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4382 called by the debug printing, thereby overwriting the timestamp buffer.
4383 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4385 PH/08 When the original message was included in an autoreply transport, it
4386 always said "this is a copy of the message, including all the headers",
4387 even if body_only or headers_only was set. It now gives an appropriate
4390 PH/09 Applied a patch from the Sieve maintainer which:
4392 o fixes some comments
4393 o adds the (disabled) notify extension core
4394 o adds some debug output for the result of if/elsif tests
4395 o points to the current vacation draft in the documentation
4396 and documents the missing references header update
4400 o fixes a bug in processing the envelope test (when testing
4401 multiple envelope elements, the last element determined the
4404 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4405 Electronic Mail") by including:
4407 Auto-submitted: auto-generated
4409 in the messages that it generates (bounce messages and others, such as
4410 warnings). In the case of bounce messages for non-SMTP messages, there was
4411 also a typo: it was using "Auto_submitted" (underscore instead of
4412 hyphen). Since every message generated by Exim is necessarily in response
4413 to another message, thes have all been changed to:
4415 Auto-Submitted: auto-replied
4417 in accordance with these statements in the RFC:
4419 The auto-replied keyword:
4421 - SHOULD be used on messages sent in direct response to another
4422 message by an automatic process,
4424 - MUST NOT be used on manually-generated messages,
4426 - MAY be used on Delivery Status Notifications (DSNs) and Message
4427 Disposition Notifications (MDNs),
4429 - MUST NOT be used on messages generated by automatic or periodic
4430 processes, except for messages which are automatic responses to
4433 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4434 to the default Received: header definition.
4436 PH/12 Added log selector acl_warn_skipped (default on).
4438 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4439 variables because (a) they are in the wrong storage pool and (b) even if
4440 they were copied, it wouldn't work properly because of the caching.
4442 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4443 checking when decoding. Apparently there are clients that generate
4444 overlong encoded strings. Why am I not surprised?
4446 PH/15 If the first argument of "${if match_address" was not empty, but did not
4447 contain an "@" character, Exim crashed. Now it writes a panic log message
4448 and treats the condition as false.
4450 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4452 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4453 extension "envelope-auth". The code is finished and in agreement with
4454 other implementations, but there is no documentation so far and in fact,
4455 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4456 not changing the active code.
4458 Print executed "if" and "elsif" statements when debugging is used. This
4459 helps a great deal to understand what a filter does.
4461 Document more things not specified clearly in RFC3028. I had all this
4462 sorted out, when out of a sudden new issues came to my mind. Oops."
4464 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4467 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4468 canonical form (as documented). However, after a host name lookup from
4469 the IP address, check_host() was doing a simple string comparison with
4470 addresses acquired from the DNS when checking that the found name did
4471 have the original IP as one of its addresses. Since any found IPv6
4472 addresses are likely to be in abbreviated form, the comparison could
4473 fail. Luckily, there already exists a function for doing the comparison
4474 by converting both addresses to binary, so now that is used instead of
4475 the text comparison.
4477 PH/20 There was another similar case to PH/19, when a complete host name was
4478 given in a host list; looking up its IP address could give an abbreviated
4479 form, whereas the current host's name might or might not be abbreviated.
4480 The same fix has been applied.
4486 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4487 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4490 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4491 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4493 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4495 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4496 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4497 Also, for "Auto-Submitted", treat anything other than "no" as
4498 non-personal, in accordance with RFC 3834. (Previously it treated
4499 anything starting "auto-" as non-personal.)
4501 TF/01 The control=submission/name=... option had a problem with syntax
4502 errors if the name included a slash character. The /name= option
4503 now slurps the rest of the string, so it can include any characters
4504 but it must come last in the list of options (after /sender_retain
4507 PH/05 Some modifications to the interface to the fake nameserver for the new
4515 TK/01 Added the "success_on_redirect" address verification option. See
4516 NewStuff for rationale and an example.
4518 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4520 PH/02 Patch to exigrep to allow it to work on syslog lines.
4522 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4523 fread() to skip over the body file's header line, because in Cygwin the
4524 header line is locked and is inaccessible.
4526 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4527 co-exist for some time) to make it clear that it is the Exim ID that is
4528 referenced, not the Message-ID: header line.
4530 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4531 string_format() function, because snprintf() does not exist on all
4534 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4535 ${stat: expansion item.
4537 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4538 protocol synchronization error", to keep the pedants happy.
4540 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4541 well as for IRIX systems, when gcc is being used. See the host.c source
4544 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4546 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4549 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4550 transport and to the smtp transport in LMTP mode.
4552 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4554 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4555 run a filter in a subprocess. This could lead to confusion in subsequent
4556 lookups in the parent process. There should also be a search_tidyup() at
4557 the end of the subprocess.
4559 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4560 only if the host matched helo_try_verify_hosts, which caused the
4561 verification to occur when the EHLO/HELO command was issued. The ACL just
4562 tested the remembered result. Now, if a previous verification attempt has
4563 not happened, "verify = helo" does it there and then.
4565 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4567 TK/03 Fix log output including CR from clamd.
4569 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4570 exist provoked a memory error which could cause a segfault.
4572 PH/15 Installed PCRE 6.2
4574 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4576 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4577 of the problem. Specifically, suggested +O2 rather than +O1 for the
4580 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4582 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4583 record, the port was not correctly logged when the outgoing_port log
4584 selector was set (it logged the transort's default port).
4586 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4587 fallback_hosts, and "hosts" in the smtp transport.
4589 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4590 host errors such as "Connection refused".
4592 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4593 authentication with radiusclient 0.4.9:
4595 - Error returned from rc_read_config was caught wrongly
4596 - Username/password not passed on to radius server due to wrong length.
4598 The presumption is that some radiusclient API changes for 4.51/PH/17
4599 were not taken care of correctly. The code is still untested by me (my
4600 Linux distribution still has 0.3.2 of radiusclient), but it was
4601 contributed by a Radius user.
4603 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4604 expanding the "port" option of the smtp transport.
4606 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4607 while reading a MIME header. Thanks to Tom Hughes for a patch.
4609 PH/24 Include config.h inside local_scan.h so that configuration settings are
4612 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4613 that in an ACL for RCPT (for example), you can examine exactly what was
4616 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4617 commands, but it was not correctly comparing the address with the actual
4618 client host address. Thus, it would show the EHLO address in Received:
4619 header lines when this was not necessary.
4621 PH/27 Added the % operator to ${eval:}.
4623 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4624 it should be ignoring failures (because with -C, for example, it has lost
4625 privilege). It wasn't ignoring creation failures other than "already
4628 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4631 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4632 because an input line was too long, either on its own, or by virtue of
4633 too many continuations, the temporary file was not being removed, and the
4634 return code was incorrect.
4636 PH/31 Missing "BOOL" in function definition in filtertest.c.
4638 PH/32 Applied Sieve patches from the maintainer.
4640 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4642 PH/33 Added "verify = not_blind".
4644 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4645 Local/Makefile (with some defaults set). These are used in built scripts
4646 such as exicyclog, but they have never been used in the exim_install
4647 script (though there are many overriding facilities there). I have
4648 arranged that the exim_install script now takes note of these two
4651 PH/35 Installed configuration files for Dragonfly.
4653 PH/36 When a locally submitted message by a trusted user did not contain a
4654 From: header, and the sender address was obtained from -f or from an SMTP
4655 MAIL command, and the trusted user did not use -F to supply a sender
4656 name, $originator_name was incorrectly used when constructing a From:
4657 header. Furthermore, $originator_name was used for submission mode
4658 messages from external hosts without From: headers in a similar way,
4659 which is clearly wrong.
4661 PH/37 Added control=suppress_local_fixups.
4663 PH/38 When log_selector = +received_sender was set, and the addition of the
4664 sender made the log line's construction buffer exactly full, or one byte
4665 less than full, an overflow happened when the terminating "\n" was
4668 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4669 when the result of a list match is failure because a DNS lookup failed.
4671 PH/40 RM_COMMAND is now used in the building process.
4673 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4674 the "build-* directories that it finds.
4676 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4677 address in a domain literal was a prefix of an interface address.
4679 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4680 when verifying a sender address, unless rewrite_headers is false.
4682 PH/44 Wrote a long comment about why errors_to addresses are verified as
4683 recipients, not senders.
4685 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4686 the ratelimit ACL was added.
4688 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4690 PH/46 Added extra information about PostgreSQL errors to the error string.
4692 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4693 avoiding the need to install special zones in a real server. This is
4694 backwards compatible; if it can't find the fake resolver, it drops back.
4695 Thus, both old and new test suites can be run.
4697 TF/02 Added util/ratelimit.pl
4699 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4700 clock is set back in time.
4702 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4703 Candler <B.Candler@pobox.com>.
4705 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4706 OK for addresses that are the result of redirections.
4708 PH/48 A number of further additions for the benefit of the new test suite,
4709 including a fake gethostbyname() that interfaces to the fake DNS resolver
4712 TF/06 The fix for widen_domains has also been applied to qualify_single and
4713 search_parents which are the other dnslookup options that can cause
4716 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4719 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4721 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4722 service" if it encountered an MX record with an empty target hostname.
4723 The message is now "an MX or SRV record indicated no SMTP service".
4725 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4726 if the DNS of the sending site is misconfigured. This is quite a
4727 common situation. This change restores the behaviour of treating a
4728 helo verification defer as a failure.
4730 PH/51 If self=fail was set on a router, the bounce message did not include the
4731 actual error message.
4737 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4739 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4740 command itself ended in error, the underlying message about the transport
4741 filter timeout was being overwritten with the pipe command error. Now the
4742 underlying error message should be appended to the second error message.
4744 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4746 PH/02 Reduce the amount of output that "make" produces by default. Full output
4747 can still be requested.
4749 PH/03 The warning log line about a condition test deferring for a "warn" verb
4750 was being output only once per connection, rather than after each
4751 occurrence (because it was using the same function as for successful
4752 "warn" verbs). This seems wrong, so I have changed it.
4754 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4755 it should not have, which might have caused a crash in the right
4756 circumstances, but probably never did.
4758 PH/04 Installed a modified version of Tony Finch's patch to make submission
4759 mode fix the return path as well as the Sender: header line, and to
4760 add a /name= option so that you can make the user's friendly name appear
4763 TF/03 Added the control = fakedefer ACL modifier.
4765 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4766 Mark Lowes for thorough testing.
4768 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4770 TK/03 Merged latest SRS patch from Miles Wilton.
4772 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4773 with the definition in sysexits.h (which is #included earlier).
4774 Fortunately, Exim does not actually use EX_OK. The code used to try to
4775 preserve the sysexits.h value, by assuming that macro definitions were
4776 scanned for macro replacements. I have been disabused of this notion,
4777 so now the code just undefines EX_OK before #including unistd.h.
4779 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4780 in the smtp transport. When a block could not be written in a single
4781 write() function, the timeout was being re-applied to each part-write.
4782 This seems wrong - if the receiver was accepting one byte at a time it
4783 would take for ever. The timeout is now adjusted when this happens. It
4784 doesn't have to be particularly precise.
4786 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4787 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4789 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4790 without a domain if the check to <postmaster@domain> fails.
4792 SC/01 Eximstats: added -xls and the ability to specify output files
4793 (patch written by Frank Heydlauf).
4795 SC/02 Eximstats: use FileHandles for outputting results.
4797 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4799 SC/04 Eximstats: fixed display of large numbers with -nvr option
4801 SC/05 Eximstats: fixed merging of reports with empty tables.
4803 SC/06 Eximstats: added the -include_original_destination flag
4805 SC/07 Eximstats: removed tabs and trailing whitespace.
4807 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4809 TK/06 MBOX spool code: Add real "From " MBOX separator line
4810 so the .eml file is really in mbox format (even though
4811 most programs do not really care). Patch from Alex Miller.
4813 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4814 The latter is generated from $received_to and is only set if the
4815 message has one envelope recipient. SA can use these headers,
4816 obviously out-of-the-box. Patch from Alex Miller.
4818 PH/08 The ${def test on a variable was returning false if the variable's
4819 value was "0", contrary to what the specification has always said!
4820 The result should be true unless the variable is empty.
4822 PH/09 The syntax error of a character other than { following "${if
4823 def:variable_name" (after optional whitespace) was not being diagnosed.
4824 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4825 accidental colon was present, for example, could give incorrect results.
4827 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4828 result is used (not including appendfile, where other changes are about
4831 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4832 This involved changing a lot of size variables from int to off_t. It
4833 should work with maildirs and everything.
4835 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4836 spamd dying while we are connected to it.
4838 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4841 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4842 function for BDB 4.3.
4844 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4846 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4847 the sources for PCRE have been re-arranged and more files are now
4850 PH/15 The code I had for printing potentially long long variables in PH/11
4851 above was not the best (it lost precision). The length of off_t variables
4852 is now inspected at build time, and an appropriate printing format (%ld
4853 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4854 to be "long long int" or "long int". This is needed for the internal
4855 formatting function string_vformat().
4857 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4858 the configuration file to be ":syslog", then the script "guesses" where
4859 the logs files are, rather than using the compiled in default. In our
4860 case the guess is not the same as the compiled default, so the script
4861 suddenly stopped working when I started to use syslog. The patch checks
4862 to see if log_file_path is "". If so, it attempts to read it from exim
4863 with no configuration file to get the compiled in version, before it
4864 falls back to the previous guessing code."
4866 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4867 implementing BATV in an Exim configuration. See NewStuff for the gory
4870 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4871 Makefile that are specific to HP-UX.
4873 PH/18 If the "use_postmaster" option was set for a recipient callout together
4874 with the "random" option, the postmaster address was used as the MAIL
4875 FROM address for the random test, but not for the subsequent recipient
4876 test. It is now used for both.
4878 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4879 patch removes a few documentation additions to RFC 3028, because the
4880 latest draft now contains them. It adds the new en;ascii-case comparator
4881 and a new error check for 8bit text in MIME parts. Comparator and
4882 require names are now matched exactly. I enabled the subaddress
4883 extension, but it is not well tested yet (read: it works for me)."
4885 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4886 rework some of the code of TK/09 above to avoid the hardwired use of
4887 "%lld" and "long long". Replaced the call to snprintf() with a call to
4890 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4891 records point to non-existent hosts", "retry timeout exceeded", and
4892 "retry time not reached for any host after a long failure period".
4894 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4895 experimental DomainKeys support:
4897 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4898 (2) On an error such as an illegally used "control", the wrong name for
4899 the control was given.
4901 These problems did NOT occur unless DomainKeys support was compiled.
4903 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4905 PH/24 Added ${if match_ip condition.
