1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
28 JH/01 Avoid costly startup code when not strictly needed. This reduces time
29 for some exim process initialisations. It does mean that the logging
30 of TLS configuration problems is only done for the daemon startup.
32 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
34 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
35 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
38 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
39 DKIM signing is being done. Previously a transport_filter would always
40 disable CHUNKING, falling back to traditional DATA.
42 JH/05 Regard command-line receipients as tainted.
44 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
46 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
47 PAM library frees one of the arguments given to it, despite the
48 documentation. Therefore a plain malloc must be used.
50 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
51 on-stack buffers were used, resulting in a taint trap when DSN information
52 copied from a received message was written into the buffer.
54 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
55 the ordering of its ARC headers. This caused a crash.
57 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
58 a new record was being constructed with information from the peer, a trap
61 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
62 installation would get error messages from DMARC verify, when it hit the
63 nonexistent file indicated by the default. Distros wanting DMARC enabled
64 should both provide the file and set the option.
65 Also enforce no DMARC verification for command-line sourced messages.
67 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
68 could, depending on the platform, hang at the STARTTLS response.
70 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
71 message on a connection. Previously if one message had ARC headers and
72 the following one did not, a crash could result when adding an
73 Authentication-Results: header.
75 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
76 Exim main code and Exim-related utities. The introduction of taint
77 tracking also did many adjustments to string handling. Since then, eximon
78 frequently terminated with an assert failure.
80 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
81 check for 452 responses. This slightly helps the inefficieny of doing
82 a large alias-expansion into a recipient-limited target. The max_rcpt
83 transport option still applies (and at the current default, will override
84 the new feature). The check is done for either cause of synch, and forces
85 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
86 connection. The new facility is not tunable at this time.
88 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
89 library live data was being used, so the results became garbage. Make
90 copies while it is still usable.
92 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
93 on delivery deferred (==) and failed (**) lines (if a delivery was
94 attemtped). Previously it was only on completion (=>) lines.
96 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
97 for the expansion of the server_scram_iter and server_scram_salt options.
99 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
100 are now specifically given a NO_DATA response without hitting the system
101 resolver. The library goes on to do the now-standard TXT lookup.
102 Use of dnsdb lookups is not affected.
104 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
105 only retrieve the errormessage once. Previously two calls to dlerror()
106 were used, and the second one (for mainlog/paniclog) retrieved null
109 JH/20 Taint checking: disallow use of tainted data for
110 - the appendfile transport file and directory options
111 - the pipe transport command
112 - the autoreply transport file, log and once options
113 - file names used by the redirect router (including filter files)
115 - paths used by single-key lookups
116 Previously this was permitted.
118 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
119 adjusted the size of a major service buffer; this failed because the
120 buffer was in use at the time. Change to a compile-time increase in the
121 buffer size, when this authenticator is compiled into exim.
123 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
124 previous fast-mode was untenable in the face of glibs using mmap to
125 support larger malloc requests.
127 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
128 New values supported, if defined on system where compiled:
129 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
130 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
132 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
133 running a limited number of proceses in parallel, a benefit is gained. The
134 amount varies with the platform hardware and load. The use of the option
135 queue_run_in_order means we cannot do this, as ordering becomes
138 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
139 had introduced a string-copy (for ensuring NUL-termination) which was not
140 appropriate for that case, which can include embedded NUL bytes in the
141 block of data. Investigation showed the copy to actually be needless, the
142 data being length-specified.
144 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
145 done during a receiving connection, and both used TLS, global info was
146 used rather than per-connection info for tracking the state of data
147 queued for transmission. This could result in a connection hang.
149 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
150 Previously, when delivering serveral messages down a single connection
151 only the first would provide a SIZE. This was due to the size information
152 not being properly tracked.
154 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
155 TAI (at 37 seconds currently), pretend to be in UTC for time-related
156 expansion and logging. Previously, spurious values such as a future
157 minute could be seen.
159 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
160 it could crash from a null-deref. This could also affect the
161 ${addresses: } operator and ${readsock } item.
163 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
164 message following a mime one, the variable was not reset.
166 JH/30 When an pipelined-connect fails at the first response, assume incorrect
167 cached capability (perhaps the peer reneged?) and immediately retry in
170 JH/31 Fix spurious detection of timeout while writing to transport filter.
172 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
173 an attempt to copy the string was made before checking it.
175 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
176 taint of the lookup key was maintained; we now regard the presence in the
177 filesystem as sufficient validation.
179 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
180 argument is supplied.
182 JH/35 The dsearch lookup now requires that the directory is an absolute path.
183 Previously this was not checked, and nonempty relative paths made an
184 access under Exim's current working directory.
186 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
187 Previously no event was raised.
189 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
190 parameter supplied by the sender MAIL FROM command. Previously it was
191 ignored, and only the check_spool_space option value for the required
194 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
195 the size of the signing public-key. Previously it was instead giving
196 the size of the signature hash.
198 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
199 the default. See the (new) dkim_verify_min_keysizes option.
201 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
202 and an ACL use a lookup for checking either the local_part or domain,
203 stale data could be accessed. Ensure that variable references are
204 dropped between messages.
206 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
207 by the client was not checked as pointing within response data before
208 being used. A malicious client could thus cause an out-of-bounds read and
209 possibly gain authentication. Fix by adding the check.
211 JH/42 Internationalisation: change the default for downconversion in the smtp
212 transport to be "if needed". Previously it was "as previously set" for
213 the message, which usually meant "if needed" for message-submission but
214 "no" for everything else. However, MTAs have been seen using SMTPUTF8
215 even when the envelope addresses did not need it, resulting in forwarding
216 failures to non-supporting MTAs. A downconvert in such cases will be
217 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
218 transport. The change does mean that addresses needing conversion will
219 be converted when previously a delivery failure would occur.
221 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
222 response was received it would be used unchecked in a fail-DSN, violating
223 standards on line-length limits. Truncate if needed.
225 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
226 doesn't work. (Additionally add a new main config option to configure the
233 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
234 side, to decode a TLS 1.3 packet capture.
236 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
237 Previously the default library behaviour applied, sending two, each in
240 JH/03 Debug output for ACL now gives the config file name and line number for
243 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
245 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
247 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
248 buffer overrun for (non-chunking) other transports.
250 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
251 TLS1.3, means that a server rejecting a client certificate is not visible
252 to the client until the first read of encrypted data (typically the
253 response to EHLO). Add detection for that case and treat it as a failed
254 TLS connection attempt, so that the normal retry-in-clear can work (if
255 suitably configured).
257 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
258 and/or domain. Found and fixed by Jason Betts.
260 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
261 configuration). If a CNAME target was not a wellformed name pattern, a
264 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
265 the OS reports them interleaved with other addresses.
267 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
268 used both for input and for a verify callout, both encrypted, SMTP
269 responses being sent by the server could be lost. This resulted in
270 dropped connections and sometimes bounces generated by a peer sending
273 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
274 string. Previously it was possible to cause undefined behaviour in a
275 library routine (usually a crash). Found by "zerons".
277 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
280 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
281 API was removed, so update to use the newer ones.
283 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
284 any timeout set, is taking a long time. Previously we would hang on to a
285 rotated logfile "forever" if the input was arriving with long gaps
286 (a previous attempt to fix addressed lack, for a long time, of initial
289 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
290 shared (NFS) environment. The length of the tempfile name is now
291 4 + 16 ("hdr.$message_exim_id") which might break on file
292 systems which restrict the file name length to lower values.
295 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
296 shared (NFS) environment.
298 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
299 did for all versions <4.90). Notably -M, -m, --invert, -I may be
302 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
303 on some platforms for bit 31.
305 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
306 to changes apparently associated with TLS1.3 handling some of the APIs
307 previously used were either nonfunctional or inappropriate. Strings
308 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
309 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
310 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
311 This affects log line X= elements, the $tls_{in,out}_cipher variables,
312 and the use of specific cipher names in the encrypted= ACL condition.
314 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
316 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
317 verification result was not updated unless hosts_require_ocsp applied.
319 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
320 queue_list_requires_admin set to false, non-admin users were denied the
323 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
324 directory-of-certs mode. Previously they were advertised despite the
327 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
328 A single TCP connection by a client will now hold a TLS connection open
329 for multiple message deliveries, by default. Previoud the default was to
332 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
333 default. If built with the facility, DANE will be used. The facility
334 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
336 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
337 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
338 must be defined and you must still, unless you define DISABLE_TLS, manage
339 the the include-dir and library-file requirements that go with that
340 choice. Non-TLS builds are still supported.
342 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
345 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
346 default. If the platform supports and has the facility enabled, it will
347 be requested on all coneections.
349 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
350 controlled by the build-time option SUPPORT_PIPE_CONNECT.
352 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
354 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
355 success-DSN messages. Previously the From: header was always the default
356 one for these; the option was ignored.
358 JH/28 Fix the timeout on smtp response to apply to the whole response.
359 Previously it was reset for every read, so a teergrubing peer sending
360 single bytes within the time limit could extend the connection for a
361 long time. Credit to Qualsys Security Advisory Team for the discovery.
363 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
364 delivery address, which leaked information of the results of local
365 forwarding. Change to the original envelope recipient address, per
368 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
369 requested. Previously not bounce was generated and a log entry of
370 error ignored was made.
372 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
374 JH/32 Introduce a general tainting mechanism for values read from the input
375 channel, and values derived from them. Refuse to expand any tainted
376 values, to catch one form of exploit.
378 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
379 was unused and the unexpanded text used for the test. Found and
380 fixed by Ruben Jenster.
382 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
383 an attempt to use a TLS library read routine dereffed a nul pointer,
386 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
387 them in our smtp response.
389 JH/36 Have the general router option retry_use_local_part default to true when
390 any of the restrictive preconditions are set (to anything). Previously it
391 was only for check_local user. The change removes one item of manual
392 configuration which is required for proper retries when a remote router
393 handles a subset of addresses for a domain.
395 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
396 link count into consideration.
398 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
399 caused the extension of big_buffer, the following lines were ignored.
401 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
402 accordance with RFC 2308. Previously there was no expiry, so a longlived
403 receive process (eg. due to ACL delays) versus a short SOA value could
406 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
408 JH/39 Promote DMARC support to mainline.
410 JH/40 Bug 2452: Add a References: header to DSNs.
412 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
413 parameters. The relevant library call is documented as "Deprecated: This
414 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
415 3.6.0, DH parameters are negotiated following RFC7919."
417 HS/06 Change the default of dnssec_request_domains to "*"
419 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
420 carried on and emitted a BDAT command, even when PIPELINING was not
423 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
424 buffer was used for the filename, resulting in a trap when tainted
425 arguments (eg. $domain) were used.
427 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
428 recommended to avoid a possible server-load attack. The feature can be
429 re-enabled via the openssl_options main cofiguration option.
431 JH/45 local_scan API: documented the current smtp_printf() call. This changed
432 for version 4.90 - adding a "more data" boolean to the arguments.
433 Bumped the ABI version number also, this having been missed previously;
434 release versions 4.90 to 4.92.3 inclusive were effectively broken in
435 respect of usage of smtp_printf() by either local_scan code or libraries
436 accessed via the ${dlfunc } expansion item. Both will need coding
437 adjustment for any calls to smtp_printf() to match the new function
438 signature; a FALSE value for the new argument is always safe.
440 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
441 the file-offset (which the Linux syscall does, and exim expects); this
442 resulted in an indefinite loop.
444 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
445 to do ARC verification. The Authentication-Results: header line added
446 by the configuration then had no ARC item.
452 JH/01 Remove code calling the customisable local_scan function, unless a new
453 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
455 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
456 non-signal-safe functions being used.
458 JH/03 Bug 2269: When presented with a received message having a stupidly large
459 number of DKIM-Signature headers, disable DKIM verification to avoid
460 a resource-consumption attack. The limit is set at twenty.
462 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
463 report of oldest_pass in ${authres } in consequence, and separate out
464 some descriptions of reasons for verification fail.
466 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
467 files in the spool were present and unlocked. A queue-runner could spot
468 them, resulting in a duplicate delivery. Fix that by doing the unlock
469 after the unlink. Investigation by Tim Stewart. Take the opportunity to
470 add more error-checking on spoolfile handling while that code is being
473 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
474 No known attacks, no CVE, this is defensive hardening.
476 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
477 a queue-runner could start a delivery while other operations were ongoing.
478 Cutthrough delivery was a common victim, resulting in duplicate delivery.
479 Found and investigated by Tim Stewart. Fix by using the open message data
480 file handle rather than opening another, and not locally closing it (which
481 releases a lock) for that case, while creating the temporary .eml format
482 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
484 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
485 $sender_verify_failure/$recipient_verify_failure to "random".
487 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
490 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
491 Previously this would segfault.
493 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
496 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
497 like zero, since the resolver should be doing this for us, But we need one
498 as a CNAME but no MX presence gets the CNAME returned; we need to check
499 that doesn't point to an MX to declare it "no MX returned" rather than
500 "error, loop". A new main option is added so the older capability of
501 following some limited number of chain links is maintained.
503 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
505 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
506 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
507 now-deprecated earlier definitions used only specified the range up to TLS
508 1.2 (in the older-version library docs).
510 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
512 JH/15 Rework TLS client-side context management. Stop using a global, and
513 explicitly pass a context around. This enables future use of TLS for
514 connections to service-daemons (eg. malware scanning) while a client smtp
515 connection is using TLS; with cutthrough connections this is quite likely.
517 JH/16 Fix ARC verification to do AS checks in reverse order.
519 JH/17 Support a "tls" option on the ${readsocket } expansion item.
521 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
522 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
523 Previously the "utf8" would be re-prepended for every additional message.
525 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
526 Previously thery were accepted, resulting in issues when attempting to
527 forward messages to a non-supporting MTA.
529 PP/02 Let -n work with printing macros too, not just options.
531 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
532 one parent address was copied, and bogus data was used at delivery-logging
533 time. Either a crash (after delivery) or bogus log data could result.
534 Discovery and analysis by Tim Stewart.
536 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
537 Previously if the string ended mid-character, we did not insert the
538 promised '?' replacement.
540 PP/04 Documentation: current string operators work on bytes, not codepoints.
542 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
543 should pack well giving a smaller memory footprint so better caching and
544 therefore performance. Group the declarations where this can't be done so
545 that the byte-sized flag variables are not interspersed among pointer
546 variables, giving a better chance of good packing by the compiler.
548 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
549 non-null, to avoid issues with sites running BATV. Previously reports were
550 sent with an empty envelope sender so looked like bounces.
552 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
553 The ignore_error flag wasn't being returned from the filter subprocess so
554 was not set for later routers. Investigation and fix by Matthias Kurz.
556 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
557 and a msg:complete for the whole, when a message is manually removed using
558 -Mrm. Developement by Matthias Kurz, hacked on by JH.
560 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
561 a "Gnu special" function, asprintf() in the DB utility binary builds; I
562 hope that is portable enough.
564 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
565 requiring a known-CA anchor certificate; make it now rely entirely on the
566 TLSA as an anchor. Checking the name on the leaf cert against the name
567 on the A-record for the host is still done for TA (but not for EE mode).
569 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
570 included in delivery lines for non-proxied connections, when compiled with
571 SUPPORT_SOCKS and running with proxy logging enabled.
573 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
574 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
575 move the existing event to fire before the normal logging of message
576 failure so that custom logging is bracketed by normal logging.
578 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
579 msg:fail:internal event. Developement by Matthias Kurz.
581 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
582 far too small for todays use of crypto signatures stored there. Go all
583 the way to the max DNS message size of 64kB, even though this might be
584 overmuch for IOT constrained device use.
586 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
587 copy a string over itself. The library routine is documented as not
588 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
590 JH/32 For main options check_spool_space and check_inode_space, where the
591 platform supports 64b integers, support more than the previous 2^31 kB
592 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
593 the previous G, M, k.
595 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
596 $authenticated_fail_id variable on authentication failure. Previously
599 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
600 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
601 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
602 GNUTLS_SEC_PARAM_MEDIUM.
604 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
605 side. Previously we would continue as if no SNI had been received.
607 JH/36 Harden the handling of string-lists. When a list consisted of a sole
608 "<" character, which should be a list-separator specification, we walked
609 off past the nul-terimation.
611 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
612 causes) even when the retry time is not yet met. Previously they were
613 not, meaning that when (say) an account was over-quota and temp-rejecting,
614 and multiple senders' messages were queued, only one sender would get
615 notified on each configured delay_warning cycle.
617 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
619 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
620 cipher-suites, an error can be left on the stack even for a succeeding
621 accept; this results in impossible error messages when a later operation
624 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
625 return error codes indicating retry. Under TLS1.3 this becomes required.
627 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
628 it only wrote the new authenticators, resulting in a lack of tracking of
629 peer changes of ESMTP extensions until the next cache flush.
631 JH/41 Fix the loop reading a message header line to check for integer overflow,
632 and more-often against header_maxsize. Previously a crafted message could
633 induce a crash of the recive process; now the message is cleanly rejected.
635 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
636 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
642 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
643 When redis_servers is set to a list of > 1 element, and the Redis servers
644 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
645 case of MOVED into a DEFER case instead, thus moving the query onto the
646 next server in the list. For a cluster of N elements, all N servers must
647 be defined in redis_servers.