4907 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4908 databases so that it will be absolutely obvious if a crash occurs in the
4909 DB library. This is a regular occurrence (often caused by mis-matched
4912 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4913 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4914 picked up on a user's system that detects such things. There doesn't seem
4915 to be a gcc warning option for this - only an attribute that has to be
4916 put on the function's prototype. It seems that in Fedora Core 4 they have
4917 set this on a number of new functions. No doubt there will be more in due
4920 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4921 specify a transport. However, if an address that was verified by such a
4922 router was the subject of a callout, Exim crashed because it tried to
4923 read the rcpt_include_affixes from the non-existent transport. Now it
4924 just assumes that the setting of that option is false. This bug was
4925 introduced by 4.51/PH/31.
4927 PH/28 Changed -d+all to exclude +memory, because that information is very
4928 rarely of interest, but it makes the output a lot bigger. People tend to
4929 do -d+all out of habit.
4931 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4932 code in os-type was giving problems when libc.so lives in lib64, like on
4935 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4936 aren't the modern standard, and it seems that some systems' include files
4937 don't always have them. Exim was already checking for some of the newer
4938 ones like T_AAAA, and defining it itself. I've added checks for all the
4939 record types that Exim uses.
4941 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4942 not automatically generating a new one, as it is supposed to. This
4943 prevented TLS from working. If the file did exist, but contained invalid
4944 data, a new version was generated, as expected. It was only the case of a
4945 non-existent file that was broken.
4947 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4948 with a change in libdomainkeys > 0.64.
4950 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4951 from DNS. If the selector record carries the flag, it now has
4952 precedence over the domain-wide flag.
4954 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4956 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4957 the use of a port name that isn't defined in /etc/services) occurred, the
4958 message was deferred as in a normal delivery, and thus remained on the
4959 spool, instead of being failed because of the mua_wrapper setting. This
4960 is now fixed, and I tidied up some of the mua_wrapper messages at the
4963 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4964 the messages in a hash of arrays rather than using individual hashes.
4965 This is a bit cleaner and results in dramatic memory savings, albeit
4966 at a slight CPU cost.
4968 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4969 as requested by Marc Sherman.
4971 SC/10 Eximstats: added histograms for user specified patterns as requested
4974 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4976 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4977 fopen() in the content-scanning modules that did not already have it.
4983 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4984 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4986 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4988 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4990 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4991 to test Sieve filters that use "vacation".
4993 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4994 that changes the way the GnuTLS parameters are stored in the cache file.
4995 The new format can be generated externally. For backward compatibility,
4996 if the data in the cache doesn't make sense, Exim assumes it has read an
4997 old-format file, and it generates new data and writes a new file. This
4998 means that you can't go back to an older release without removing the
5001 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5002 work if there are any delivery delays because "one_time" forces the
5003 parent to be marked "delivered", so its unseen clone is never tried
5004 again. For this reason, Exim now forbids the simultaneous setting of
5007 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5008 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5009 entry if you want to know the details. The fix, however, neglected to
5010 consider the case when local delivery batching is involved. The test for
5011 "previously delivered" was not happening when checking to see if an
5012 address could be batched with a previous (undelivered) one; under
5013 certain circumstances this could lead to multiple deliveries to the same
5016 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5017 in its include files, and this causes problems building Exim.
5019 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5020 header_syntax) but Exim was just ignoring anything given after a slash.
5021 In particular, this caused confusion with an attempt to use "verify =
5022 reverse_host_lookup/defer_ok". An error is now given when options are
5023 supplied for verify items that do not have them. (Maybe reverse_host_
5024 lookup should have a defer_ok option, but that's a different point.)
5026 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5027 defined by RFC 821) to 2048, because there were problems with some AUTH
5028 commands, and RFC 1869 says the size should be increased for extended
5029 SMTP commands that take arguments.
5031 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5034 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5035 "unknown" error; now it says that the functionality isn't in the binary.
5037 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5038 an address' error message when a string expansion fails (syntax or
5039 whatever). Otherwise the password may appear in the log. Following change
5040 PH/42 below, there is no longer a chance of it appearing in a bounce
5043 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5045 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5046 of its addresses was ever tried. (Bugzilla bug #2.)
5048 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5049 the result incorrectly in the debug output. (It correctly added a newline
5050 to what was transported.)
5052 TF/01 Added $received_time.
5054 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5055 commented out examples of how to interface to a virus scanner and to
5056 SpamAssassin. Also added commented examples of av_scanner and
5057 spamd_address settings.
5059 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5060 and controls are allowed in which ACLs. There were a couple of minor
5061 errors. Some of the entries in the conditions table (which is a table of
5062 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5063 negation of where the condition IS allowed.
5065 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5067 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5068 header file does not have a version number, so I've had to invent a new
5069 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5070 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5071 radiusclient), but it was contributed by a Radius user.
5073 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5074 files or directories, for OpenSSL.
5076 PH/19 When an Exim process that is running as root has to create an Exim log
5077 file, it does so in a subprocess that runs as exim:exim so as to get the
5078 ownership right at creation (otherwise, other Exim processes might see
5079 the file with the wrong ownership). There was no test for failure of this
5080 fork() call, which would lead to the process getting stuck as it waited
5081 for a non-existent subprocess. Forks do occasionally fail when resources
5082 run out. I reviewed all the other calls to fork(); they all seem to check
5085 PH/20 When checking for unexpected SMTP input at connect time (before writing
5086 the banner), Exim was not dealing correctly with a non-positive return
5087 from the read() function. If the client had disconnected by this time,
5088 the result was a log entry for a synchronization error with an empty
5089 string after "input=" when read() returned zero. If read() returned -1
5090 (an event I could not check), uninitialized data bytes were printed.
5091 There were reports of junk text (parts of files, etc) appearing after
5094 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5096 PH/22 Added support for macro redefinition, and (re)definition in between
5097 driver and ACL definitions.
5099 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5100 forgetting to use the resulting value; it was using the unexpanded value.
5102 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5103 hadn't been configured. The fix is from Juergen Kreileder, who
5104 understands it better than I do:
5106 "Here's what I see happening with three configured cyrus_sasl
5107 authenticators configured (plain, login, cram-md5):
5109 On startup auth_cyrus_sasl_init() gets called for each of these.
5110 This means three calls to sasl_listmech() without a specified mech_list.
5111 => SASL tests which mechs of all available mechs actually work
5112 => three warnings about OTP not working
5113 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5115 With the patch, sasl_listmech() also gets called three times. But now
5116 SASL's mech_list option is set to the server_mech specified in the the
5117 authenticator. Or in other words, the answer from sasl_listmech()
5118 gets limited to just the mech you're testing for (which is different
5120 => the return list contains just 'plain' or 'login', 'cram-md5' or
5121 nothing depending on the value of ob->server_mech.
5123 I've just tested the patch: Authentication still works fine,
5124 unavailable mechs specified in the exim configuration are still
5125 caught, and the auth.log warnings about OTP are gone."
5127 PH/25 When debugging is enabled, the contents of the command line are added
5128 to the debugging output, even when log_selector=+arguments is not
5131 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5132 answer is "GNU", and only if the return is "GNU/something" is the answer
5135 PH/27 $acl_verify_message is now set immediately after the failure of a
5136 verification in an ACL, and so is available in subsequent modifiers. In
5137 particular, the message can be preserved by coding like this:
5139 warn !verify = sender
5140 set acl_m0 = $acl_verify_message
5142 Previously, $acl_verify_message was set only while expanding "message"
5143 and "log_message" when a very denied access.
5145 PH/28 Modified OS/os.c-Linux with
5147 -#ifndef OS_LOAD_AVERAGE
5148 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5150 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5151 nomenclature these days.)
5153 PH/29 Installed patch from the Sieve maintainer that adds the options
5154 sieve_useraddress and sieve_subaddress to the redirect router.
5156 PH/30 In these circumstances:
5157 . Two addresses routed to the same list of hosts;
5158 . First host does not offer TLS;
5159 . First host accepts first address;
5160 . First host gives temporary error to second address;
5161 . Second host offers TLS and a TLS session is established;
5162 . Second host accepts second address.
5163 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5164 and peerdn, if requested) that were in fact used only for the second
5167 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5168 attention to any local part prefix or suffix that was matched by the
5169 router that accepted the address. It now behaves in the same way as it
5170 does for delivery: the affixes are removed from the local part unless
5171 rcpt_include_affixes is set on the transport.
5173 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5174 timeout during the DATA phase of an incoming message.
5176 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5177 applied a patch sanctioned by the Sieve maintainer.
5179 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5180 the uid or gid is negative. A case of a negative gid caused this to be
5181 noticed. The fix allows for either to be negative.
5183 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5184 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5187 PH/36 The change PH/12 above was broken. Fixed it.
5189 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5190 the grounds that routing the same address twice would always produce the
5191 same answer. This might have been true once, but it is certainly no
5192 longer true now. Routing a child address may depend on the previous
5193 routing that produced that child. Some complicated redirection strategies
5194 went wrong when messages had multiple recipients, and made Exim's
5195 behaviour dependent on the order in which the addresses were given.
5197 I have moved the duplicate checking until after the routing is complete.
5198 Exim scans the addresses that are assigned to local and remote
5199 transports, and removes any duplicates. This means that more work will be
5200 done, as duplicates will always all be routed, but duplicates are
5201 presumably rare, so I don't expect this is of any significance.
5203 For deliveries to pipes, files, and autoreplies, the duplicate checking
5204 still happens during the routing process, since they are not going to be
5207 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5208 It corrects a timeout issue with spamd. This is Ian's comment: "The
5209 background is that sometimes spamd either never reads data from a
5210 connection it has accepted, or it never writes response data. The exiscan
5211 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5212 blindly assumes that writes won't block so it may never time out."
5214 PH/39 Allow G after quota size as well as K and M.
5216 PH/40 The value set for $authenticated_id in an authenticator may not contain
5217 binary zeroes or newlines because the value is written to log lines and
5218 to spool files. There was no check on this. Now the value is run through
5219 the string_printing() function so that such characters are converted to
5220 printable escape sequences.
5222 PH/41 $message_linecount is a new variable that contains the total number of
5223 lines in the message. Compare $body_linecount, which is the count for the
5226 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5227 bounce and delay warning messages, except in certain special cases, which
5230 (a) An SMTP error message from a remote host;
5231 (b) A message specified in a :fail: redirection;
5232 (c) A message specified in a "fail" command in a system filter;
5233 (d) A message specified in a FAIL return from the queryprogram router;
5234 (e) A message specified by the cannot_route_message router option.
5236 In these cases only, Exim does include the error details in bounce and
5237 warning messages. There are also a few cases where bland messages such
5238 as "unrouteable address" or "local delivery error" are given.
5240 PH/43 $value is now also set for the "else" part of a ${run expansion.
5242 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5243 being worked on, but at least Exim now implements the latest version to
5246 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5247 process to complete was treated as a delivery failure, a timeout while
5248 writing the message to the pipe was logged, but erroneously treated as a
5249 successful delivery. Such timeouts include transport filter timeouts. For
5250 consistency with the overall process timeout, these timeouts are now
5251 treated as errors, giving rise to delivery failures by default. However,
5252 there is now a new Boolean option for the pipe transport called
5253 timeout_defer, which, if set TRUE, converts the failures into defers for
5254 both kinds of timeout. A transport filter timeout is now identified in
5257 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5258 systems where "make" and "gmake" are different, calling "gmake" at top
5259 level broke things. I've arranged for the value of $(MAKE) to be passed
5260 from the Makefile to this script so that it can call the same version of
5264 A note about Exim versions 4.44 and 4.50
5265 ----------------------------------------
5267 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5268 changes of various kinds. As a consequence, a big documentation update was
5269 needed. This delayed the release for rather longer than seemed good, especially
5270 in the light of a couple of (minor) security issues. Therefore, the changes
5271 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5272 release. So 4.44 and 4.50 are in effect two different branches that both start
5275 I have left the 4.50 change log unchanged; it contains all the changes since
5276 4.43. The change log for 4.44 is below; many of its items are identical to
5277 those for 4.50. This seems to be the most sensible way to preserve the
5278 historical information.
5284 1. Minor wording change to the doc/README.SIEVE file.
5286 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5287 computation of the current number of files was incorrect.
5289 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5290 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5293 4. Give more explanation in the error message when the command for a transport
5294 filter fails to execute.
5296 5. There are several places where Exim runs a non-Exim command in a
5297 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5298 was being done only for the command run by the queryprogram router. It is
5299 now done for all such subprocesses. The other cases are: ${run, transport
5300 filters, and the commands run by the lmtp and pipe transports.
5302 6. Added CONFIGURE_GROUP build-time option.
5304 7. Some older OS have a limit of 256 on the maximum number of file
5305 descriptors. Exim was using setrlimit() to set 1000 as a large value
5306 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5307 systems. I've change it so that if it can't get 1000, it tries for 256.
5309 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5310 was an oversight, and furthermore, ever since the addition of extra
5311 controls (e.g. 4.43/32), the checks on when to allow different forms of
5312 "control" were broken. There should now be diagnostics for all cases when a
5313 control that does not make sense is encountered.
5315 9. Added the /retain_sender option to "control=submission".
5317 10. $recipients is now available in the predata ACL (oversight).
5319 11. Tidy the search cache before the fork to do a delivery from a message
5320 received from the command line. Otherwise the child will trigger a lookup
5321 failure and thereby defer the delivery if it tries to use (for example) a
5322 cached ldap connection that the parent has called unbind on.
5324 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5325 of $address_data from the recipient verification was clobbered by the
5326 sender verification.
5328 13. The value of address_data from a sender verification is now available in
5329 $sender_address_data in subsequent conditions in the ACL statement.
5331 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5333 15. Added a new option "connect=<time>" to callout options, to set a different
5336 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5337 was its contents. (It was OK if the option was not defined at all.)
5339 17. A "Completed" log line is now written for messages that are removed from
5340 the spool by the -Mrm option.
5342 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5343 information about exactly what failed.
5345 19. Added -dd to debug only the daemon process.
5347 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5348 handles timeouts, both on the server side and network timeouts. Renamed the
5349 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5351 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5352 It is now set to "smtps".
5354 22. $host_address is now set to the target address during the checking of
5355 ignore_target_hosts.
5357 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5358 being passed; this would have caused $sender_host_name to have been used if
5359 matching the list had actually called for a host name (not very likely,
5360 since this list is usually IP addresses). A host name is now passed as
5363 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5364 code that specifies a non-restarting handler (typically sigaction() in
5365 modern systems) in an attempt to fix a rare and obscure crash bug.