649 GF/02 Catch and remove uninitialized value warning in exiqsumm
650 Check for existence of @ARGV before looking at $ARGV[0]
652 JH/01 Replace the store_release() internal interface with store_newblock(),
653 which internalises the check required to safely use the old one, plus
654 the allocate and data copy operations duplicated in both (!) of the
655 extant use locations.
657 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
658 modifier. This matches the restriction on the commandline.
660 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
661 Previously only the last row was returned.
663 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
664 we assumed that tags in the header were well-formed, and parsed the
665 element content after inspecting only the first char of the tag.
666 Assumptions at that stage could crash the receive process on malformed
669 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
670 While running the DKIM ACL we operate on the Permanent memory pool so that
671 variables created with "set" persist to the DATA ACL. Also (at any time)
672 DNS lookups that fail create cache records using the Permanent pool. But
673 expansions release any allocations made on the current pool - so a dnsdb
674 lookup expansion done in the DKIM ACL releases the memory used for the
675 DNS negative-cache, and bad things result. Solution is to switch to the
676 Main pool for expansions.
677 While we're in that code, add checks on the DNS cache during store_reset,
678 active in the testsuite.
679 Problem spotted, and debugging aided, by Wolfgang Breyha.
681 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
682 When none of the hosts presented to a transport match an already-open
683 connection, close it and proceed with the list. Previously we would
684 queue the message. Spotted by Lena with Yahoo, probably involving
687 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
688 Previously a spurious "250 OK id=" response was appended to the proper
691 JH/08 The "support for" informational output now, which built with Content
692 Scanning support, has a line for the malware scanner interfaces compiled
693 in. Interface can be individually included or not at build time.
695 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
696 by the template makefile "src/EDITME". The "STREAM" support for an older
697 ClamAV interface method is removed.
699 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
700 rows affected is given instead).
702 JH/11 The runtime Berkeley DB library version is now additionally output by
703 "exim -d -bV". Previously only the compile-time version was shown.
705 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
706 SMTP connection. Previously, when one had more receipients than the
707 first, an abortive onward connection was made. Move to full support for
708 multiple onward connections in sequence, handling cutthrough connection
709 for all multi-message initiating connections.
711 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
712 routers. Previously, a multi-recipient message would fail to match the
713 onward-connection opened for the first recipient, and cause its closure.
715 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
716 a timeout on read on a GnuTLS initiating connection, resulting in the
717 initiating connection being dropped. This mattered most when the callout
718 was marked defer_ok. Fix to keep the two timeout-detection methods
721 JH/15 Relax results from ACL control request to enable cutthrough, in
722 unsupported situations, from error to silently (except under debug)
723 ignoring. This covers use with PRDR, frozen messages, queue-only and
726 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
728 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
729 metadata, resulting in a crash in free().
731 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
732 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
733 Broken also in d185889f4, with init system revamp.
735 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
736 Previously we abruptly closed the connection after reading a malware-
737 found indication; now we go on to read the "scan ok" response line,
740 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
741 ACL. Previously, a crash would result.
743 JH/19 Speed up macro lookups during configuration file read, by skipping non-
744 macro text after a replacement (previously it was only once per line) and
745 by skipping builtin macros when searching for an uppercase lead character.
747 JH/20 DANE support moved from Experimental to mainline. The Makefile control
748 for the build is renamed.
750 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
751 was allocated for every new TLS startup, meaning one per message. Fix
752 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
754 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
755 reported the original. Fix to report (as far as possible) the ACL
756 result replacing the original.
758 JH/23 Fix memory leak during multi-message connections using STARTTLS under
759 OpenSSL. Certificate information is loaded for every new TLS startup,
760 and the resources needed to be freed.
762 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
764 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
767 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
768 DATA response info to the (existing) per-recipient response info for
769 the "C=" log element. It can have useful tracking info from the
770 destination system. Patch from Simon Arlott.
772 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
773 length value. Previously this would segfault.
775 HS/02 Support Avast multiline protoocol, this allows passing flags to
776 newer versions of the scanner.
778 JH/28 Ensure that variables possibly set during message acceptance are marked
779 dead before release of memory in the daemon loop. This stops complaints
780 about them when the debug_store option is enabled. Discovered specifically
781 for sender_rate_period, but applies to a whole set of variables.
782 Do the same for the queue-runner and queue-list loops, for variables set
783 from spool message files. Do the same for the SMTP per-message loop, for
784 certain variables indirectly set in ACL operations.
786 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
787 as a multi-recipient message from a mailinglist manager). The coding had
788 an arbitrary cutoff number of characters while checking for more input;
789 enforced by writing a NUL into the buffer. This corrupted long / fast
790 input. The problem was exposed more widely when more pipelineing of SMTP
791 responses was introduced, and one Exim system was feeding another.
792 The symptom is log complaints of SMTP syntax error (NUL chars) on the
793 receiving system, and refused recipients seen by the sending system
794 (propating to people being dropped from mailing lists).
795 Discovered and pinpointed by David Carter.
797 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
798 replaced by the ${authresults } expansion.
800 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
802 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
803 allows proper process termination in container environments.
805 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
806 Previously the "final dot" had a newline after it; ensure it is CR,LF.
808 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
809 and "err_perm", deprecated since 4.83 when the RFC-defined words
810 "temperror" and "permerror" were introduced.
812 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
813 transport-filters or DKIM-signing. The restriction was lost in the
814 consolidation of verify-callout and delivery SMTP handling.
815 Extend the restriction to also cover ARC-signing.
817 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
818 in defer=pass mode supply a 450 to the initiator. Previously the message
821 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
822 tls_require_ciphers is used as before.
824 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
825 "pass_unscanned". Only tmpfails from the scanner are written to
826 the paniclog, as they may require admin intervention (permission
827 denied, license issues). Other scanner errors (like decompression
828 bombs) do not cause a paniclog entry.
830 JH/36 Fix reinitialisation of DKIM logging variable between messages.
831 Previously it was possible to log spurious information in receive log
834 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
835 triggered odd behaviour from Outlook Express clients.
837 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
840 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
841 since the IETF WG has not yet settled on that versus the original
842 "bare" representation.
844 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
845 Previously the millisecond value corrupted the output.
846 Fix also for syslog_pid=no and log_selector +pid, for which the pid
847 corrupted the output.
853 JH/01 Rework error string handling in TLS interface so that the caller in
854 more cases is responsible for logging. This permits library-sourced
855 string to be attached to addresses during delivery, and collapses
856 pairs of long lines into single ones.
858 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
859 during configuration. Wildcards are allowed and expanded.
861 JH/02 Rework error string handling in DKIM to pass more info back to callers.
862 This permits better logging.
864 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
865 do not close it down and have the child transport start it up again on
866 the passed-on TCP connection. Instead, proxy the child (and any
867 subsequent ones) for TLS via a unix-domain socket channel. Logging is
868 affected: the continued delivery log lines do not have any DNSSEC, TLS
869 Certificate or OCSP information. TLS cipher information is still logged.
871 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
872 identical IP addresses on different listening ports. Will also affect
875 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
876 add noisy ifdef guards to special-case this sillyness.
877 Patch from Bernd Kuhls.
879 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
880 than 255 are no longer allowed.
882 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
883 Disable the session-cache too, which might reduce our load. Since we
884 currrectly use a new context for every connection, both as server and
885 client, there is no benefit for these.
886 GnuTLS appears to not support tickets server-side by default (we don't
887 call gnutls_session_ticket_enable_server()) but client side is enabled
888 by default on recent versions (3.1.3 +) unless the PFS priority string
891 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
892 <https://reproducible-builds.org/specs/source-date-epoch/>.
894 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
895 the check for any unsuccessful recipients did not notice the limit, and
896 erroneously found still-pending ones.
898 JH/08 Pipeline CHUNKING command and data together, on kernels that support
899 MSG_MORE. Only in-clear (not on TLS connections).
901 JH/09 Avoid using a temporary file during transport using dkim. Unless a
902 transport-filter is involved we can buffer the headers in memory for
903 creating the signature, and read the spool data file once for the
904 signature and again for transmission.
906 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
907 4.77 as the kernel support then wasn't solid, having issues in 64bit
908 mode. Now, it's been long enough. Add support for FreeBSD also.
910 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
911 case where the routing stage had gathered several addresses to send to
912 a host before calling the transport for the first, we previously failed
913 to close down TLS in the old transport process before passing the TCP
914 connection to the new process. The new one sent a STARTTLS command
915 which naturally failed, giving a failed delivery and bloating the retry
916 database. Investigation and fix prototype from Wolfgang Breyha.
918 JH/12 Fix check on SMTP command input synchronisation. Previously there were
919 false-negatives in the check that the sender had not preempted a response
920 or prompt from Exim (running as a server), due to that code's lack of
921 awareness of the SMTP input buffering.