5367 25. Narrowed the window for a race in the daemon that could cause it to ignore
5368 SIGCHLD signals. This is not a major problem, because they are used only to
5369 wake it up if nothing else does.
5371 26. A malformed maildirsize file could cause Exim to calculate negative values
5372 for the mailbox size or file count. Odd effects could occur as a result.
5373 The maildirsize information is now recalculated if the size or filecount
5376 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5377 support for a long time. Removed HAVE_SYS_VFS_H.
5379 28. Installed the latest version of exipick from John Jetmore.
5381 29. In an address list, if the pattern was not a regular expression, an empty
5382 subject address (from a bounce message) matched only if the pattern was an
5383 empty string. Non-empty patterns were not even tested. This was the wrong
5384 because it is perfectly reasonable to use an empty address as part of a
5385 database query. An empty address is now tested by patterns that are
5386 lookups. However, all the other forms of pattern expect the subject to
5387 contain a local part and a domain, and therefore, for them, an empty
5388 address still always fails if the pattern is not itself empty.
5390 30. Exim went into a mad DNS loop when attempting to do a callout where the
5391 host was specified on an smtp transport, and looking it up yielded more
5392 than one IP address.
5394 31. Re-factored the code for checking spool and log partition space into a
5395 function that finds that data and another that does the check. The former
5396 is then used to implement four new variables: $spool_space, $log_space,
5397 $spool_inodes, and $log_inodes.
5399 32. The RFC2047 encoding function was originally intended for short strings
5400 such as real names; it was not keeping to the 75-character limit for
5401 encoded words that the RFC imposes. It now respects the limit, and
5402 generates multiple encoded words if necessary. To be on the safe side, I
5403 have increased the buffer size for the ${rfc2047: expansion operator from
5406 33. It is now permitted to omit both strings after an "if" condition; if the
5407 condition is true, the result is "true". As before, when the second string
5408 is omitted, a false condition yields an empty string. This makes it less
5409 cumbersome to write custom ACL and router conditions.
5411 34. Failure to deliver a bounce message always caused it to be frozen, even if
5412 there was an errors_to setting on the router. The errors_to setting is now
5415 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5416 canonical form (fully expanded) before being placed in
5417 $sender_host_address.
5419 36. The table in the code that translates DNS record types into text (T_A to
5420 "A" for instance) was missing entries for NS and CNAME. It is just possible
5421 that this could have caused confusion if both these types were looked up
5422 for the same domain, because the text type is used as part of Exim's
5423 per-process caching. But the chance of anyone hitting this buglet seems
5426 37. The dnsdb lookup has been extended in a number of ways.
5428 (1) There is a new type, "zns", which walks up the domain tree until it
5429 finds some nameserver records. It should be used with care.
5431 (2) There is a new type, "mxh", which is like "mx" except that it returns
5432 just the host names, not the priorities.
5434 (3) It is now possible to give a list of domains (or IP addresses) to be
5435 looked up. The behaviour when one of the lookups defers can be
5436 controlled by a keyword.
5438 (4) It is now possible to specify the separator character for use when
5439 multiple records are returned.
5441 38. The dnslists ACL condition has been extended: it is now possible to supply
5442 a list of IP addresses and/or domains to be looked up in a particular DNS
5445 39. Added log_selector=+queue_time_overall.
5447 40. When running the queue in the test harness, wait just a tad after forking a
5448 delivery process, to get repeatability of debugging output.
5450 41. Include certificate and key file names in error message when GnuTLS fails
5451 to set them up, because the GnuTLS error message doesn't include the name
5452 of the failing file when there is a problem reading it.
5454 42. Allow both -bf and -bF in the same test run.
5456 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5458 44. The "Exiscan patch" is now merged into the mainline Exim source.
5460 45. Sometimes the final signoff response after QUIT could fail to get
5461 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5462 < 0 before doing a fflush(). This bug looks as though it goes back to the
5463 introduction of TLS in release 3.20, but "sometimes" must have been rare
5464 because the tests only now provoked it.
5466 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5467 (this can affect the format of dates).
5469 47. exim_tidydb, when checking for the continued existence of a message for
5470 which it has found a message-specific retry record, was not finding
5471 messages that were in split spool directories. Consequently, it was
5472 deleting retry records that should have stayed in existence.
5474 48. Steve fixed some bugs in eximstats.
5476 49. The SPA authentication driver was not abandoning authentication and moving
5477 on to the next authenticator when an expansion was forced to fail,
5478 contradicting the general specification for all authenticators. Instead it
5479 was generating a temporary error. It now behaves as specified.
5481 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5482 (the order specifies the preference for clients). The order is now AES256,
5483 AES128, 3DES, ARCFOUR128.
5485 51. Small patch to Sieve code - explicitly set From: when generating an
5488 52. Exim crashed if a remote delivery caused a very long error message to be
5489 recorded - for instance if somebody sent an entire SpamAssassin report back
5490 as a large number of 550 error lines. This bug was coincidentally fixed by
5491 increasing the size of one of Exim's internal buffers (big_buffer) that
5492 happened as part of the Exiscan merge. However, to be on the safe side, I
5493 have made the code more robust (and fixed the comments that describe what
5496 53. Now that there can be additional text after "Completed" in log lines (if
5497 the queue_time_overall log selector is set), a one-byte patch to exigrep
5498 was needed to allow it to recognize "Completed" as not the last thing in
5501 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5502 patch that reportedly fixes this has been added. I am not expert enough to
5503 create a test for it. This is what the patch creator wrote:
5505 "I found a little strange behaviour of ldap code when working with
5506 Windows 2003 AD Domain, where users was placed in more than one
5507 Organization Units. When I tried to give exim partial DN, the exit code
5508 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5509 But simultaneously result of request was absolutely normal ldap result,
5510 so I produce this patch..."
5512 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5513 REFERENCE, so I have modified the code to exclude the patch when that macro
5516 55. Some experimental protocols are using DNS PTR records for new purposes. The
5517 keys for these records are domain names, not reversed IP addresses. The
5518 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5519 leaves it alone. Component reversal etc. now happens only for IP addresses.
5522 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5524 57. Double the size of the debug message buffer (to 2048) so that more of very
5525 long debug lines gets shown.
5527 58. The exicyclog utility now does better if the number of log files to keep
5528 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5530 59. Two changes related to the smtp_active_hostname option:
5532 (1) $smtp_active_hostname is now available as a variable.
5533 (2) The default for smtp_banner uses $smtp_active_hostname instead
5534 of $primary_hostname.
5536 60. The host_aton() function is supposed to be passed a string that is known
5537 to be a valid IP address. However, in the case of IPv6 addresses, it was
5538 not checking this. This is a hostage to fortune. Exim now panics and dies
5539 if the condition is not met. A case was found where this could be provoked
5540 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5541 components; fortuitously, this particular loophole had already been fixed
5542 by change 4.50/55 above.
5544 If there are any other similar loopholes, the new check in host_aton()
5545 itself should stop them being exploited. The report I received stated that
5546 data on the command line could provoke the exploit when Exim was running as
5547 exim, but did not say which command line option was involved. All I could
5548 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5549 running as the user.
5552 61. There was a buffer overflow vulnerability in the SPA authentication code
5553 (which came originally from the Samba project). I have added a test to the
5554 spa_base64_to_bits() function which I hope fixes it.
5557 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5558 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5560 63. The daemon start-up calls getloadavg() while still root for those OS that
5561 need the first call to be done as root, but it missed one case: when
5562 deliver_queue_load_max is set with deliver_drop_privilege. This is
5563 necessary for the benefit of the queue runner, because there is no re-exec
5564 when deliver_drop_privilege is set.
5566 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5567 This has been fixed.
5569 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5570 was in use, was not putting the data itself into the right store pool;
5571 consequently, it could be overwritten for a subsequent message in the same
5572 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5575 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5577 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5578 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5579 the function were treating the return as a boolean value, which happened to
5580 work because 0=false and not-0=true, but is not correct code.
5582 68. The host_aton() function was not handling scoped IPv6 addresses (those
5583 with, for example, "%eth0" on the end) correctly.
5585 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5586 negated items (that is, ~something) in unsigned ints. Some compilers
5587 apparently mutter when there is no cast.
5589 70. If an address verification called from an ACL failed, and did not produce a
5590 user-specific message (i.e. there was only a "system" message), nothing was
5591 put in $acl_verify_message. In this situation, it now puts the system
5594 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5595 session; change 4.31/43 added the unwanted input to the log line - except
5596 that it did not do this in the start of session case. It now does.
5598 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5599 This is wrong and can cause the other end to generate a synchronization
5600 error if it is another Exim or anything else that does the synchronization
5601 check. A QUIT command is no longer sent after a timeout.
5603 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5604 during host lookups.
5606 74. The defer_ok option of callout verification was not working if it was used
5607 when verifying addresses in header lines, that is, for this case:
5609 verify = header_sender/callout=defer_ok
5611 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5612 those file descriptors could be used for SMTP connections. If anything
5613 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5614 could be sent to the SMTP client, causing chaos. The daemon now opens
5615 stdin, stdout, and stderr to /dev/null when it puts itself into the
5618 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5619 log by default. The user can override this with suitable Perl magic.
5621 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5622 the log message when discard triggers, was not working for the DATA ACL or
5623 for the non-SMTP ACL.
5625 78. Error message wording change in sieve.c.
5627 79. If smtp_accept_max_per_host was set, the number of connections could be
5628 restricted to fewer than expected, because the daemon was trying to set up
5629 a new connection before checking whether the processes handling previous
5630 connections had finished. The check for completed processes is now done
5631 earlier. On busy systems, this bug wouldn't be noticed because something
5632 else would have woken the daemon, and it would have reaped the completed
5635 80. If a message was submitted locally by a user whose login name contained one
5636 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5637 It caused a spool format error. I have fixed the spool reading code. A
5638 related problem was that the "from" clause in the Received: line became
5639 illegal because of the space(s). It is now covered by ${quote_local_part.
5641 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5644 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5646 83. Patch from Sieve maintainer to fix unterminated string problem in
5647 "vacation" handling.
5649 84. Some minor changes to the Linux configuration files to help with other
5650 OS variants using glibc.
5652 85. One more patch for Sieve to update vacation handling to latest spec.
5655 ----------------------------------------------------
5656 See the note above about the 4.44 and 4.50 releases.
5657 ----------------------------------------------------
5663 1. Change 4.43/35 introduced a bug that caused file counts to be
5664 incorrectly computed when quota_filecount was set in an appendfile
5667 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5668 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5671 3. Give more explanation in the error message when the command for a transport
5672 filter fails to execute.
5674 4. There are several places where Exim runs a non-Exim command in a
5675 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5676 was being done only for the command run by the queryprogram router. It is
5677 now done for all such subprocesses. The other cases are: ${run, transport
5678 filters, and the commands run by the lmtp and pipe transports.
5680 5. Some older OS have a limit of 256 on the maximum number of file
5681 descriptors. Exim was using setrlimit() to set 1000 as a large value
5682 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5683 systems. I've change it so that if it can't get 1000, it tries for 256.
5685 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5686 was an oversight, and furthermore, ever since the addition of extra
5687 controls (e.g. 4.43/32), the checks on when to allow different forms of
5688 "control" were broken. There should now be diagnostics for all cases when a
5689 control that does not make sense is encountered.
5691 7. $recipients is now available in the predata ACL (oversight).
5693 8. Tidy the search cache before the fork to do a delivery from a message
5694 received from the command line. Otherwise the child will trigger a lookup
5695 failure and thereby defer the delivery if it tries to use (for example) a
5696 cached ldap connection that the parent has called unbind on.
5698 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5699 of $address_data from the recipient verification was clobbered by the
5700 sender verification.
5702 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5703 was its contents. (It was OK if the option was not defined at all.)
5705 11. A "Completed" log line is now written for messages that are removed from
5706 the spool by the -Mrm option.
5708 12. $host_address is now set to the target address during the checking of
5709 ignore_target_hosts.
5711 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5712 being passed; this would have caused $sender_host_name to have been used if
5713 matching the list had actually called for a host name (not very likely,
5714 since this list is usually IP addresses). A host name is now passed as
5717 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5718 code that specifies a non-restarting handler (typically sigaction() in
5719 modern systems) in an attempt to fix a rare and obscure crash bug.
5721 15. Narrowed the window for a race in the daemon that could cause it to ignore
5722 SIGCHLD signals. This is not a major problem, because they are used only to
5723 wake it up if nothing else does.
5725 16. A malformed maildirsize file could cause Exim to calculate negative values
5726 for the mailbox size or file count. Odd effects could occur as a result.
5727 The maildirsize information is now recalculated if the size or filecount
5730 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5731 support for a long time. Removed HAVE_SYS_VFS_H.
5733 18. Updated exipick to current release from John Jetmore.
5735 19. Allow an empty sender to be matched against a lookup in an address list.
5736 Previously the only cases considered were a regular expression, or an
5739 20. Exim went into a mad DNS lookup loop when doing a callout where the
5740 host was specified on the transport, if the DNS lookup yielded more than
5743 21. The RFC2047 encoding function was originally intended for short strings
5744 such as real names; it was not keeping to the 75-character limit for
5745 encoded words that the RFC imposes. It now respects the limit, and
5746 generates multiple encoded words if necessary. To be on the safe side, I
5747 have increased the buffer size for the ${rfc2047: expansion operator from
5750 22. Failure to deliver a bounce message always caused it to be frozen, even if
5751 there was an errors_to setting on the router. The errors_to setting is now
5754 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5755 canonical form (fully expanded) before being placed in
5756 $sender_host_address.
5758 24. Updated eximstats to version 1.33
5760 25. Include certificate and key file names in error message when GnuTLS fails
5761 to set them up, because the GnuTLS error message doesn't include the name
5762 of the failing file when there is a problem reading it.
5764 26. Expand error message when OpenSSL has problems setting up cert/key files.
5767 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5768 (this can affect the format of dates).
5770 28. exim_tidydb, when checking for the continued existence of a message for
5771 which it has found a message-specific retry record, was not finding
5772 messages that were in split spool directories. Consequently, it was
5773 deleting retry records that should have stayed in existence.
5775 29. eximstats updated to version 1.35
5776 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5777 1.35 - bugfix such that pie charts by volume are generated correctly
5779 30. The SPA authentication driver was not abandoning authentication and moving
5780 on to the next authenticator when an expansion was forced to fail,
5781 contradicting the general specification for all authenticators. Instead it
5782 was generating a temporary error. It now behaves as specified.