923 PP/04 Add commandline_checks_require_admin option.
924 Exim drops privileges sanely, various checks such as -be aren't a
925 security problem, as long as you trust local users with access to their
926 own account. When invoked by services which pass untrusted data to
927 Exim, this might be an issue. Set this option in main configuration
928 AND make fixes to the calling application, such as using `--` to stop
931 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
932 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
933 responses to those, into a single TLS record each way (this usually means
934 a single packet). As a side issue, smtp_enforce_sync now works on TLS
937 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
938 affects you only if you're dancing at the edge of the param size limits.
939 If you are, and this message makes sense to you, then: raise the
940 configured limit or use OpenSSL 1.1. Nothing we can do for older
943 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
944 cmdline element to get the documented default one. Previously it was
947 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
948 are unsuable for non-message-specific reasons - eg. connection timeout,
949 banner-time rejection.
951 JH/16 Fix logging of delivery remote port, when specified by router, under
954 PP/06 Repair manualroute's ability to take options in any order, even if one
955 is the name of a transport.
958 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
960 JH/17 Change the list-building routines interface to use the expanding-string
961 triplet model, for better allocation and copying behaviour.
963 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
964 Previously it was constructed the first time a possibly-matching string
965 was met in the configuration file input during startup; now it is done
968 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
969 compatible one, to avoid the (poorly documented) possibility of a config
970 file in the working directory redirecting the DB files, possibly correpting
971 some existing file. CVE-2017-10140 assigned for BDB.
973 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
974 cache-hot. Previously, although the result was properly cached, the
975 initial verify call returned a defer.
977 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
978 the main verify for receipient in uncached-mode.
980 JH/22 Retire historical build files to an "unsupported" subdir. These are
981 defined as "ones for which we have no current evidence of testing".
983 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
984 if present. Previously it was ignored.
986 JH/24 Start using specified-initialisers in C structure init coding. This is
987 a C99 feature (it's 2017, so now considered safe).
989 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
990 if was a fixed-sized field and bitmask ops via macros; it is now more
993 PP/07 GitHub PR 56: Apply MariaDB build fix.
994 Patch provided by Jaroslav Škarvada.
996 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
997 during Coverity cleanups [4.87 JH/47]
998 Diagnosis and fix provided by Michael Fischer v. Mollard.
1000 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1001 the right size to place the terminating semicolon on its own folded
1002 line, the header hash was calculated to an incorrect value thanks to
1003 the (relaxed) space the fold became.
1005 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1006 and confused the parent.
1008 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1009 which could crash as a result. This could lead to undeliverable messages.
1011 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1014 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1015 matters on fast-turnover and PID-randomising systems, which were getting
1016 out-of-order delivery.
1018 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1019 a possibly-overlapping copy. The symptom was that "Remote host closed
1020 connection in response to HELO" was logged instead of the actual 4xx
1023 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1024 Previously only that bufferd was discarded, resulting in SYMTP command
1027 JH/32 DKIM: when a message has multiple signatures matching an identity given
1028 in dkim_verify_signers, run the dkim acl once for each. Previously only
1029 one run was done. Bug 2189.
1031 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1032 "panic the current process" to "deliberately defer". The panic log is
1033 still written with the problem list name; the mail and reject logs now
1034 get a temp-reject line for the message that was being handled, saying
1035 something like "domains check lookup or other defer". The SMTP 451
1036 message is still "Temporary local problem".
1038 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1039 A crafted sequence of BDAT commands could result in in-use memory beeing
1040 freed. CVE-2017-16943.
1042 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1043 from SMTP input. Previously it was always done; now only done for DATA
1044 and not BDAT commands. CVE-2017-16944.
1046 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1047 to the message (such as an overlong header line). Previously this was
1048 not done and we did not exit BDAT mode. Followon from the previous item
1049 though a different problem.
1055 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1056 than -2003 did; needs libidn2 in addition to libidn.
1058 JH/02 The path option on a pipe transport is now expanded before use.
1060 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1061 Patch provided by "Björn", documentation fix added too.
1063 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1064 missing a wire-to-host endian conversion.
1066 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1067 close after a BDAT command line could be taken as a following command,
1068 giving a synch failure. Fix by only checking for synch immediately
1069 before acknowledging the chunk.
1071 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1072 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1073 macro. Patches provided by Josh Soref.
1075 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1076 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1077 by IANA as a service extension. However, John Klensin suggests that we
1080 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1081 the dkim code may be unix-mode line endings rather than smtp wire-format
1082 CRLF, so prepend a CR to any bare LF.
1084 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1085 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1087 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1088 the first were themselves being wrongly included in the feed into dkim
1089 processing; with most chunk sizes in use this resulted in an incorrect
1090 body hash calculated value.
1092 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1093 DKIM signature block, for verification. Although advised against by
1094 standards it is specifically not ruled illegal.
1096 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1098 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1099 missing a body hash (the bh= tag).
1101 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1102 It seems that HAProxy sends the Proxy Protocol information in clear and
1103 only then does a TLS startup, so do the same.
1105 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1106 TCP connections (such as for Spamd) unless the daemon successfully set
1107 Fast Open mode on its listening sockets. This fixes breakage seen on
1108 too-old kernels or those not configured for Fast Open, at the cost of
1109 requiring both directions being enabled for TFO, and TFO never being used
1110 by non-daemon-related Exim processes.
1112 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1113 endings, at least on the first header line. Try to canonify any that get
1114 past that check, despite the cost.
1116 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1117 now limited to an arbitrary five deep, while parsing addresses with the
1118 strip_excess_angle_brackets option enabled.
1120 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1121 instead leave the unprompted TLS handshake in socket buffer for the
1122 TLS library to consume.
1124 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1126 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1128 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1129 region in one operation, for speed, and this leaves assigned pointers
1130 dangling. Add checks run only under the testsuite which checks all
1131 variables at a store-reset and panics on a dangling pointer; add code
1132 explicitly nulling out all the variables discovered. Fixes one known
1133 bug: a transport crash, where a dangling pointer for $sending_ip_address
1134 originally assigned in a verify callout, is re-used.
1136 PP/06 Drop '.' from @INC in various Perl scripts.
1138 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1140 PP/08 Reduce a number of compilation warnings under clang; building with
1141 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1142 should be warning-free.
1144 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1146 HS/01 Fix portability problems introduced by PP/08 for platforms where
1147 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1149 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1150 chunk. This allows us to accept broken chunked messages. We need a more
1151 general solution here.
1153 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1154 already-broken messages in the queue.
1156 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1158 JH/19 Fix reference counting bug in routing-generated-address tracking.
1164 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1165 supports it and a size is available (ie. the sending peer gave us one).
1167 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1168 years of being deprecated). The replacements are the ACLs
1169 acl_smtp_mime and acl_not_smtp_mime.
1171 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1172 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1173 or even an in-clear connection were permitted. Now, if the host lookup
1174 was dnssec and dane was requested then the host is only used if the
1175 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1176 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1177 if one fails this test.
1178 This means that a poorly-configured remote DNS will make it incommunicado;
1179 but it protects against a DNS-interception attack on it.
1181 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1182 non-noisy when a race steals the message being considered.
1184 JH/05 If main configuration option tls_certificate is unset, generate a
1185 self-signed certificate for inbound TLS connections.
1187 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1188 in rewrites and routers.
1190 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1191 and logged a warning sing 4.83; now they are a configuration file error.
1193 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1194 (lacking @domain). Apply the same qualification processing as RCPT.
1196 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1198 JH/10 Support ${sha256:} applied to a string (as well as the previous
1201 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1202 a cutthrough deliver is pending, as we always want to make a connection.
1203 This also avoids re-routing the message when later placing the cutthrough
1204 connection after a verify cache hit.
1205 Do not update it with the verify result either.
1207 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1208 when routing results in more than one destination address.
1210 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1211 signing (which inhibits the cutthrough capability). Previously only
1212 the presence of an option was tested; now an expansion evaluating as
1213 empty is permissible (obviously it should depend only on data available
1214 when the cutthrough connection is made).
1216 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1217 the relevant preceding SMTP command did not note the pipelining mode.
1219 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1220 Previously they were not counted.
1222 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1223 as one having no matching records. Previously we deferred the message
1224 that needed the lookup.
1226 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1227 distinguished as "(=".
1229 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1230 for missing MX records. Previously it only worked for missing A records.
1232 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1234 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1235 after the data-go-ahead and data-ack. Patch from Jason Betts.
1237 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1238 even for a "none" policy. Patch from Tony Meyer.
1240 JH/22 Fix continued use of a connection for further deliveries. If a port was
1241 specified by a router, it must also match for the delivery to be
1244 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1245 When one of the recipients of a message was unsuitable for the connection
1246 (has no matching addresses), we lost track of needing to mark it
1247 deferred. As a result mail would be lost.