5784 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5785 (the order specifies the preference for clients). The order is now AES256,
5786 AES128, 3DES, ARCFOUR128.
5788 31. Small patch to Sieve code - explicitly set From: when generating an
5791 32. Exim crashed if a remote delivery caused a very long error message to be
5792 recorded - for instance if somebody sent an entire SpamAssassin report back
5793 as a large number of 550 error lines. This bug was coincidentally fixed by
5794 increasing the size of one of Exim's internal buffers (big_buffer) that
5795 happened as part of the Exiscan merge. However, to be on the safe side, I
5796 have made the code more robust (and fixed the comments that describe what
5799 33. Some experimental protocols are using DNS PTR records for new purposes. The
5800 keys for these records are domain names, not reversed IP addresses. The
5801 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5802 leaves it alone. Component reversal etc. now happens only for IP addresses.
5805 34. The host_aton() function is supposed to be passed a string that is known
5806 to be a valid IP address. However, in the case of IPv6 addresses, it was
5807 not checking this. This is a hostage to fortune. Exim now panics and dies
5808 if the condition is not met. A case was found where this could be provoked
5809 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5810 components; fortuitously, this particular loophole had already been fixed
5811 by change 4.50/55 or 4.44/33 above.
5813 If there are any other similar loopholes, the new check in host_aton()
5814 itself should stop them being exploited. The report I received stated that
5815 data on the command line could provoke the exploit when Exim was running as
5816 exim, but did not say which command line option was involved. All I could
5817 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5818 running as the user.
5821 35. There was a buffer overflow vulnerability in the SPA authentication code
5822 (which came originally from the Samba project). I have added a test to the
5823 spa_base64_to_bits() function which I hope fixes it.
5826 36. The daemon start-up calls getloadavg() while still root for those OS that
5827 need the first call to be done as root, but it missed one case: when
5828 deliver_queue_load_max is set with deliver_drop_privilege. This is
5829 necessary for the benefit of the queue runner, because there is no re-exec
5830 when deliver_drop_privilege is set.
5832 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5833 was in use, was not putting the data itself into the right store pool;
5834 consequently, it could be overwritten for a subsequent message in the same
5835 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5838 38. Sometimes the final signoff response after QUIT could fail to get
5839 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5840 < 0 before doing a fflush(). This bug looks as though it goes back to the
5841 introduction of TLS in release 3.20, but "sometimes" must have been rare
5842 because the tests only now provoked it.
5848 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5849 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5850 Now it returns an integer. A number of calls were still expecting a T/F
5851 return. Fortuitously, in all cases, the tests worked in OK situations,
5852 which is the norm. However, things would have gone wrong on any write
5853 failures on the smtp file descriptor. This function is used when sending
5854 messages over SMTP and also when doing verify callouts.
5856 2. When Exim is called to do synchronous delivery of a locally submitted
5857 message (the -odf or -odi options), it no longer closes stderr before doing
5860 3. Implemented the mua_wrapper option.
5862 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5864 5. Implemented the functions header_remove(), header_testname(),
5865 header_add_at_position(), and receive_remove_recipient(), and exported them
5868 6. If an ACL "warn" statement specified the addition of headers, Exim already
5869 inserted X-ACL-Warn: at the start if there was no header name. However, it
5870 was not making this test for the second and subsequent header lines if
5871 there were newlines in the string. This meant that an invalid header could
5872 be inserted if Exim was badly configured.
5874 7. Allow an ACL "warn" statement to add header lines at the start or after all
5875 the Received: headers, as well as at the end.
5877 8. Added the rcpt_4xx retry error code.
5879 9. Added postmaster_mailfrom=xxx to callout verification option.
5881 10. Added mailfrom=xxxx to the callout verification option, for verify=
5884 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5885 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5887 12. Inserted some casts to stop certain compilers warning when using pointer
5888 differences as field lengths or precisions in printf-type calls (mostly
5889 affecting debugging statements).
5891 13. Added optional readline() support for -be (dynamically loaded).
5893 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5894 same clock tick as a message's arrival, so that its received time was the
5895 same as the "first fail" time on the retry record, and that message
5896 remained on the queue past the ultimate address timeout, every queue runner
5897 would try a delivery (because it was past the ultimate address timeout) but
5898 after another failure, the ultimate address timeout, which should have then
5899 bounced the address, did not kick in. This was a "< instead of <=" error;
5900 in most cases the first failure would have been in the next clock tick
5901 after the received time, and all would be well.
5903 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5904 being recognized when the domain list was tested by the match_domain
5905 condition in an expansion string.
5907 16. Added the ${str2b64: operator.
5909 17. Exim was always calling setrlimit() to set a large limit for the number of
5910 processes, without checking whether the existing limit was already
5911 adequate. (It did check for the limit on file descriptors.) Furthermore,
5912 errors from getrlimit() and setrlimit() were being ignored. Now they are
5913 logged to the main and panic logs, but Exim does carry on, to try to do its
5914 job under whatever limits there are.
5916 18. Imported PCRE 5.0.
5918 19. Trivial typo in log message " temporarily refused connection" (the leading
5921 20. If the log selector return_path_on_delivery was set and an address was
5922 redirected to /dev/null, the delivery process crashed because it assumed
5923 that a return path would always be set for a "successful" delivery. In this
5924 case, the whole delivery is bypassed as an optimization, and therefore no
5927 21. Internal re-arrangement: the function for sending a challenge and reading
5928 a response while authentication was assuming a zero-terminated challenge
5929 string. It's now changed to take a pointer and a length, to allow for
5930 binary data in such strings.
5932 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5934 23. Exim was not respecting finduser_retries when seeking the login of the
5935 uid under which it was called; it was always trying 10 times. (The default
5936 setting of finduser_retries is zero.) Also, it was sleeping after the final
5937 failure, which is pointless.
5939 24. Implemented tls_on_connect_ports.
5941 25. Implemented acl_smtp_predata.
5943 26. If the domain in control=submission is set empty, Exim assumes that the
5944 authenticated id is a complete email address when it generates From: or
5945 Sender: header lines.
5947 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5948 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5949 chown and chgrp in /bin and hostname in /usr/bin.
5951 28. Exim was keeping the "process log" file open after each use, just as it
5952 does for the main log. This opens the possibility of it remaining open for
5953 long periods when the USR1 signal hits a daemon. Occasional processlog
5954 errors were reported, that could have been caused by this. Anyway, it seems
5955 much more sensible not to leave this file open at all, so that is what now
5958 29. The long-running daemon process does not normally write to the log once it
5959 has entered its main loop, and it closes the log before doing so. This is
5960 so that log files can straightforwardly be renamed and moved. However,
5961 there are a couple of unusual error situations where the daemon does write
5962 log entries, and I had neglected to close the log afterwards.
5964 30. The text of an SMTP error response that was received during a remote
5965 delivery was being truncated at 512 bytes. This is too short for some of
5966 the long messages that one sometimes sees. I've increased the limit to
5969 31. It is now possible to make retry rules that apply only when a message has a
5970 specific sender, in particular, an empty sender.
5972 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5973 it possible to be selective about when SMTP synchronization is enforced.
5975 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5977 32. Implemented hosts_connection_nolog.
5979 33. Added an ACL for QUIT.
5981 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5984 35. Added mailbox_size and mailbox_filecount to appendfile.
5986 36. Added control = no_multiline_responses to ACLs.
5988 37. There was a bug in the logic of the code that waits for the clock to tick
5989 in the case where the clock went backwards by a substantial amount such
5990 that the microsecond fraction of "now" was more than the microsecond
5991 fraction of "then" (but the whole seconds number was less).
5993 38. Added support for the libradius Radius client library this is found on
5994 FreeBSD (previously only the radiusclient library was supported).
6000 1. When certain lookups returned multiple values in the form name=value, the
6001 quoting of the values was not always being done properly. Specifically:
6002 (a) If the value started with a double quote, but contained no whitespace,
6004 (b) If the value contained whitespace other than a space character (i.e.
6005 tabs or newlines or carriage returns) it was not quoted.
6006 This fix has been applied to the mysql and pgsql lookups by writing a
6007 separate quoting function and calling it from the lookup code. The fix
6008 should probably also be applied to nisplus, ibase and oracle lookups, but
6009 since I cannot test any of those, I have not disturbed their existing code.
6011 2. A hit in the callout cache for a specific address caused a log line with no
6012 reason for rejecting RCPT. Now it says "Previous (cached) callout
6013 verification failure".
6015 3. There was an off-by-one bug in the queryprogram router. An over-long
6016 return line was truncated at 256 instead of 255 characters, thereby
6017 overflowing its buffer with the terminating zero. As well as fixing this, I
6018 have increased the buffer size to 1024 (and made a note to document this).
6020 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6021 when Exim is waiting for an SMTP response from a remote server, Exim
6022 restarts its select() call on the socket, thereby resetting its timeout.
6023 This is not a problem when such interrupts are rare. Somebody set up a cron
6024 job to run exiwhat every 2 minutes, which is less than the normal select()
6025 timeout (5 or 10 minutes). This meant that the select() timeout never
6026 kicked in because it was always reset. I have fixed this by comparing the
6027 time when an interrupt arrives with the time at the start of the first call
6028 to select(). If more time than the timeout has elapsed, the interrupt is
6029 treated as a timeout.
6031 5. Some internal re-factoring in preparation for the addition of Sieve
6032 extensions (by MH). In particular, the "personal" test is moved to a
6033 separate function, and given an option for scanning Cc: and Bcc: (which is
6034 not set for Exim filters).
6036 6. When Exim created an email address using the login of the caller as the
6037 local part (e.g. when creating a From: or Sender: header line), it was not
6038 quoting the local part when it contained special characters such as @.
6040 7. Installed new OpenBSD configuration files.
6042 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6043 try to make them clearer.
6045 9. Callout options, other than the timeout value, were being ignored when
6046 verifying sender addresses in header lines. For example, when using
6048 verify = header_sender/callout=no_cache
6050 the cache was (incorrectly) being used.
6052 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6053 only the Cygwin environment.
6055 11. When return_path_on_delivery was set as a log selector, if different remote
6056 addresses in the same message used different return paths and parallel
6057 remote delivery occurred, the wrong values would sometimes be logged.
6058 (Whenever a remote delivery process finished, the return path value from
6059 the most recently started remote delivery process was logged.)
6061 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6062 lines when AUTH and/or TLS are in use. This is the "received protocol"
6063 field. Exim used to use "asmtp" for authenticated SMTP, without any
6064 indication (in the protocol name) for TLS use. Now it follows the RFC and
6065 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6066 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6067 names appear in log lines as well as in Received: header lines.
6069 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6070 extensions, and comparison tests, and to fix some bugs.
6072 14. Changes to the "personal" filter test:
6074 (1) The test was buggy in that it was just doing the equivalent of
6075 "contains" tests on header lines. For example, if a user's address was
6076 anne@some.where, the "personal" test would incorrectly be true for
6078 To: susanne@some.where
6080 This test is now done by extracting each address from the header in turn,
6081 and checking the entire address. Other tests that are part of "personal"
6082 are now done using regular expressions (for example, to check local parts
6083 of addresses in From: header lines).
6085 (2) The list of non-personal local parts in From: addresses has been
6086 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6087 taken from the Sieve specification recommendations.
6089 (3) If the message contains any header line starting with "List-" it is
6090 treated as non-personal.
6092 (4) The test for "circular" in the Subject: header line has been removed
6093 because it now seems ill-conceived.
6095 15. Minor typos in src/EDITME comments corrected.
6097 16. Installed latest exipick from John Jetmore.
6099 17. If headers_add on a router specified a text string that was too long for
6100 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6101 of string_sprintf() is now avoided.
6103 18. $message_body_size was not set (it was always zero) when running the DATA
6104 ACL and the local_scan() function.
6106 19. For the "mail" command in an Exim filter, no default was being set for
6107 the once_repeat time, causing a random time value to be used if "once" was
6108 specified. (If the value happened to be <= 0, no repeat happened.) The
6109 default is now 0s, meaning "never repeat". The "vacation" command was OK
6110 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6111 (I found it when inspecting the code).
6113 20. There is now an overall timeout for performing a callout verification. It
6114 defaults to 4 times the callout timeout, which applies to individual SMTP
6115 commands during the callout. The overall timeout applies when there is more
6116 than one host that can be tried. The timeout is checked before trying the
6117 next host. This prevents very long delays if there are a large number of
6118 hosts and all are timing out (e.g. when the network connections are timing
6119 out). The value of the overall timeout can be changed by specifying an
6120 additional sub-option for "callout", called "maxwait". For example:
6122 verify = sender/callout=5s,maxwait=20s
6124 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6125 to the end before writing, but this should make it even safer).
6127 22. Exim was forgetting that it had advertised PIPELINING for the second and
6128 subsequent messages on an SMTP connection. It was also not resetting its
6129 memory on STARTTLS and an internal HELO.
6131 23. When Exim logs an SMTP synchronization error within a session, it now
6132 records whether PIPELINING has been advertised or not.
6134 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6135 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6136 rather than long int.
6138 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6140 26. Added the never_mail option to autoreply.
6146 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6147 crash if the getsockname() call failed; this can happen if a connection is
6148 closed very soon after it is established. The problem was simply in the
6149 order in which certain operations were done, causing Exim to try to write
6150 to the SMTP stream before it had set up the file descriptor. The bug has
6151 been fixed by making things happen in the correct order.
6157 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6158 before the connection was closed, thus losing the rejection response.
6160 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6161 some early Solaris releases, but causes trouble in current releases where
6162 socklen_t is defined.
6164 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6167 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6170 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6171 that releasing the top of it at the end releases what was used for sub-
6172 expansions (unless the block got too big). However, discard this block if
6173 the first thing is a variable or header, so that we can use its block when
6174 it is dynamic (useful for very large $message_headers, for example).
6176 6. Lookups now cache *every* query, not just the most recent. A new, separate
6177 store pool is used for this. It can be recovered when all lookup caches are
6178 flushed. Lookups now release memory at the end of their result strings.
6179 This has involved some general refactoring of the lookup sources.
6181 7. Some code has been added to the store_xxx() functions to reduce the amount
6182 of flapping under certain conditions.
6184 8. log_incoming_interface used to affect only the <= reception log lines. Now
6185 it causes the local interface and port to be added to several more SMTP log
6186 lines, for example "SMTP connection from", and rejection lines.
6188 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6190 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6192 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6193 could be overwritten at the end of the current message (or the start of a
6194 new message if it was set in a HELO ACL). The value is now preserved for
6195 the duration of the SMTP connection.
6197 12. If a transport had a headers_rewrite setting, and a matching header line
6198 contained an unqualified address, that address was qualified, even if it
6199 did not match any rewriting rules. The underlying bug was that the values
6200 of the flags that permit the existence of unqualified sender and recipient
6201 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6202 non-local messages, and by -bnq for local messages) were not being
6203 preserved with the message after it was received.
6205 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6206 "next input=" as part of the text comprising the host identity instead of
6207 the correct text. The code was using the same buffer for two different
6208 strings. However, depending on which order the printing function evaluated
6209 its arguments, the bug did not always show up. Under Linux, for example, my
6210 test suite worked just fine.
6212 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6213 doesn't work with some older versions of Perl. It has been changed to "my",
6214 which in any case is probably the better facility to use.
6216 15. A really picky compiler found some instances of statements for creating
6217 error messages that either had too many or two few arguments for the format
6220 16. The size of the buffer for calls to the DNS resolver has been increased
6221 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6222 for addresses that have a lot of PTR records. This alleviates a problem; it
6223 does not fully solve it.
6225 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6226 buffer now truncates the list and logs the incident, which is the same
6227 action as happens when Exim is looking up a host name and its aliases.
6228 Previously in this situation something unpredictable would happen;
6229 sometimes it was "internal error: store_reset failed".
6231 18. If a server dropped the connection unexpectedly when an Exim client was
6232 using GnuTLS and trying to read a response, the client delivery process
6233 crashed while trying to generate an error log message.
6235 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6236 string, for example:
6238 warn message = H1: something\nH2: something
6240 the text was added as a single header line from Exim's point of view
6241 though it ended up OK in the delivered message. However, searching for the
6242 second and subsequent header lines using $h_h2: did not work. This has been
6243 fixed. Similarly, if a system filter added multiple headers in this way,
6244 the routers could not see them.
6246 20. Expanded the error message when iplsearch is called with an invalid key to
6247 suggest using net-iplsearch in a host list.
6249 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6250 ACLs are no longer actually imposed (and a message to that effect is
6253 22. If a "gecos" field in a passwd entry contained escaped characters, in
6254 particular, if it contained a \" sequence, Exim got it wrong when building
6255 a From: or a Sender: header from that name. A second bug also caused
6256 incorrect handling when an unquoted " was present following a character
6257 that needed quoting.
6259 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6260 was not being matched caselessly.
6262 24. Arranged for all hyphens in the exim.8 source to be escaped with
6265 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6266 properly. Recipient callout cache records were still being keyed to include
6267 the sender, even when use_sender was set false. This led to far more
6268 callouts that were necessary. The sender is no longer included in the key
6269 when use_sender is false.
6271 26. Added "control = submission" modifier to ACLs.
6273 27. Added the ${base62d: operator to decode base 62 numbers.
6275 28. dnsdb lookups can now access SRV records.
6277 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6278 the configuration file.
6280 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6281 (-v) mode. This makes the output for a verbose queue run more intelligible.
6283 31. Added a use_postmaster feature to recipient callouts.
6285 32. Added the $body_zerocount variable, containing the number of binary zero
6286 bytes in the message body.
6288 33. The time of last modification of the "new" subdirectory is now used as the
6289 "mailbox time last read" when there is a quota error for a maildir
6292 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6294 35. Added +ignore_unknown as a special item in host lists.
6296 36. Code for decoding IPv6 addresses in host lists is now included, even if
6297 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6298 address was recognized as an IP address, but was then not correctly decoded
6299 into binary, causing unexpected and incorrect effects when compared with
6306 1. Very minor rewording of debugging text in manualroute to say "list of
6307 hosts" instead of "hostlist".
6309 2. If verify=header_syntax was set, and a header line with an unqualified
6310 address (no domain) and a large number of spaces between the end of the
6311 name and the colon was received, the reception process suffered a buffer
6312 overflow, and (when I tested it) crashed. This was caused by some obsolete
6313 code that should have been removed. The fix is to remove it!
6315 3. When running in the test harness, delay a bit after writing a bounce
6316 message to get a bit more predictability in the log output.
6318 4. Added a call to search_tidyup() just before forking a reception process. In
6319 theory, someone could use a lookup in the expansion of smtp_accept_max_
6320 per_host which, without the tidyup, could leave open a database connection.
6322 5. Added the variables $recipient_data and $sender_data which get set from a
6323 lookup success in an ACL "recipients" or "senders" condition, or a router
6324 "senders" option, similar to $domain_data and $local_part_data.
6326 6. Moved the writing of debug_print from before to after the "senders" test
6329 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6330 problems for message scanning, either using a data ACL, or using
6331 local_scan() because the Received: header was not generated till after they
6332 were called (in order to set the time as the time of reception completion).
6333 I have revised the way this works. The header is now generated after the
6334 body is received, but before the ACL or local_scan() are called. After they
6335 are run, the timestamp in the header is updated.
6341 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6342 before starting a queue runner without re-exec. This happened only when
6343 deliver_drop_privilege was set or when the Exim user was set to root. The
6344 effect of the bug was that timeouts during subsequent deliveries caused
6345 crashes instead of being properly handled. The handler is now left at its
6346 default (and expected) setting.
6348 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6349 message, again when deliver_drop_privilege is set or Exim is run as root.
6350 The bug described in (1) was not present in this case, but the tidying up
6351 of the other signals was missing. I have made the two cases consistent.
6353 3. The ignore_target_hosts setting on a manualroute router was being ignored
6354 for hosts that were looked up using the /MX notation.
6356 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6359 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6360 operated on the sender address. After changing the $sender_address to <>
6361 for the sender address verify, Exim was re-instated it as the original
6362 (before rewriting) address, but remembering that it had rewritten it, so it
6363 wasn't rewriting it again. This bug also had the effect of breaking the
6364 sender address verification caching when the sender address was rewritten.
6366 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6367 This has been changed so that if the ip literal address matches
6368 ignore_target_hosts, the router declines.
6370 7. Added expansion conditions match_domain, match_address, and match_local_
6371 part (NOT match_host).
6373 8. The placeholder for the Received: header didn't have a length field set.
6375 9. Added code to Exim itself and to exim_lock to test for a specific race
6376 condition that could lead to file corruption when using MBX delivery. The
6377 issue is with the lockfile that is created in /tmp. If this file is removed
6378 after a process has opened it but before that process has acquired a lock,
6379 there is the potential for a second process to recreate the file and also
6380 acquire a lock. This could lead to two Exim processes writing to the file
6381 at the same time. The added code performs the same test as UW imapd; it
6382 checks after acquiring the lock that its file descriptor still refers to
6383 the same named file.
6385 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6386 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6389 11. Added the smtp_active_hostname option. If used, this will typically be made
6390 to depend on the incoming interface address. Because $interface_address is
6391 not set up until the daemon has forked a reception process, error responses
6392 that can happen earlier (such as "too many connections") no longer contain
6395 12. If an expansion in a condition on a "warn" statement fails because a lookup
6396 defers, the "warn" statement is abandoned, and the next ACL statement is
6397 processed. Previously this caused the whole ACL to be aborted.
6399 13. Added the iplsearch lookup type.
6401 14. Added ident_timeout as a log selector.
6403 15. Added tls_certificate_verified as a log selector.
6405 16. Added a global option tls_require_ciphers (compare the smtp transport
6406 option of the same name). This controls incoming TLS connections.
6408 17. I finally figured out how to make tls_require_ciphers do a similar thing
6409 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6410 before starting the TLS session.
6412 18. Tabs are now shown as \t in -bP output.
6414 19. If the log selector return_path_on_delivery was set, Exim crashed when
6415 bouncing a message because it had too many Received: header lines.
6417 20. If two routers both had headers_remove settings, and the first one included
6418 a superfluous trailing colon, the final name in the first list and the
6419 first name in the second list were incorrectly joined into one item (with a
6420 colon in the middle).
6426 1. Added -C and -D options to the exinext utility, mainly to make it easier
6427 to include in the automated testing, but these could be helpful when
6428 multiple configurations are in use.
6430 2. The exinext utility was not formatting the output nicely when there was
6431 an alternate port involved in the retry record key, nor when there was a
6432 message id as well (for retries that were specific to a specific message
6433 and a specific host). It was also confused by IPv6 addresses, because of
6434 the additional colons they contain. I have fixed the IPv4 problem, and
6435 patched it up to do a reasonable job for IPv6.
6437 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6438 delivery, the log line now contains "pipelined" if PIPELINING was used.
6440 4. An SMTP transport process used to panic and die if the bind() call to set
6441 an explicit outgoing interface failed. This has been changed; it is now
6442 treated in the same way as a connect() failure.
6444 5. A reference to $sender_host_name in the part of a conditional expansion
6445 that was being skipped was still causing a DNS lookup. This no longer
6448 6. The def: expansion condition was not recognizing references to header lines
6449 that used bh_ and bheader_.
6451 7. Added the _cache feature to named lists.
6453 8. The code for checking quota_filecount in the appendfile transport was
6454 allowing one more file than it should have been.
6456 9. For compatibility with Sendmail, the command line option
6464 and sets the incoming protocol and host name (for trusted callers). The
6465 host name and its colon can be omitted when only the protocol is to be set.
6466 Note the Exim already has two private options, -pd and -ps, that refer to
6467 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6468 "s", but I don't think that's a major issue.
6470 10. A number of refactoring changes to the code, none of which should affect
6473 (a) The number of logging options was getting close to filling up the
6474 32-bit word that was used as a bit map. I have split them into two classes:
6475 those that are passed in the argument to log_write(), and those that are
6476 only ever tested independently outside of that function. These are now in
6477 separate 32-bit words, so there is plenty of room for expansion again.
6478 There is no change in the user interface or the logging behaviour.
6480 (b) When building, for example, log lines, the code previously used a
6481 macro that called string_cat() twice, in order to add two strings. This is
6482 not really sufficiently general. Furthermore, there was one instance where
6483 it was actually wrong because one of the argument was used twice, and in
6484 one call a function was used. (As it happened, calling the function twice
6485 did not affect the overall behaviour.) The macro has been replaced by a
6486 function that can join an arbitrary number of extra strings onto a growing
6489 (c) The code for expansion conditions now uses a table and a binary chop
6490 instead of a serial search (which was left over from when there were very
6491 few conditions). Also, it now recognizes conditions like "pam" even when
6492 the relevant support is not compiled in: a suitably worded error message is
6493 given if an attempt is made to use such a condition.
6495 11. Added ${time_interval:xxxxx}.
6497 12. A bug was causing one of the ddress fields not to be passed back correctly
6498 from remote delivery subprocesses. The field in question was not being
6499 subsequently used, so this caused to problems in practice.
6501 13. Added new log selectors queue_time and deliver_time.
6503 14. Might have fixed a bug in maildirsizefile handling that threw up
6504 "unexpected character" debug warnings, and recalculated the data
6505 unnecessarily. In any case, I expanded the warning message to give more
6508 15. Added the message "Restricted characters in address" to the statements in
6509 the default ACL that block characters like @ and % in local parts.
6511 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6512 Three changes have been made:
6514 (a) There was a serious bug; a negative response to MAIL caused the whole
6515 recipient domain to be cached as invalid, thereby blocking all messages
6516 to all local parts at the same domain, from all senders. This bug has
6517 been fixed. The domain is no longer cached after a negative response to
6518 MAIL if the sender used is not empty.
6520 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6523 (c) A new callout option, "use_sender" has been added for people who want
6524 the modified behaviour.
6530 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6533 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6534 indeed breaks things for older releases.
6536 3. Added additional logging to the case where there is a problem reading data
6537 from a filter that is running in a subprocess using a pipe, in order to
6538 try to track down a specific problem.
6540 4. Testing facility fudge: when running in the test harness and attempting
6541 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6542 getting "No route to host". Convert this to a timeout.
6544 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6547 6. Some OS don't have socklen_t but use size_t instead. This affects the
6548 fifth argument of getsockopt() amongst other things. This is now
6549 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6550 can be set for individual OS. I have set it for SunOS5, OSF1, and
6551 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6552 some earlier ones do not.
6554 7. Change 4.30/15 was not doing the test caselessly.
6556 8. The standard form for an IPv6 address literal was being rejected by address
6557 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6558 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6559 this, as well as the form without the "IPv6" on the front (but only when
6560 address literals are enabled, of course).
6562 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6564 10. Exim crashed if a message with an empty sender address specified by -f
6565 encountered a router with an errors_to setting. This could be provoked only
6566 by a command such as
6570 where an empty string was supplied; "<>" did not hit this bug.
6572 11. Installed PCRE release 4.5.
6574 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6575 remained set. It is now erased.
6577 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6578 times from message ids (which are base 36 rather than the normal 62).
6580 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6581 were being counted as actual protocol errors, and logged if the log
6582 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6583 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6584 and DATA following a set of rejected RCPTs do not count as protocol errors.
6585 In other words, Exim assumes they were pipelined, though this may not
6586 actually be the case. Of course, in all cases the client gets an
6587 appropriate error code.
6589 15. If a lookup fails in an ACL condition, a message about the failure may
6590 be available; it is used if testing the ACL cannot continue, because most
6591 such messages specify what the cause of the deferral is. However, some
6592 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6593 that caused an old message to be retained and used if a later statement
6594 caused a defer, replacing the real cause of the deferral.
6596 16. If an IP address had so many PTR records that the DNS lookup buffer
6597 was not large enough to hold them, Exim could crash while trying to process
6598 the truncated data. It now detects and logs this case.
6600 17. Further to 4.21/58, another change has been made: if (and only if) the
6601 first line of a message (the first header line) ends with CRLF, a bare LF
6602 in a subsequent header line has a space inserted after it, so as not to
6603 terminate the header.
6605 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6606 unnecessarily, used atoi() instead of strtol(), and didn't check the
6607 termination when getting file sizes from file names by regex.
6609 19. Completely re-implemented the support for maildirsize files, in the light
6610 of a number of problems with the previous contributed implementation
6611 (4.30/29). In particular:
6613 . If the quota is zero, the maildirsize file is maintained, but no quota is
6616 . If the maildir directory does not exist, it is created before any attempt
6617 to write a maildirsize file.
6619 . The quota value in the file is just a cache; if the quota is changed in
6620 the transport, the new value overrides.
6622 . A regular expression is available for excluding directories from the
6625 20. The autoreply transport checks the characters in options that define the
6626 message's headers; it allows continued headers, but it was checking with
6627 isspace() after an embedded newline instead of explicitly looking for a
6630 21. If all the "regular" hosts to which an address was routed had passed their
6631 expiry times, and had not reached their retry times, the address was
6632 bounced, even if fallback hosts were defined. Now Exim should go on to try
6635 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6636 equivalent code in the SMTP transport. Some hosts send humungous responses
6637 to HELO/EHLO, more than 1024 it seems.
6639 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6640 gives compiler warnings in some environments. I've now done it "properly",
6643 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6644 (because of problems with the built-in one) was declared to return uschar *
6645 instead of char *, causing compiler failure.