1249 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1251 JH/25 Decoding ACL controls is now done using a binary search; the source code
1252 takes up less space and should be simpler to maintain. Merge the ACL
1253 condition decode tables also, with similar effect.
1255 JH/26 Fix problem with one_time used on a redirect router which returned the
1256 parent address unchanged. A retry would see the parent address marked as
1257 delivered, so not attempt the (identical) child. As a result mail would
1260 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1261 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1262 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1265 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1266 Main config options check_{log,spool}_{inodes,space} are now
1267 100 inodes, 10MB unless set otherwise in the configuration.
1269 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1270 Previously it only applied to the main-section connection policy
1273 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1275 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1276 by me. Added RFC7919 DH primes as an alternative.
1278 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1279 are not in the system include path.
1281 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1282 GnuTLS, when a session startup failed (eg because the client disconnected)
1283 Exim did stdio operations after fclose. This was exposed by a recent
1284 change which nulled out the file handle after the fclose.
1286 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1287 signed directly by the cert-signing cert, rather than an intermediate
1288 OCSP-signing cert. This is the model used by LetsEncrypt.
1290 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1292 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1293 an incoming connection.
1295 HS/02 Bug 1802: Do not half-close the connection after sending a request
1298 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1299 fallback to "prime256v1".
1301 JH/34 SECURITY: Use proper copy of DATA command in error message.
1302 Could leak key material. Remotely exploitable. CVE-2016-9963.
1308 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1309 and 3.4.4 - once the server is enabled to respond to an OCSP request
1310 it does even when not requested, resulting in a stapling non-aware
1311 client dropping the TLS connection.
1313 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1314 support variable-length bit vectors. No functional change.
1316 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1317 The I= interface field on outgoing lines is now after the H= remote
1318 host field, same as incoming lines. There is a separate
1319 outgoing_interface log selector which allows you to disable the
1322 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1323 If not running log_selector +smtp_connection the mainlog would be held
1324 open indefinitely after a "too many connections" event, including to a
1325 deleted file after a log rotate. Leave the per net connection logging
1326 leaving it open for efficiency as that will be quickly detected by the
1327 check on the next write.
1329 HS/01 Bug 1671: Fix post transport crash.
1330 Processing the wait-<transport> messages could crash the delivery
1331 process if the message IDs didn't exist for some reason. When
1332 using 'split_spool_directory=yes' the construction of the spool
1333 file name failed already, exposing the same netto behaviour.
1335 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1336 mime_regex ACL conditions.
1338 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1339 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1340 to HELO, local diagnostic string.
1342 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1343 log line to debug. Even when configured with a tls authenticator many
1344 client connections are expected to not authenticate in this way, so
1345 an authenticate fail is not an error.
1347 HS/02 Add the Exim version string to the process info. This way exiwhat
1348 gives some more detail about the running daemon.
1350 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1351 matter for fast-change records such as DNSBLs.
1353 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1354 retry record, even if constant. There may be multiple transports with
1355 different interface settings and the retry behaviour needs to be kept
1358 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1360 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1362 JH/10 Bug 840: fix log_defer_output option of pipe transport
1364 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1365 hosts_randomize. This matters a lot when combined with mua_wrapper.
1367 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1368 ${quote_pgsql:<string>} operator.
1370 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1372 JH/14 Change header file name for freeradius-client. Relevant if compiling
1373 with Radius support; from the Gentoo tree and checked under Fedora.
1375 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1377 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1378 option transport_filter, meaning no filtering.
1380 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1382 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1383 defaults to "*" (all hosts). The variable is now available when not built
1384 with TLS, default unset, mainly to enable keeping the testsuite sane.
1385 If a server certificate is not supplied (via tls_certificate) an error is
1386 logged, and clients will find TLS connections fail on startup. Presumably
1387 they will retry in-clear.
1388 Packagers of Exim are strongly encouraged to create a server certificate
1389 at installation time.
1391 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1392 with the $config_file variable.
1394 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1395 in transport context, after the attempt, and per-recipient. The latter type
1396 is per host attempted. The event data is the error message, and the errno
1397 information encodes the lookup type (A vs. MX) used for the (first) host,
1398 and the trailing two digits of the smtp 4xx response.
1400 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1401 to write to mainlog (or rejectlog, paniclog) in the window between file
1402 creation and permissions/ownership being changed. Particularly affects
1403 installations where exicyclog is run as root, rather than exim user;
1404 result is that the running daemon panics and dies.
1406 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1408 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1409 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1410 "pri" and "weight". Note that the previous implicit priority given by the
1411 list order is no longer honoured.
1413 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1414 for DKIM processing.
1416 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1417 by defining SUPPORT_SOCKS.
1419 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1420 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1421 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1422 variables are renamed to proxy_{local,external}_{address,port}.
1424 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1425 for a build by defining SUPPORT_I18N
1427 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1428 of the query string, and make ${quote_redis:} do that quoting.
1430 JH/29 Move Events support from Experimental to mainline, enabled by default
1431 and removable for a build by defining DISABLE_EVENT.
1433 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1435 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1436 cached by the daemon.
1438 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1439 by defining LOOKUP_REDIS. The libhiredis library is required.
1441 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1442 keys are given for lookup.
1444 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1445 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1446 only supported when built with TLS support. The PolarSSL SHA routines
1447 are still used when the TLS library is too old for convenient support.
1449 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1450 openssl_options), for security. OpenSSL forces this from version 1.1.0
1451 server-side so match that on older versions.
1453 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1454 allocation for $value could be released as the expansion processing
1455 concluded, but leaving the global pointer active for it.
1457 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1458 and to use the domains and local_parts ACL conditions.
1460 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1461 incorrectly not doubled on cutthrough transmission, hence seen as a
1462 body-termination at the receiving system - resulting in truncated mails.
1463 Commonly the sender saw a TCP-level error, and retransmitted the message
1464 via the normal store-and-forward channel. This could result in duplicates
1465 received - but deduplicating mailstores were liable to retain only the
1466 initial truncated version.
1468 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1470 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1472 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1473 we're in there, support oversigning also; bug 1309.
1475 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1477 HS/04 Add support for keep_environment and add_environment options.
1479 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1480 either intentional arithmetic overflow during PRNG, or testing config-
1483 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1484 delivery resulted in actual delivery. Cancel cutthrough before DATA
1487 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1488 rejects a recipient: pass the reject to the originator.
1490 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1491 Many were false-positives and ignorable, but it's worth fixing the
1494 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1495 for the new environment-manipulation done at startup. Move the routines
1496 from being local to tls.c to being global via the os.c file.
1498 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1499 an extract embedded as result-arg for a map, the first arg for extract
1500 is unavailable so we cannot tell if this is a numbered or keyed
1501 extraction. Accept either.
1507 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1510 JH/02 The smtp transport option "multi_domain" is now expanded.
1512 JH/03 The smtp transport now requests PRDR by default, if the server offers
1515 JH/04 Certificate name checking on server certificates, when exim is a client,
1516 is now done by default. The transport option tls_verify_cert_hostnames
1517 can be used to disable this per-host. The build option
1518 EXPERIMENTAL_CERTNAMES is withdrawn.
1520 JH/05 The value of the tls_verify_certificates smtp transport and main options
1521 default to the word "system" to access the system default CA bundle.
1522 For GnuTLS, only version 3.0.20 or later.
1524 JH/06 Verification of the server certificate for a TLS connection is now tried
1525 (but not required) by default. The verification status is now logged by
1526 default, for both outbound TLS and client-certificate supplying inbound
1529 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1532 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1533 Status Notification (bounce) messages are now MIME format per RFC 3464.
1534 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1535 under the control of the dsn_advertise_hosts option, and routers may
1536 have a dsn_lasthop option.
1538 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1539 default, modifiable by a malware= option. The list separator for
1540 the options can now be changed in the usual way. Bug 68.
1542 JH/10 The smtp_receive_timeout main option is now expanded before use.
1544 JH/11 The incoming_interface log option now also enables logging of the
1545 local interface on delivery outgoing connections.
1547 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1548 if the interface and destination host and port all match.
1550 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1553 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1554 Patch from Andrew Lewis.
1556 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1557 now supports optional time-restrictions, weighting, and priority
1558 modifiers per server. Patch originally by <rommer@active.by>.
1560 JH/16 The spamd_address main option now supports a mixed list of local
1561 and remote servers. Remote servers can be IPv6 addresses, and
1562 specify a port-range.
1564 JH/17 Bug 68: The spamd_address main option now supports an optional
1565 timeout value per server.
1567 JH/18 Bug 1581: Router and transport options headers_add/remove can
1568 now have the list separator specified.
1570 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1573 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1576 JH/21 Support for the A6 type of dns record is withdrawn.
1578 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1579 rather than the verbs used.
1581 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1582 from 255 to 1024 chars.