6647 25. Fixed a file descriptor leak when processing alias/forward files.
6649 26. Fixed a minor format string issue in dbfn.c.
6651 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6653 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6654 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6657 29. When a sender address is verified, it is cached, to save repeating the test
6658 when there is more than one recipient in a message. However, when the
6659 verification involves a callout, it is possible for different callout
6660 options to be set for different recipients. It is too complicated to keep
6661 track of this in the cache, so now Exim always runs a verification when a
6662 callout is required, relying on the callout cache for the optimization.
6663 The overhead is duplication of the address routing, but this should not be
6666 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6667 to be verified with callout=postmaster, and the main callout worked but the
6668 postmaster check failed, the verification correctly failed. However, if a
6669 subsequent RCPT command asked for sender verification *without* the
6670 postmaster check, incorrect caching caused this verification also to fail,
6673 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6674 it was not caching the DNS options (qualify_single, search_parents) that
6675 were used when the lookup failed. A subsequent lookup with different
6676 options therefore always gave the same answer, though there were cases
6677 where it should not have. (Example: a "domains = !$mx_any" option on a
6678 dnslookup router: the "domains" option is always processed without any
6679 widening, but the router might have qualify_single set.) Now Exim uses the
6680 cached value only when the same options are set.
6682 32. Added John Jetmore's "exipick" utility to the distribution.
6684 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6685 than a timeout (e.g. a certificate is required, and is not provided), an
6686 Exim server now closes the connection immediately. Previously it waited for
6687 the client to close - but if the client is SSL, it seems that they each
6688 wait for each other, leading to a delay before one of them times out.
6690 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6691 maintained 0.8.x compatibility because I don't think many are using it, and
6692 it is clearly obsolete.
6694 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6697 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6698 client certificate was expired. A simple patch fixes this, though I don't
6699 understand the full logic of why the verify callback is called multiple
6702 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6703 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6704 which causes problems with some clients (such as the Certicom SSL Plus
6705 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6706 disables the coutermeasure allowing Eudora to connect."
6708 38. Exim was not checking that a write() to a log file succeeded. This could
6709 lead to Bad Things if a log got too big, in particular if it hit a file
6710 size limit. Exim now panics and dies if it cannot write to a log file, just
6711 as it does if it cannot open a log file.
6713 39. Modified OS/Makefile-Linux so that it now contains
6715 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6717 The two -D definitions ensure that Exim is compiled with large file
6718 support, which makes it possible to handle log files that are bigger than
6721 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6722 instance) a domain was checked against a named list that involved a lookup,
6723 causing $domain_data to be set, then another domain was checked against the
6724 same list, then the first domain was re-checked, the value of $domain_data
6725 after the final check could be wrong. In particular, if the second check
6726 failed, it could be set empty. This bug probably also applied to
6729 41. The strip_trailing_dot option was not being applied to the address given
6730 with the -f command-line option.
6732 42. The code for reading a message's header from the spool was incrementing
6733 $received_count, but never initializing it. This meant that the value was
6734 incorrect (doubled) while delivering a message in the same process in which
6735 it was received. In the most common configuration of Exim, this never
6736 happens - a fresh exec is done - but it can happen when
6737 deliver_drop_privilege is set.
6739 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6740 it now includes up to 150 characters of the unexpected data in the log
6743 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6744 and building data strings. The size of both of these buffers was 10 000
6745 bytes - far larger than anybody would *ever* want, thought I. Needless to
6746 say, somebody hit the limit. I have increased the maximum line length to
6747 20 000 and the maximum data length of concatenated lines to 100 000. I have
6748 also fixed two bugs, because there was no checking on these buffers. Tsk,
6749 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6750 buffer is too small.
6752 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6753 lsearch lookups. Now it does.
6755 46. When parsing a route_list item in a manualroute router, a fixed-length
6756 buffer was used for the list of hosts. I made this 1024 bytes long,
6757 thinking that nobody would ever have a list of hosts that long. Wrong.
6758 Somebody had a whole pile of complicated expansion conditions, and the
6759 string was silently truncated, leading to an expansion error. It turns out
6760 that it is easier to change to an unlimited length (owing to other changes
6761 that have happened since this code was originally written) than to build
6762 structure for giving a limitation error. The length of the item that
6763 expands into the list of hosts is now unlimited.
6765 47. The lsearch lookup could not handle data where the length of text line was
6766 more than 4095 characters. Such lines were truncated, leading to shortened
6767 data being returned. It should now handle lines of any length.
6769 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6770 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6773 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6774 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6775 no longer happens in comment lines. A list of replacements is now placed
6776 at the head of all of the source files, except those whose only change is
6777 to replace PERL_COMMAND in the very first #! line.
6779 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6780 messages on the queue, with a bottom-up merge sort, using code contributed
6781 by Michael Haardt. This should make operations like -bp somewhat faster on
6782 large queues. It won't affect queue runners, except when queue_run_in_order
6785 51. Installed eximstats 1.31 in the distribution.
6787 52. Added support for SRV lookups to the dnslookup router.
6789 53. If an ACL referred to $message_body or $message_body_end, the value was not
6790 reset for any messages that followed in the same SMTP session.
6792 54. The store-handling optimization for building very long strings was not
6793 differentiating between the different store pools. I don't think this
6794 actually made any difference in practice, but I've tidied it.
6796 55. While running the routers to verify a sender address, $sender_address
6797 was still set to the sender address. This is wrong, because when routing to
6798 send a bounce to the sender, it would be empty. Therefore, I have changed
6799 it so that, while verifying a sender address, $sender_address is set to <>.
6800 (There is no change to what happens when verifying a recipient address.)
6802 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6803 target A or AAAA records (if not already returned) without resetting the
6804 qualify_single or search_parents options of the DNS resolver. These are
6805 inappropriate in this case because the targets of MX and SRV records must
6806 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6807 target that, when qualified, matched something in the local domain. These
6808 two options are now turned off when doing these lookups.
6810 57. It seems that at least some releases of Reiserfs (which does not have the
6811 concept of a fixed number of inodes) returns zero and not -1 for the
6812 number of available inodes. This interacted badly with check_spool_inodes,
6813 which assumed that -1 was the "no such thing" setting. What I have done is
6814 to check that the total number of inodes is greater than zero before doing
6815 the test of how many are available.
6817 58. When a "warn" ACL statement has a log_message modifier, the message is
6818 remembered, and not repeated. This is to avoid a lot of repetition when a
6819 message has many recipients that cause the same warning to be written.
6820 However, Exim was preserving the list of already written lines for an
6821 entire SMTP session, which doesn't seem right. The memory is now reset if a
6822 new message is started.
6824 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6825 debugging output unless log_rewrite was also set.
6827 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6828 of (int)(handle) when we know that handle contains (void *)(-1).
6830 61. The Exim daemon panic-logs an error return when it closes the incoming
6831 connection. However "connection reset by peer" seems to be common, and
6832 isn't really an error worthy of noting specially, so that particular error
6835 62. When Exim is trying to find all the local interfaces, it used to panic and
6836 die if the ioctl to get the interface flags failed. However, it seems that
6837 on at least one OS (Solaris 9) it is possible to have an interface that is
6838 included in the list of interfaces, but for which you get a failure error
6839 for this call. This happens when the interface is not "plumbed" into a
6840 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6841 failure of the "get flags" call assumes that the interface is down.
6843 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6844 makes life easier for people who are doing arithmetic on fields extracted
6845 from dates, where you often get leading zeros that should not be
6846 interpreted as octal.
6848 64. Added qualify_domain to the redirect router, to override the global
6851 65. If a pathologically long header line contained very many addresses (the
6852 report of this problem mentioned 10 000) and each of them was rewritten,
6853 Exim could use up a very large amount of memory. (It kept on making new
6854 copies of the header line as it rewrote, and never released the old ones.)
6855 At the expense of a bit more processing, the header rewriting function has
6856 been changed so that it no longer eats memory in this way.
6858 66. The generation of the Received: header has been moved from the time that a
6859 message starts to be received, to the time that it finishes. The timestamp
6860 in the Received: header should now be very close to that of the <= log
6861 line. There are two side-effects of this change:
6863 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6864 logged header lines no longer include the local Received: line, because
6865 it has not yet been created. The same applies to a copy of the message
6866 that is returned to a non-SMTP sender when a message is rejected.
6868 (b) When a filter file is tested using -bf, no additional Received: header
6869 is added to the test message. After some thought, I decided that this
6872 This change does not affect the value of $received_for. It is still set
6873 after address rewriting, but before local_scan() is called.
6875 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6877 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6878 gave an unhelpful panic error message, and a defer error. I have managed to
6879 change this behaviour so that it now rejects any supplied certificate,
6880 which seems right, as the list of acceptable certificates is empty.
6882 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6883 gave an unhelpful defer error. I have not managed to make this reject any
6884 supplied certificates, but the error message it gives is "no certificate
6885 supplied", which is not helpful.
6887 70. exigrep's output now also includes lines that are not associated with any
6888 message, but which match the given pattern. Implemented by a patch from
6889 Martin Sluka, which also tidied up the Perl a bit.
6891 71. Recipient callout verification, like sender verification, was using <> in
6892 the MAIL FROM command. This isn't really the right thing, since the actual
6893 sender may affect whether the remote host accepts the recipient or not. I
6894 have changed it to use the actual sender in the callout; this means that
6895 the cache record is now keyed on a recipient/sender pair, not just the
6896 recipient address. There doesn't seem to be a real danger of callout loops,
6897 since a callout by the remote host to check the sender would use <>.
6898 [SEE ABOVE: changed after hitting problems.]
6900 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6901 temporary errors. However, in the case of such a code being given after
6902 the end of a data transmission (i.e. after ".") Exim was failing to write
6903 a retry record for the message. (Yes, there was some broken host that was
6904 actually sending 8xx at this point.)
6906 73. An unknown lookup type in a host list could cause Exim to panic-die when
6907 the list was checked. (An example that provoked this was putting <; in the
6908 middle of a list instead of at the start.) If this happened during a DATA
6909 ACL check, a -D file could be left lying around. This kind of configuration
6910 error no longer causes Exim to die; instead it causes a defer error. The
6911 incident is still logged to the main and panic logs.
6913 74. Buglet left over from Exim 3 conversion. The message "too many messages
6914 in one connection" was written to the rejectlog but not the mainlog, except
6915 when address rewriting (yes!) was being logged.
6917 75. Added write_rejectlog option.
6919 76. When a system filter was run not as root (that is, when system_filter_user
6920 was set), the values of the $n variables were not being returned to the
6921 main process; thus, they were not subsequently available in the $sn
6924 77. Added +return_path_on_delivery log selector.
6926 78. A connection timeout was being treated differently from recipients deferred
6927 when testing hosts_max_try with a message that was older than the host's
6928 retry timeout. (The host should not be counted, thus allowing all hosts to
6929 be tried at least once before bouncing.) This may have been the cause of an
6930 occasionally reported bug whereby a message would remain on the queue
6931 longer than the retry timeout, but would be bounced if a delivery was
6932 forced. I say "may" because I never totally pinned down the problem;
6933 setting up timeout/retry tests is difficult. See also the next item.
6935 79. The ultimate address timeout was not being applied to errors that involved
6936 a combination of host plus message (for example, a timeout on a MAIL
6937 command). When an address resolved to a number of possible hosts, and they
6938 were not all tried for each delivery (e.g. because of hosts_max_try), a
6939 message could remain on the queue longer than the retry timeout.
6941 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6944 81. Fixed an obscure SMTP outgoing bug which required at least the following
6945 conditions: (a) there was another message waiting for the same server;
6946 (b) the server returned 5xx to all RCPT commands in the first message so
6947 that the message was not completed; (c) the server dropped the connection
6948 or gave a negative response to the RSET that Exim sends to abort the
6949 transaction. The observed case was a dropped connection after DATA that had
6950 been sent in pipelining mode. That is, the server had advertised PIPELINING
6951 but was not implementing it correctly. The effect of the bug was incorrect
6952 behaviour, such as trying another host, and this could lead to a crash.
6958 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6959 and daemon.c were passed as pointers to ints; they should have been
6960 pointers to socklen_t variables (which are typically unsigned ints).
6962 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6965 3. Fixed a really odd bug that affected only the testing scheme; patching a
6966 certain fixed string in the binary changed the value of another string that
6967 happened to be identical to the end of the original first string.
6969 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6970 name", it returns that address as the IP address. On some operating
6971 systems (e.g. Solaris), it also passes back the IP address string as the
6972 "host name". However, on others (e.g. Linux), it passes back an empty
6973 string. Exim wasn't checking for this, and was changing the host name to an
6974 empty string, assuming it had been canonicalized.
6976 5. Although rare, it is permitted to have more than one PTR record for a given
6977 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6978 all the names associated with an address, because they do in Solaris.
6979 However, it seems that they do not in Linux for data that comes from the
6980 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6981 I found this out when I moved to a new Linux workstation and tried to run
6982 the Exim test suite.
6984 To get round this problem I have changed the code so that it now does its
6985 own call to the DNS to look up PTR records when searching for a host name.
6986 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6987 addresses that are only in /etc/hosts are still found.
6989 This behaviour is, however, controlled by an option called host_lookup_
6990 order, which defaults to "bydns:byaddr". If people want to use the other
6991 order, or indeed, just use one or the other means of lookup, they can
6992 specify it in this variable.
6994 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6995 some operating systems, this comes back from gethostbyaddr() as an empty
6996 string, and this is what Exim used to test for. However, it seems that in
6997 other systems, "." is yielded. Exim now tests for this case too.
6999 7. The values of check_spool_space and check_log_space are now held internally
7000 as a number of kilobytes instead of an absolute number of bytes. If a
7001 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7002 kilobyte. This means that much larger values can be stored.
7004 8. Exim monitor: an attempt to get the action menu when not actually pointing
7005 at a message produces an empty menu entitled "No message selected". This
7006 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7007 no entries in it ("Shell widget menu has zero width and/or height"). So I
7008 have added a single, blank menu entry in this case.
7010 9. Added ${quote_local_part.
7012 10. MIME decoding is now applied to the contents of Subject: header lines when
7015 11. Now that a reference to $sender_host_address automatically causes a reverse
7016 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7017 host lookup before query-style lookups in lists that might use this
7018 variable. This has therefore been abolished, and the "net-" prefix is no
7019 longer necessary for query-style lookups.
7021 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7022 have been a typo for LFLAGS, so it has been changed.