1584 JH/24 Verification callouts now attempt to use TLS by default.
1586 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1587 are generic router options now. The defaults didn't change.
1589 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1590 Original patch from Alexander Shikoff, worked over by JH.
1592 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1593 files if ZCAT_COMMAND is not executable.
1595 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1597 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1599 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1600 Normally benign, it bites when the pair was led to by a CNAME;
1601 modern usage is to not canonicalize the domain to a CNAME target
1602 (and we were inconsistent anyway for A-only vs AAAA+A).
1604 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1606 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1607 when evaluating $sender_host_dnssec.
1609 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1610 $sender_helo_dnssec variable.
1612 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1614 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1616 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1618 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1619 documented as working, but never had. Support all but $spam_report.
1621 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1622 added for tls authenticator.
1624 HS/03 Add perl_taintmode main config option
1630 TL/01 When running the test suite, the README says that variables such as
1631 no_msglog_check are global and can be placed anywhere in a specific
1632 test's script, however it was observed that placement needed to be near
1633 the beginning for it to behave that way. Changed the runtest perl
1634 script to read through the entire script once to detect and set these
1635 variables, reset to the beginning of the script, and then run through
1636 the script parsing/test process like normal.
1638 TL/02 The BSD's have an arc4random API. One of the functions to induce
1639 adding randomness was arc4random_stir(), but it has been removed in
1640 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1641 function when detected.
1643 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1644 cause callback expansion.
1646 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1647 syntax errors in an expansion can be treated as a string instead of
1648 logging or causing an error, due to the internal use of bool_lax
1649 instead of bool when processing it.
1651 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1652 server certificates when making smtp deliveries.
1654 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1656 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1658 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1660 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1661 Merged patch from Sebastian Wiedenroth.
1663 JH/05 Fix results-pipe from transport process. Several recipients, combined
1664 with certificate use, exposed issues where response data items split
1665 over buffer boundaries were not parsed properly. This eventually
1666 resulted in duplicates being sent. This issue only became common enough
1667 to notice due to the introduction of connection certificate information,
1668 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1670 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1671 size buffer was used, resulting in syntax errors when an expansion
1674 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1675 version 3.3.6 or later.
1677 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1678 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1679 both become "event_action", the variables become $event_name, $event_data
1680 and $event_defer_errno. There is a new variable $verify_mode, usable in
1681 routers, transports and related events. The tls:cert event is now also
1682 raised for inbound connections, if the main configuration event_action
1685 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1686 early OCSP support, but no stapling (appears to be less than 1.0.0).
1688 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1689 server certificate names available under the smtp transport option
1690 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1693 JH/10 Time-related extraction expansions from certificates now use the main
1694 option "timezone" setting for output formatting, and are consistent
1695 between OpenSSL and GnuTLS compilations. Bug 1541.
1697 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1698 encoded parameter in the incoming message. Bug 1558.
1700 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1701 include certificate info, eximon was claiming there were spoolfile
1704 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1706 JH/14 Log delivery-related information more consistently, using the sequence
1707 "H=<name> [<ip>]" wherever possible.
1709 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1710 are problematic for Debian distribution, omit them from the release
1713 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1715 JH/16 Fix string representation of time values on 64bit time_t architectures.
1718 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1719 output list separator was used.
1724 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1725 checkers that were complaining about end of non-void function with no
1728 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1729 This was a regression introduced in 4.83 by another bugfix.
1731 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1733 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1734 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1740 TF/01 Correctly close the server side of TLS when forking for delivery.
1742 When a message was received over SMTP with TLS, Exim failed to clear up
1743 the incoming connection properly after forking off the child process to
1744 deliver the message. In some situations the subsequent outgoing
1745 delivery connection happened to have the same fd number as the incoming
1746 connection previously had. Exim would try to use TLS and fail, logging
1747 a "Bad file descriptor" error.
1749 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1750 utilities have not been installed.
1752 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1753 temporary space as the ACL may create new global variables.
1755 TL/01 LDAP support uses per connection or global context settings, depending
1756 upon the detected version of the libraries at build time.
1758 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1759 to extract and use the src ip:port in logging and expansions as if it
1760 were a direct connection from the outside internet. PPv2 support was
1761 updated based on HAProxy spec change in May 2014.
1763 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1765 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1766 Properly escape header and check for NULL return.
1768 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1771 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1773 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1774 characters in header names, implemented as a verify condition.
1775 Contributed by Michael Fischer v. Mollard.
1777 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1778 results permerror and temperror. Previous values are deprecated but
1779 still accepted. In a future release, err_perm and err_temp will be
1780 completely removed, which will be a backward incompatibility if the
1781 ACL tests for either of these two old results. Patch contributed by
1782 user bes-internal on the mailing list.
1784 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1786 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1787 selectors, in both main and reject logs.
1789 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1792 JH/07 Add malware type "sock" for talking to simple daemon.
1794 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1796 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1797 routers/transports under cutthrough routing.
1799 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1800 numbers. Touch up "bool" conditional to keep the same definition.
1802 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1804 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1806 JH/12 Expand items in router/transport headers_add or headers_remove lists
1807 individually rather than the list as a whole. Bug 1452.
1809 Required for reasonable handling of multiple headers_ options when
1810 they may be empty; requires that headers_remove items with embedded
1811 colons must have them doubled (or the list-separator changed).
1813 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1814 view the policy declared in the DMARC record. Currently, $dmarc_status
1815 is a combined value of both the record presence and the result of the
1818 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1820 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1821 dnslookup router and the smtp transport (applying to the forward
1824 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1825 of ldap servers used for a specific lookup. Patch provided by Heiko
1828 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1829 New variable $lookup_dnssec_authenticated for observability.
1831 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1832 Patch submitted by Lars Timman.
1834 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1836 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1837 Requires trusted mode and valid format message id, aborts otherwise.
1838 Patch contributed by Heiko Schlichting.
1840 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1841 certextract with support for various fields. Bug 1358.
1843 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1844 is requested by default, modifiable by smtp transport option
1847 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1848 operate on certificate variables to give certificate fingerprints
1849 Also new ${sha256:cert_variable}.
1851 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1853 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1854 Christian Aistleitner.
1856 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1858 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1859 file. Patch from Wolfgang Breyha.
1861 JH/25 Expand the coverage of the delivery $host and $host_address to
1862 client authenticators run in verify callout. Bug 1476.
1864 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1865 align with daemon_smtp_ports. Bug 72.
1867 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1868 support and error reporting did not work properly.
1870 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1871 and is readable. Patch from Andrew Colin Kissa.
1873 TL/14 Enhance documentation of ${run expansion and how it parses the
1874 commandline after expansion, particularly in the case when an
1875 unquoted variable expansion results in an empty value.
1877 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1879 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1880 by renaming away from it. Addresses GH issue 15, reported by
1883 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1885 TL/15 SECURITY: prevent double expansion in math comparison functions
1886 (can expand unsanitized data). Not remotely exploitable.
1893 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1895 PP/02 Make -n do something, by making it not do something.
1896 When combined with -bP, the name of an option is not output.
1898 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1901 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1902 $sender_host_name and config options to manage this, and basic check
1905 PP/05 DSCP support for outbound connections and control modifier for inbound.
1907 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1908 (Only plugin which currently uses this is kerberos4, which nobody should
1909 be using, but we should make it available and other future plugins might
1910 conceivably use it, even though it would break NAT; stuff *should* be
1911 using channel bindings instead).
1913 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1914 name; added for Sendmail compatibility; requires admin caller.
1915 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1916 just ignore it); requires trusted caller.
1917 Also parse but ignore: -Ac -Am -X<logfile>
1920 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1922 TL/02 Add +smtp_confirmation as a default logging option.
1924 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1925 Patch by Magnus Holmgren from 2007-02-20.
1927 TL/04 Bugzilla 1281 - Spec typo.
1928 Bugzilla 1283 - Spec typo.
1929 Bugzilla 1290 - Spec grammar fixes.
1931 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1933 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1935 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1936 mailing list by by Dmitry Isaikin.
1938 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1940 JH/02 Support "G" suffix to numbers in ${if comparisons.
1942 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1944 NM/01 Bugzilla 1197 - Spec typo
1945 Bugzilla 1196 - Spec examples corrections
1947 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1949 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1950 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1953 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1954 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1956 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1957 "acl {{name}{arg}...}", and optional args on acl condition
1960 JH/05 Permit multiple router/transport headers_add/remove lines.
1962 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1964 JH/07 Avoid using a waiting database for a single-message-only transport.
1965 Performance patch from Paul Fisher. Bugzilla 1262.
1967 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1970 JH/09 Add $headers_added variable, with content from use of ACL modifier
1971 add_header (but not yet added to the message). Bugzilla 199.
1973 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1974 Pulled from Bugzilla 817 by Wolfgang Breyha.
1976 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1978 (nb: this is the same fix as in Exim 4.80.1)
1980 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1983 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1986 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1988 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1989 advertise SMTP AUTH mechanism to us, instead of a generic
1990 protocol violation error. Also, make Exim more robust to bad
1991 data from the Dovecot auth socket.
1993 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1995 When a queue runner is handling a message, Exim first routes the
1996 recipient addresses, during which it prunes them based on the retry
1997 hints database. After that it attempts to deliver the message to
1998 any remaining recipients. It then updates the hints database using
2001 So if a recipient address works intermittently, it can get repeatedly
2002 deferred at routing time. The retry hints record remains fresh so the
2003 address never reaches the final cutoff time.
2005 This is a fairly common occurrence when a user is bumping up against
2006 their storage quota. Exim had some logic in its local delivery code
2007 to deal with this. However it did not apply to per-recipient defers
2008 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2010 This change adds a proper retry rule check during routing so that the
2011 final cutoff time is checked against the message's age. We only do
2012 this check if there is an address retry record and there is not a
2013 domain retry record; this implies that previous attempts to handle
2014 the address had the retry_use_local_parts option turned on. We use
2015 this as an approximation for the destination being like a local
2016 delivery, as in LMTP.
2018 I suspect this new check makes the old local delivery cutoff check
2019 redundant, but I have not verified this so I left the code in place.
2021 TF/02 Correct gecos expansion when From: is a prefix of the username.
2023 Test 0254 submits a message to Exim with the header
2027 When I ran the test suite under the user fanf2, Exim expanded
2028 the header to contain my full name, whereas it should have added
2029 a Resent-Sender: header. It erroneously treats any prefix of the
2030 username as equal to the username.
2032 This change corrects that bug.
2034 GF/01 DCC debug and logging tidyup
2035 Error conditions log to paniclog rather than rejectlog.
2036 Debug lines prefixed by "DCC: " to remove any ambiguity.
2038 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2040 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2041 Bug spotted by Jeremy Harris; was flawed since initial commit.
2042 Would have resulted in OCSP responses post-SNI triggering an Exim
2043 NULL dereference and crash.
2045 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2047 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2048 Bug detection, analysis and fix by Samuel Thibault.
2049 Bugzilla 1331, Debian bug #698092.
2051 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2053 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2054 Server implementation by Todd Lyons, client by JH.
2055 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2056 config variable "prdr_enable" controls whether the server
2057 advertises the facility. If the client requests PRDR a new
2058 acl_data_smtp_prdr ACL is called once for each recipient, after
2059 the body content is received and before the acl_smtp_data ACL.
2060 The client is controlled by both of: a hosts_try_prdr option
2061 on the smtp transport, and the server advertisement.
2062 Default client logging of deliveries and rejections involving
2063 PRDR are flagged with the string "PRDR".
2065 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2066 fclose(). Diagnosis by Todd Lyons.
2068 PP/17 Update configure.default to handle IPv6 localhost better.
2069 Patch by Alain Williams (plus minor tweaks).
2072 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2073 This is now consistent with GnuTLS, and is now documented: the
2074 previous undocumented portable approach to treating the option as
2075 unset was to force an expansion failure. That still works, and
2076 an empty string is now equivalent.
2078 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2079 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2080 not performing validation itself.
2082 PP/20 Added force_command boolean option to pipe transport.
2083 Patch from Nick Koston, of cPanel Inc.
2085 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2088 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2090 PP/21 Fix eximon continuous updating with timestamped log-files.
2091 Broken in a format-string cleanup in 4.80, missed when I repaired the
2092 other false fix of the same issue.
2093 Report and fix from Heiko Schlichting.
2096 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2097 Report from Prashanth Katuri.
2099 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2100 It's SecureTransport, so affects any MacOS clients which use the
2101 system-integrated TLS libraries, including email clients.
2103 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2104 using a MIME ACL for non-SMTP local injection.
2105 Report and assistance in diagnosis by Warren Baker.
2107 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2109 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2111 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2112 last id that failed so it may be referenced in subsequent ACL's.
2114 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2117 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2118 ldap library initialization, allowing self-signed CA's to be
2119 used. Also properly sets require_cert option later in code by
2120 using NULL (global ldap config) instead of ldap handle (per
2121 session). Bug diagnosis and testing by alxgomz.
2123 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2124 the src/util/ subdirectory.
2126 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2127 renamed to Transport Post Delivery Action by Jeremy Harris, as
2130 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2131 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2132 redis_servers = needs to be configured which will be used by the redis
2133 lookup. Patch from Warren Baker, of The Packet Hub.
2135 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2137 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2138 hostname or reverse DNS when processing a host list. Used suggestions
2139 from multiple comments on this bug.
2141 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2143 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2144 Missed a few lines, added it to make the runtest require no keyboard
2147 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2148 contains upper case chars. Make router use caseful_local_part.
2150 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2151 support when GnuTLS has been built with p11-kit.
2157 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2159 This, or similar/improved, will also be change PP/11 of 4.82.
2165 PP/01 Handle short writes when writing local log-files.
2166 In practice, only affects FreeBSD (8 onwards).
2167 Bugzilla 1053, with thanks to Dmitry Isaikin.
2169 NM/01 Bugzilla 949 - Documentation tweak
2171 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2174 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2176 PP/02 Implemented gsasl authenticator.
2178 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2180 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2181 `pkg-config foo` for cflags/libs.
2183 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2184 with rest of GSASL and with heimdal_gssapi.
2186 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2187 `pkg-config foo` for cflags/libs for the TLS implementation.
2189 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2190 properties get this fed in as external SSF. A number of robustness
2191 and debugging improvements to the cyrus_sasl authenticator.
2193 PP/08 cyrus_sasl server now expands the server_realm option.
2195 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2196 Patch by Jeremy Harris.
2198 PP/10 Added dbmjz lookup type.
2200 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2202 PP/12 MAIL args handles TAB as well as SP, for better interop with
2203 non-compliant senders.
2204 Analysis and variant patch by Todd Lyons.
2206 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2207 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2208 Patch from Dirk Mueller <dmueller@suse.com>
2210 PP/13 tls_peerdn now print-escaped for spool files.
2211 Observed some $tls_peerdn in wild which contained \n, which resulted
2212 in spool file corruption.
2214 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2215 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2216 or write after TLS renegotiation, which otherwise led to messages
2219 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2220 as a tracking header (ie: a signed header comes before the signature).
2221 Patch from Wolfgang Breyha.
2223 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2224 comma-sep list; embedded commas doubled.
2226 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2228 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2230 Report and patch from Dmitry Banschikov.
2232 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2233 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2234 failures. We appear to now support TLS1.1+ with Exim.
2236 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2237 lets Exim select keys and certificates based upon TLS SNI from client.
2238 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2239 before an outbound SMTP session. New log_selector, +tls_sni.
2241 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2242 NULL dereference. Report and patch from Alun Jones.
2244 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2245 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2246 Not seeing resolver debug output on NetBSD, but suspect this is a
2247 resolver implementation change.
2249 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2250 Left warnings. Added "eximon gdb" invocation mode.
2252 PP/21 Defaulting "accept_8bitmime" to true, not false.
2254 PP/22 Added -bw for inetd wait mode support.
2256 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2257 locate the relevant includes and libraries. Made this the default.
2259 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2260 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2262 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2263 This may cause build issues on older platforms.
2265 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2266 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2267 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2268 Added SNI support via GnuTLS too.
2269 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2271 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2273 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2274 Applied second patch from Janne, implementing suggestion to default
2275 multiple-strings-in-record handling to match SPF spec.
2277 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2279 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2280 read-only, out of scope).
2281 Patch from Wolfgang Breyha, report from Stuart Northfield.
2283 PP/29 Fix three issues highlighted by clang analyser static analysis.
2284 Only crash-plausible issue would require the Cambridge-specific
2285 iplookup router and a misconfiguration.
2286 Report from Marcin Mirosław.
2288 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2290 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2291 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2292 As part of this, removing so much warning spew let me fix some minor
2293 real issues in debug logging.
2295 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2296 assignment on my part. Fixed.
2298 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2299 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2300 Janne Snabb (who went above and beyond: thank you).
2302 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2303 string otherwise requires a connection and a bunch more work and it's
2304 relatively easy to get wrong. Should also expose TLS library linkage
2307 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2308 64-bit ${eval} (JH/03).
2310 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2311 GNU libc to support some of the 64-bit stuff, should not lead to
2312 conflicts. Defined before os.h is pulled in, so if a given platform
2313 needs to override this, it can.
2315 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2316 protection layer was required, which is not implemented.
2317 Bugzilla 1254, patch from Wolfgang Breyha.
2319 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2320 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2321 tls_dhparam take prime identifiers. Also unbreak combination of
2322 OpenSSL+DH_params+TLSSNI.
2324 PP/39 Disable SSLv2 by default in OpenSSL support.
2330 PP/01 Solaris build fix for Oracle's LDAP libraries.
2331 Bugzilla 1109, patch from Stephen Usher.
2333 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2335 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2338 TF/02 Fix a couple more cases where we did not log the error message
2339 when unlink() failed. See also change 4.74-TF/03.
2341 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2342 lock up or crash if it happened to be inside a call to libc when it
2343 got a SIGUSR1 from exiwhat.
2345 The SIGUSR1 handler appends the current process status to the process
2346 log which is later printed by exiwhat. It used to use the general
2347 purpose logging code to do this, but several functions it calls are
2348 not safe for signals.
2350 The new output code in the SIGUSR1 handler is specific to the process
2351 log, and simple enough that it's easy to inspect for signal safety.
2352 Removing some special cases also simplifies the general logging code.
2353 Removing the spurious timestamps from the process log simplifies
2356 TF/04 Improved ratelimit ACL condition.
2358 The /noupdate option has been deprecated in favour of /readonly which
2359 has clearer semantics. The /leaky, /strict, and /readonly update modes
2360 are mutually exclusive. The update mode is no longer included in the
2361 database key; it just determines when the database is updated. (This
2362 means that when you upgrade Exim will forget old rate measurements.)
2364 Exim now checks that the per_* options are used with an update mode that
2365 makes sense for the current ACL. For example, when Exim is processing a
2366 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2367 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2368 must specify per_mail/readonly. If you omit the update mode it defaults to
2369 /leaky where that makes sense (as before) or /readonly where required.
2371 The /noupdate option is now undocumented but still supported for
2372 backwards compatibility. It is equivalent to /readonly except that in
2373 ACLs where /readonly is required you may specify /leaky/noupdate or
2374 /strict/noupdate which are treated the same as /readonly.
2376 A useful new feature is the /count= option. This is a generalization
2377 of the per_byte option, so that you can measure the throughput of other
2378 aggregate values. For example, the per_byte option is now equivalent
2379 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2381 The per_rcpt option has been generalized using the /count= mechanism
2382 (though it's more complicated than the per_byte equivalence). When it is
2383 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2384 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2385 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2386 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2387 Note that using per_rcpt with a non-readonly update mode in more than
2388 one ACL will cause the recipients to be double-counted. (The per_mail
2389 and per_byte options don't have this problem.)
2391 The handling of very low rates has changed slightly. If the computed rate
2392 is less than the event's count (usually one) then this event is the first
2393 after a long gap. In this case the rate is set to the same as this event's
2394 count, so that the first message of a spam run is counted properly.
2396 The major new feature is a mechanism for counting the rate of unique
2397 events. The new per_addr option counts the number of different
2398 recipients that someone has sent messages to in the last time period. It
2399 behaves like per_rcpt if all the recipient addresses are different, but
2400 duplicate recipient addresses do not increase the measured rate. Like
2401 the /count= option this is a general mechanism, so the per_addr option
2402 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2403 example, measure the rate that a client uses different sender addresses
2404 with the options per_mail/unique=$sender_address. There are further
2405 details in the main documentation.
2407 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2409 TF/06 Removed a few PCRE remnants.
2411 TF/07 Automatically extract Exim's version number from tags in the git
2412 repository when doing development or release builds.
2414 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2415 Bugzilla 879. Patch from Paul Fisher.
2417 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2418 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2421 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2423 PP/05 Variable $av_failed, true if the AV scanner deferred.
2424 Bugzilla 1078. Patch from John Horne.
2426 PP/06 Stop make process more reliably on build failure.
2427 Bugzilla 1087. Patch from Heiko Schlittermann.
2429 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2430 Bugzilla 1089. Patch from Heiko Schlittermann.
2432 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2433 Bugzilla 1131. Patch from Holger Weiß.
2435 PP/09 Handle IPv6 addresses with SPF.
2436 Bugzilla 860. Patch from Wolfgang Breyha.
2438 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2440 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2443 PP/11 match_* no longer expand right-hand-side by default.
2444 New compile-time build option, EXPAND_LISTMATCH_RHS.
2445 New expansion conditions, "inlist", "inlisti".
2447 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2449 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2451 PP/14 fix log_write() format string regression from TF/03.
2452 Bugzilla 1152. Patch from Dmitry Isaikin.
2458 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2460 PP/02 Harmonised TLS library version reporting; only show if debugging.
2461 Layout now matches that introduced for other libraries in 4.74 PP/03.
2463 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2465 PP/04 New "dns_use_edns0" global option.
2467 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2470 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2471 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2473 TK/01 Updated PolarSSL code to 0.14.2.
2474 Bugzilla 1097. Patch from Andreas Metzler.
2476 PP/07 Catch divide-by-zero in ${eval:...}.
2477 Fixes bugzilla 1102.
2479 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2482 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2483 format-string attack -- SECURITY: remote arbitrary code execution.
2485 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2486 time unintentionally subject to list matching rules, letting the header
2487 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2488 arbitrary string expansion). This allowed for information disclosure.
2490 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2491 INT_MIN/-1 -- value coerced to INT_MAX.
2497 NM/01 Workaround for PCRE version dependency in version reporting
2500 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2501 This fixes portability to compilers other than gcc, notably
2502 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2504 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2505 makefiles for portability to HP-UX and POSIX correctness.
2507 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2508 Also via indented variable definition in the Makefile.
2509 (Debugging by Oliver Heesakkers).
2511 PP/02 Restore caching of spamd results with expanded spamd_address.
2512 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2514 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2515 Improves build reliability. Fix from: Frank Elsner
2517 NM/02 Fix wide character breakage in the rfc2047 coding
2518 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2520 NM/03 Allow underscore in dnslist lookups
2521 Fixes bug 1026. Patch from Graeme Fowler
2523 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2524 Code patches from Adam Ciarcinski of NetBSD.
2526 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2529 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2530 is logged, to avoid truncation. Patch from John Horne.
2532 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2533 Patch from Jakob Hirsch.
2535 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2536 SQL string expansion failure details.
2537 Patch from Andrey Oktyabrski.
2539 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2540 Patch from Simon Arlott.
2542 PP/09 New lookups functionality failed to compile on old gcc which rejects
2543 extern declarations in function scope.
2544 Patch from Oliver Fleischmann
2546 PP/10 Use sig_atomic_t for flags set from signal handlers.
2547 Check getgroups() return and improve debugging.
2548 Fixed developed for diagnosis in bug 927 (which turned out to be
2551 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2552 Patch from Mark Zealey.
2554 PP/12 Bugzilla 1056: Improved spamd server selection.
2555 Patch from Mark Zealey.
2557 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2558 Based on patch from Heiko Schlittermann.
2560 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2561 Patch from Uwe Doering, sign-off by Michael Haardt.
2563 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2564 variable declaration deep within a block. Bug and patch from
2567 PP/15 lookups-Makefile IRIX compatibility coercion.
2569 PP/16 Make DISABLE_DKIM build knob functional.
2571 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2572 Patch by Simon Arlott
2574 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2575 variable argument macros. Our copy now differs from upstream.
2581 TF/01 Failure to get a lock on a hints database can have serious
2582 consequences so log it to the panic log.
2584 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2585 controlled using the smtp_confirmation log selector.
2587 TF/03 Include the error message when we fail to unlink a spool file.
2589 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2590 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2591 for maintaining out-of-tree patches for some time.
2593 PP/01 Bugzilla 139: Documentation and portability issues.
2594 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2595 Handle per-OS dynamic-module compilation flags.
2597 PP/02 Let /dev/null have normal permissions.
2598 The 4.73 fixes were a little too stringent and complained about the
2599 permissions on /dev/null. Exempt it from some checks.
2600 Reported by Andreas M. Kirchwitz.
2602 PP/03 Report version information for many libraries, including
2603 Exim version information for dynamically loaded libraries. Created
2604 version.h, now support a version extension string for distributors
2605 who patch heavily. Dynamic module ABI change.
2607 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2608 privilege escalation vulnerability whereby the Exim run-time user
2609 can cause root to append content of the attacker's choosing to
2612 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2615 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2616 If dropping privileges for untrusted macros, we disabled normal logging
2617 on the basis that it would fail; for the Exim run-time user, this is not
2618 the case, and it resulted in successful deliveries going unlogged.
2619 Fixed. Reported by Andreas Metzler.
2625 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2626 only prepend for the Resent-* case. Fixes regression introduced in
2627 Exim 4.70 by NM/22 for Bugzilla 607.
2629 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2630 increasing numbers of administrators be bitten by this.
2632 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2634 PP/03 Bugzilla 994: added openssl_options main configuration option.
2636 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2638 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2640 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2642 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2643 without a peer certificate, leading to a segfault because of an
2644 assumption that peers always have certificates. Be