7024 13. The install script calls Exim with "-C /dev/null" in order to find the
7025 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7026 to be output. However, since Exim outputs its version number before the
7027 error, it didn't break the script. It just looked ugly. I fixed this by
7028 always allowing "-C /dev/null" if the caller is root.
7030 14. Ignore overlarge ACL variable number when reading spool file - insurance
7031 against a later release with more variables having written the file.
7033 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7034 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7035 this, as well as the form without the "IPv6" on the front.
7037 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7038 OS/Makefile-Darwin file.
7040 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7041 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7043 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7044 variable. However, some OS use ints for inodes. I've added cast to long int
7045 to get rid of the compiler warning.
7047 19. I had forgotten to lock out "/../" in configuration file names when
7048 ALT_CONFIG_PREFIX was set.
7050 20. Routers used for verification do not need to specify transports. However,
7051 if such a router generated a host list, and callout was configured, Exim
7052 crashed, because it could not find a port number from the (non-existent)
7053 transport. It now assumes port 25 in this circumstance.
7055 21. Added the -t option to exigrep.
7057 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7058 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7059 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7060 them is set, LOOKUP_LSEARCH is forced.
7062 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7064 24. Added sender and host information to the "rejected by local_scan()" log
7065 line; previously there was no indication of these.
7067 25. Added .include_if_exists.
7069 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7070 Linux. It turns out that not all file systems support this. Apparently some
7071 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7072 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7073 not supported on the file descriptor, is now ignored when Exim is trying to
7074 sync a directory. This applies only to Linux.
7076 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7078 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7079 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7082 29. Added support for maildirsize files from supplied patch (modified a bit).
7084 30. The use of :fail: followed by an empty string could lead Exim to respond to
7085 sender verification failures with (e.g.):
7087 550 Verification failed for <xxx>
7088 550 Sender verify failed
7090 where the first response line was missing the '-' that indicates it is not
7091 the final line of the response.
7093 31. The loop for finding the name of the user that called Exim had a hardwired
7094 limit of 10; it now uses the value of finduser_retries, which is used for
7095 all other user lookups.
7097 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7100 33. Exim was neglecting to zero errno before one call of strtol() when
7101 expanding a string and expecting an integer value. On some systems this
7102 resulted in spurious "integer overflow" errors. Also, it was casting the
7103 result into an int without checking.
7105 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7106 did not work. The code looks as if it has *never* worked, though it appears
7107 to have been documented since at least release 1.62. I have made it work.
7109 35. The "timeout_DNS" error in retry rules, also documented since at least
7110 1.62, also never worked. As it isn't clear exactly what this means, and
7111 clearly it isn't a major issue, I have abolished the feature by treating it
7112 as "timeout", and writing a warning to the main and panic logs.
7114 36. The display of retry rules for -brt wasn't always showing the error code
7117 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7118 timeout_connect_A, timeout_connect_MX.
7120 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7121 to the empty sender.
7123 39. The daemon was not analysing the content of -oX till after it had closed
7124 stderr and disconnected from the controlling terminal. This meant that any
7125 syntax errors were only noted on the panic log, and the return code from
7126 the command was 0. By re-arranging the code a little, I've made the
7127 decoding happen first, so such errors now appear on stderr, and the return
7128 code is 1. However, the actual setting up of the sockets still happens in
7129 the disconnected process, so errors there are still only recorded on the
7132 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7133 connections (as happens on some IP stacks) was logged at start up time as
7134 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7135 it from "IPv6 and IPv4", which means that two separate sockets are being
7138 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7139 says whether AF_INET or AF_INET6 was passed as an argument.
7141 42. Exiwhat output was messed up when time zones were included in log
7144 43. Exiwhat now gives more information about the daemon's listening ports,
7145 and whether -tls-on-connect was used.
7147 44. The "port" option of the smtp transport is now expanded.
7149 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7150 silently ignored. Now an error message is written to the main and panic
7153 46. There's a new ACL modifier called "logwrite" which writes to a log file
7154 as soon as it is encountered.
7156 47. Added $local_user_uid and $local_user_gid at routing time.
7158 48. Exim crashed when trying to verify a sender address that was being
7161 49. Exim was recognizing only a space character after ".include". It now also
7162 recognizes a tab character.
7164 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7165 extracting the relevant information from the specification. The man page no
7166 longer contains scrambled data for the -d option, and I've added a section
7167 at the front about calling Exim under different names.
7169 51. Added "extra_headers" argument to the "mail" command in filter files.
7171 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7174 53. Installed eximstats 1.29.
7176 54. Added transport_filter_timeout as a generic transport option.
7178 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7179 Cc: header lines. This was required by RFC 822, but it not required by RFC
7182 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7183 incoming messages that did not have them. Now it does so only if the
7184 message originates locally, that is, if there is no associated remote host
7185 address. When Resent- header lines are present, this applies to the Resent-
7186 lines rather than the non-Resent- lines.
7188 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7189 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7191 58. Messages for configuration errors now include the name of the main
7192 configuration file - useful now that there may be more than one file in a
7193 list (.included file names were always shown).
7195 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7196 for those rare installations that do not start the daemon as root or run it
7197 setuid root. I've cut out the call to initgroups() if the daemon is not
7200 60. The Exim user and group can now be bound into the binary as text strings
7201 that are looked up at the start of Exim's processing.
7203 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7205 62. Added $mailstore_basename variable.
7207 63. Installed patch to sieve.c from Michael Haardt.
7209 64. When Exim failed to open the panic log after failing to open the main log,
7210 the original message it was trying to log was written to stderr and debug
7211 output, but if they were not available (the usual case in production), it
7212 was lost. Now it is written to syslog before the two lines that record the
7213 failures to open the logs.
7215 65. Users' Exim filters run in subprocesses under the user's uid. It is
7216 possible for a "deliver" command or an alias in a "personal" command to
7217 provoke an address rewrite. If logging of address rewriting is configured,
7218 this fails because the process is not running as root or exim. There may be
7219 a better way of dealing with this, but for the moment (because 4.30 needs
7220 to be released), I have disabled address rewrite logging when running a
7221 filter in a non-root, non-exim process.
7227 1. The buildconfig auxiliary program wasn't quoting the value set for
7228 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7229 not defined. This bug was masked in 4.22 by the effect that was fixed in
7232 2. Some messages that were rejected after a message id was allocated were
7233 shown as "incomplete" by exigrep. It no longer does this for messages that
7234 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7236 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7237 have allow_domain_literals set, the ID did not get logged in the <= line.
7238 Domain literals are now always recognized in Message-ID: header lines.
7240 4. The first argument for a ${extract expansion item is the key name or field
7241 number. Leading and trailing spaces in this item were not being ignored,
7242 causing some misleading effects.
7244 5. When deliver_drop_privilege was set, single queue runner processes started
7245 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7246 same command in the process it spins off) were not dropping privilege.
7248 6. When the daemon running as "exim" started a queue runner, it always
7249 re-executed Exim in the spun-off process. This is a waste of effort when
7250 deliver_drop_privilege is set. The new process now just calls the
7251 queue-runner function directly.
7257 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7260 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7261 ignored. Though the use of -oP was forcing the writing of a pid file, it
7262 was always written to the default place.
7264 3. If the message "no IP address found for host xxxx" is generated during
7265 incoming verification, it is now followed by identification of the incoming
7266 connection (so you can more easily find what provoked it).
7268 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7270 5. Added some features to "harden" Exim a bit more against certain attacks:
7272 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7273 be put in Local/Makefile. This is like the never_users runtime option,
7274 but it cannot be overridden. The default setting is "root".
7276 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7277 prefix string with which any file named in a -C command line option
7280 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7281 is retained for -C and -D only if the caller of Exim is root. Without
7282 it, the exim user may also use -C and -D and retain privilege.
7284 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7285 command line option is disabled.
7287 6. Macro names set by the -D option must start with an upper case letter, just
7288 like macro names defined in the configuration file.
7290 7. Added "dereference=" facility to LDAP.
7292 8. Two instances of the typo "uknown" in the source files are fixed.
7294 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7295 the Configure-Makefile script screwed up while processing it.
7297 10. Incorporated PCRE 4.4.
7299 11. The SMTP synchronization check was not operating right at the start of an
7300 SMTP session. For example, it could not catch a HELO sent before the client
7301 waited for the greeting. There is now a check for outstanding input at the
7302 point when the greeting is written. Because of the duplex, asynchronous
7303 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7304 way, but not yet received, when the check is performed.
7306 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7307 on TCP/IP sockets, because this apparently causes some broken clients to
7310 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7311 unchanged) from the Cygwin maintainer.
7313 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7314 was supported instead of testing for mbx. Effectively a typo.
7316 15. The spa authenticator server code was not checking that the input it
7317 received was valid base64.
7319 16. The debug output line for the "set" modifier in ACLs was not showing the
7320 name of the variable that was being set.
7322 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7324 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7325 lookup on its own. Something else was needed to trigger the lookup. For
7326 example, a match in host_lookup or the need for a host name in a host list.
7327 Now, if $sender_host_name is referenced and the host name has not yet been
7328 looked up, a lookup is performed. If the lookup fails, the variable remains
7329 empty, and $host_lookup_failed is set to "1".
7331 19. Added "eqi" as a case-independent comparison operator.
7333 20. The saslauthd authentication condition could segfault if neither service
7334 nor realm was specified.
7336 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7337 error message that was logged was misleading, and incoming SMTP
7338 connections were dropped. The message is now more accurate, and temporary
7339 errors are given to SMTP connections.
7341 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7342 (except RSET) with a 421 error, until QUIT is received. However, it was
7343 failing to send a response to QUIT.
7345 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7346 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7347 ACL is now run after the helo_try_verify_hosts code.
7349 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7350 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7351 case-independent, so other case variants are also recognized). Apparently
7352 some systems use these upper case variants.
7354 25. If more than two messages were waiting for the same host, and a transport
7355 filter was specified for the transport, Exim sent two messages over the
7356 same TCP/IP connection, and then failed with "socket operation on non-
7357 socket" when it tried to send the third.
7359 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7361 27. The extern definition of crypt16() in expand.c was not being excluded when
7362 the OS had its own crypt16() function.
7364 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7365 as a preferred synonym for return_size_limit, both as an option and as an
7368 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7370 30. Changed the default configuration ACL to relax the local part checking rule
7371 for addresses that are not in any local domains. For these addresses,
7372 slashes and pipe symbols are allowed within local parts, but the sequence
7373 /../ is explicitly forbidden.
7375 31. SPA server authentication was not clearing the challenge buffer before
7378 32. log_message in a "warn" ACL statement was writing to the reject log as
7379 well as to the main log, which contradicts the documentation and doesn't
7380 seem right (because no rejection is happening). So I have stopped it.
7382 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7383 However, I am unable to do any testing of this.
7385 34. Fixed an infelicity in the appendfile transport. When checking directories
7386 for a mailbox, to see if any needed to be created, it was accidentally
7387 using path names with one or more superfluous leading slashes; tracing
7388 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7390 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7391 contents are added to the log line that is written for every discarded
7392 recipient. (Previously a log_message setting was ignored.)
7394 36. The ${quote: operator now quotes the string if it is empty.
7396 37. The install script runs exim in order to find its version number. If for
7397 some reason other than non-existence or emptiness, which it checks, it
7398 could not run './exim', it was installing it with an empty version number,
7399 i.e. as "exim-". This error state is now caught, and the installation is
7402 38. An argument was missing from the function that creates an error message
7403 when Exim fails to connect to the socket for saslauthd authentication.
7404 This could cause Exim to crash, or give a corrupted message.
7406 39. Added isip, isip4, and isip6 to ${if conditions.
7408 40. The ACL variables $acl_xx are now saved with the message, and can be
7409 accessed later in routers, transports, and filters.
7411 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7412 strings in the file are not string-expanded.
7414 42. If a MAIL command specified a SIZE value that was too large to fit into an
7415 int variable, the check against message_size_limit failed. Such values are
7416 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7417 day this will have to be increased, but I don't think I want to be around
7418 when emails are that large.
7425 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7426 iconv() is not standard in FreeBSD.
7428 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7429 IPv6 enabled. The observed symptom was a segmentation fault on return from
7430 the function os_common_find_running_interfaces() in src/os.c.
7432 3. In the check_special_case() function in daemon.c I had used "errno" as an
7433 argument name, which causes warnings on some systems. This was basically a
7434 typo, since it was named "eno" in the comments!
7436 4. The code that waits for the clock to tick (at a resolution of some fraction
7437 of a second) so as to ensure message-id uniqueness was always waiting for
7438 at least one whole tick, when it could have waited for less. [This is
7439 almost certainly not relevant at current processor speeds, where it is
7440 unlikely to ever wait at all. But we try to future-proof.]
7442 5. The function that sleeps for a time interval that includes fractions of a
7443 second contained a race. It did not block SIGALRM between setting the
7444 timer, and suspending (a couple of lines later). If the interval was short
7445 and the sigsuspend() was delayed until after it had expired, the suspension
7446 never ended. On busy systems this could lead to processes getting stuck for
7449 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7450 process, before it forks any delivery processes. The open lookup caching
7451 mechanism meant that the open file or database connection was passed into
7452 the delivery process. The problem was that delivery processes always tidy
7453 up cached lookup data. This could cause a problem for the next delivery
7454 process started by the queue runner, because the external queue runner
7455 process does not know about the closure. So the next delivery process
7456 still has data in the lookup cache. In the case of a file lookup, there was
7457 no problem because closing a file descriptor in a subprocess doesn't affect
7458 the parent. However, if the lookup was caching a connection to a database,
7459 the connection was closed, and the second delivery process was likely to
7460 see errors such as "PGSQL: query failed: server closed the connection
7461 unexpectedly". The problem has been fixed by closing all cached lookups
7462 in a queue runner before running a delivery process.
7464 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7465 seem to have the "const" qualifier which it has on other OS. I've
7468 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7469 *of the same type* (client or server) with the same public name that an
7470 error should be diagnosed.
7472 9. When Exim looked up a host name for an IP address, but failed to find the
7473 original IP address when looking up the host name (a safety check), it
7474 output the message "<ip address> does not match any IP for NULL", which was
7475 confusing, to say the least. The bug was that the host name should have
7476 appeared instead of "NULL".
7478 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7479 user that is built into the binary, and the -C or -D options is used, root
7480 privilege is dropped before the configuration file is read. In addition,
7481 logging is switched to stderr instead of the normal log files. If the
7482 configuration then re-defines the Exim user, the unprivileged environment
7483 is probably not what is expected, so Exim logs a panic warning message (but
7486 However, if deliver_drop_privilege is set, the unprivileged state may well
7487 be exactly what is intended, so the warning has been cut out in that case,
7488 and Exim is allowed to try to write to its normal log files.
7494 1. smtp_return_error_details was not giving details for temporary sender
7495 or receiver verification errors.
7497 2. Diagnose a configuration error if two authenticators have the same public
7500 3. Exim used not to create the message log file for a message until the first
7501 delivery attempt. This could be confusing when incoming messages were held
7502 for policy or load reasons. The message log file is now created at the time
7503 the message is received, and an initial "Received" line is written to it.
7505 4. The automatically generated man page for command line options had a minor
7506 bug that caused no ill effects; however, a more serious problem was that
7507 the procedure for building the man page automatically didn't always
7508 operate. Consequently, release 4.20 contains an out-of-date version. This
7509 shouldn't happen again.
7511 5. When building Exim with embedded Perl support, the script that builds the
7512 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7513 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7515 6. The freeze_tell option was not being used for messages that were frozen on
7516 arrival, either by an ACL or by local_scan().
7518 7. Added the smtp_incomplete_transaction log selector.
7520 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7521 was accepting AUTH without a new EHLO.
7523 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7524 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7527 10. Logging of TCP/IP connections (when configured) now happens in the main
7528 daemon process instead of the child process, so that the TCP/IP connection
7529 count is more accurate (but it can never be perfect).
7531 11. The use of "drop" in a nested ACL was not being handled correctly in the
7532 outer ACL. Now, if condition failure induced by the nested "drop" causes
7533 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7534 or "require"), the connection is dropped.
7536 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7537 that yield "discard" can now be used with an "accept" or a "discard" verb,
7538 but an error is generated for any others (because I can't see a useful way
7539 to define what should happen).
7541 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7542 are now processed in the same way as lines in the Exim configuration file.
7543 In particular, continuation lines are supported.
7545 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7547 15. Added -ti meaning -t -i.
7549 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7550 domains, and warn by logging if others are found.
7552 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7553 returned from GIFCONF when Exim is trying to find the list of interfaces on
7554 a host. The code in os.c has been modified to copy each ifreq to an aligned
7555 structure in all cases.
7557 Also, in some cases, the returned ifreq's were being copied to a 'struct
7558 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7559 means the last couple of bytes of an IPv6 address could be chopped if the
7560 ifreq contained only a normal sockaddr (14 bytes storage).
7562 18. Named domain lists were not supported in the hosts_treat_as_local option.
7563 An entry such as +xxxx was not recognized, and was treated as a literal
7566 19. Ensure that header lines added by a DATA ACL are included in the reject log
7567 if the ACL subsequently rejects the message.
7569 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7570 MD5 (which is deprecated).
7572 21. When testing a filter file using -bf, Exim was writing a message when it
7573 took the sender from a "From " line in the message, but it was not doing so
7574 when it took $return_path from a Return-Path: header line. It now does.
7576 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7577 with a valid header line field name (a series of printing characters
7578 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7580 23. Changed "disc" in the source to "disk" to conform to the documentation and
7581 the book and for uniformity.
7583 24. Ignore Sendmail's -Ooption=value command line item.
7585 25. When execve() failed while trying to run a command in a pipe transport,
7586 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7587 could be confused with a return value of 69 from the command itself. This
7588 has been changed to 127, the value the shell returns if it is asked to run
7589 a non-existent command. The wording for the related log line suggests a
7590 non-existent command as the problem.
7592 26. If received_header_text expands to an empty string, do not add a Received:
7593 header line to the message. (Well, it adds a token one on the spool, but
7594 marks it "old" so that it doesn't get used or transmitted.)
7596 27. Installed eximstats 1.28 (addition of -nt option).
7598 28. There was no check for failure on the call to getsockname() in the daemon
7599 code. This can fail if there is a shortage of resources on the system, with
7600 ENOMEM, for example. A temporary error is now given on failure.
7602 29. Contrary to the C standard, it seems that in some environments, the
7603 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7604 program. Exim now does this explicitly; it affects the formatting of
7605 timestamps using strftime().
7607 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7608 complaints. I've now initialized all the variables, to avoid this.
7610 32. Header lines added by a system filter were not being "seen" during
7611 transport-time rewrites.
7613 33. The info_callback() function passed to OpenSSL is set up with type void
7614 (*)(SSL *, int, int), as described somewhere. However, when calling the
7615 function (actually a macro) that sets it up, the type void(*)() is
7616 expected. I've put in a cast to prevent warnings from picky compilers.
7618 34. If a DNS black list lookup found a CNAME record, but there were no A
7619 records associated with the domain it pointed at, Exim crashed.
7621 35. If a DNS black list lookup returned more than one A record, Exim ignored
7622 all but the first. It now scans all returned addresses if a particular IP
7623 value is being sought. In this situation, the contents of the
7624 $dnslist_value variable are a list of all the addresses, separated by a
7627 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7628 to accept a host name and all its aliases if the forward lookup for any of
7629 them yielded the IP address of the incoming connection. Now it accepts only
7630 those names whose forward lookup yields the correct IP address. Any other
7631 names are discarded. This closes a loophole whereby a rogue DNS
7632 administrator could create reverse DNS records to break through a
7633 wildcarded host restriction in an ACL.
7635 37. If a user filter or a system filter that ran in a subprocess used any of
7636 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7637 the wrong values were passed to the pipe command ($thisaddress had the
7638 value of $0, $0 had the value of $1, etc). This bug was introduced by
7639 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7641 38. Improved the line breaking for long SMTP error messages from ACLs.
7642 Previously, if there was no break point between 40 and 75 characters, Exim
7643 left the rest of the message alone. Two changes have been made: (a) I've
7644 reduced the minimum length to 35 characters; (b) if it can't find a break
7645 point between 35 and 75 characters, it looks ahead and uses the first one
7646 that it finds. This may give the occasional overlong line, but at least the
7647 remaining text gets split now.
7649 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7650 file descriptors might be low, and that setting 1000 would always raise it.
7651 It turns out that in some environments, the limit is already over 1000 and
7652 that lowering it causes trouble. So now Exim takes care not to decrease it.
7654 40. When delivering a message, the value of $return_path is set to $sender_
7655 address at the start of routing (routers may change the value). By an
7656 oversight, this default was not being set up when an address was tested by
7657 -bt or -bv, which affected the outcome if any router or filter referred to
7660 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7661 log only when "message" or "log_message" are set. However, if one of the
7662 conditions was an address verification, or a call to a nested ACL, the
7663 messages generated by the underlying test were being passed through. This
7664 no longer happens. The underlying message is available in $acl_verify_
7665 message for both "message" and "log_message" expansions, so it can be
7666 passed through if needed.
7668 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7669 new expansion $bh_ to give the encoded byte string without charset
7670 translation. Translation happens only if iconv() is available; HAVE_ICONV
7671 indicates this at build time. HEADERS_CHARSET gives the charset to
7672 translate to; headers_charset can change it in the configuration, and
7673 "headers charset" can change it in an individual filter file.
7675 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7676 that creates RFC 2047 encoded "words" labels them as that charset instead
7677 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7678 expansion operator; (ii) when Exim creates a From: line for a local
7679 message; (iii) when a header line is rewritten to include a "phrase" part.
7681 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7682 buggy, causing it to skip the first lines of messages whose message ID
7683 ended in 'D'. This would not have bitten before Exim release 4.14, because
7684 message IDs were unlikely to end in 'D' before then. The effect was to have
7685 incorrect size information for certain domains.
7687 45. #include "config.h" was missing at the start of the crypt16.c module. This
7688 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7691 46. If there was a timeout during a "random" callout check, Exim treated it as
7692 a failure of the random address, and carried on sending RSET and the real
7693 address. If the delay was just some slowness somewhere, the response to the
7694 original RCPT would be taken as a response to RSET and so on, causing
7695 mayhem of various kinds.
7697 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7698 when I implemented it. It didn't allow for the fact that some option values
7699 may legitimately be negative (e.g. size_addition), and it didn't even do
7700 the right test for positive values.
7702 48. Domain names in DNS records are case-independent. Exim always looks them up
7703 in lower case. Some resolvers return domain names in exactly the case they
7704 appear in the zone file, that is, they may contain uppercase letters. Not
7705 all resolvers do this - some return always lower case. Exim was treating a
7706 change of case by a resolver as a change of domain, similar to a widening
7707 of a domain abbreviation. This triggered its re-routing code and so it was
7708 trying to route what was effectively the same domain again. This normally
7709 caused routing to fail (because the router wouldn't handle the domain
7710 twice). Now Exim checks for this case specially, and just changes the
7711 casing of the domain that it ultimately uses when it transmits the message
7714 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7717 50. If a filter generated a file delivery with a non-absolute name (possible if
7718 no home directory exists for the router), the forbid_file option was not
7721 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7722 the existing equality matching.
7724 52. Exim was using ints instead of ino_t variables in some places where it was
7725 dealing with inode numbers.
7727 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7728 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7729 called TMPDIR, and if it finds it is different, it changes its value.
7731 54. The smtp_printf() function is now made available to local_scan() so
7732 additional output lines can be written before returning. There is also an
7733 smtp_fflush() function to enable the detection of a dropped connection.
7734 The variables smtp_input and smtp_batched_input are exported to
7737 55. Changed the default runtime configuration: the message "Unknown user"
7738 has been removed from the ACL, and instead placed on the localuser router,
7739 using the cannot_route_message feature. This means that any verification
7740 failures that generate their own messages won't get overridden. Similarly,
7741 the "Unrouteable address" message that was in the ACL for unverifiable
7742 relay addresses has also been removed.
7744 56. Added hosts_avoid_esmtp to the smtp transport.
7746 57. The exicyclog script was not checking for the esoteric option
7747 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7748 will work only if exicyclog is run under the appropriate euid.
7750 58. Following a discussion on the list, the rules by which Exim recognises line
7751 endings on incoming messages have been changed. The -dropcr and drop_cr
7752 options are now no-ops, retained only for backwards compatibility. The
7753 following line terminators are recognized: LF CRLF CR. However, special
7754 processing applies to CR:
7756 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7757 nor a local message in the state where . is a terminator.
7759 (ii) If a bare CR is encountered in a header line, an extra space is added
7760 after the line terminator so as not to end the header. The reasoning
7761 behind this is that bare CRs in header lines are most likely either
7762 to be mistakes, or people trying to play silly games.
7764 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7765 was being incorrectly given as 18 bytes larger than it should have been.
7766 This is a VOB (very old bug).
7768 60. This may never have affected anything current, but just in case it has:
7769 When the local host is found other than at the start of a list of hosts,
7770 the local host, those with the same MX, and any that follow, are discarded.
7771 When the list in question was part of a longer list of hosts, the following
7772 hosts (not currently being processed) were also being discarded. This no
7773 longer happens. I'm not sure if this situation could ever has previously
7776 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7779 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7780 header. This is something that is recommended in a new Internet Draft, and
7781 is something that is documented as being done by Sendmail. There are two
7782 possible values. For messages generated by the autoreply transport, Exim
7785 Auto-Submitted: auto-replied
7787 whereas for all other generated messages (e.g. bounces) it adds
7789 Auto-Submitted: auto-generated
7791 63. The "personal" condition in filters now includes a test for the
7792 Auto-Submitted: header. If it contains the string "auto-" the message it
7793 not considered personal.
7795 64. Added rcpt_include_affixes as a generic transport option.
7797 65. Added queue_only_override (default true).
7799 66. Added the syslog_duplication option.
7801 67. If what should have been the first header line of a message consisted of
7802 a space followed by a colon, Exim was mis-interpreting it as a header line.
7803 It isn't of course - it is syntactically invalid and should therefore be
7804 treated as the start of the message body. The misbehaviour could have
7805 caused a number of strange effects, including loss of data in subsequent
7806 header lines, and spool format errors.
7808 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7809 client host had authenticated. This control can now be exercised by an ACL
7810 for more flexibility.
7812 69. By default, callouts do not happen when testing with -bh. There is now a
7813 variant, -bhc, which does actually run the callout code, including
7814 consulting and updating the callout cache.
7816 70. Added support for saslauthd authentication, courtesy of Alexander
7819 71. If statvfs() failed on the spool or log directories while checking their
7820 size for availability, Exim confusingly gave the error "space shortage".
7821 Furthermore, in debugging mode it crashed with a floating point exception.
7822 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7823 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7824 serious problem, Exim now writes to the main and panic logs when this
7825 happens, with details of the failure. It then refuses to accept the
7826 incoming message, giving the message "spool directory problem" or "log
7827 directory problem" with a 421 code for SMTP messages.
7829 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7830 0, 1, and 2 exist, because some OS complain for execs without them (see
7831 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7832 descriptors. However, the code omitted to check that the open succeeded,
7833 causing mysterious errors if for some reason the permissions on /dev/null
7834 got screwed. Now Exim writes a message to the main and panic logs, and
7835 bombs out if it can't open /dev/null.
7837 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7838 interact so that it is all more flexible. It is supposed to remain
7839 backwards compatible. Also added extra_local_interfaces.
7841 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7842 to bomb out with an assertion failure - to the client this appears as a
7843 connection drop. This problem occurs in the part of the code that was taken
7844 from the Samba project. Fortunately, the assertion is in a very simple
7845 function, so I have fixed this by reproducing the function inline in the
7846 one place where it is called, and arranging for authentication to fail
7847 instead of killing the process with assert().
7849 75. The SPA client code was not working when the server requested OEM rather
7850 than Unicode encoding.
7852 76. Added code to make require_files with a specific uid setting more usable in
7853 the case where statting the file as root fails - usually a non-root-mounted
7854 NFS file system. When this happens and the failure is EACCES, Exim now
7855 forks a subprocess and does the per-uid checking as the relevant uid.
7857 77. Added process_log_path.
7859 78. If log_file_path was not explicitly set, a setting of check_log_space or
7860 check_log_inodes was ignored.
7862 79. If a space check for the spool or log partitions fails, the incident is now
7863 logged. Of course, in the latter case the data may get lost...
7865 80. Added the %p formatting code to string_format() so that it can be used to
7866 print addresses in debug_print(). Adjusted all the address printing in the
7867 debugging in store.c to use %p rather than %d.
7869 81. There was a concern that a line of code in smtp_in.c could overflow a
7870 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7871 initially expressed, the concern was not well-founded, because trailing
7872 spaces are removed early. However, if the trailing spaces were followed by
7873 a NULL, they did not get removed, so the overflow was possible. Two fixes
7876 (a) I re-wrote the offending code in a cleaner fashion.
7877 (b) If an incoming SMTP command contains a NULL character, it is rejected
7880 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7881 runs initgroups(), so that if the Exim user is in any additional groups,
7882 they will be used during message reception.
7888 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob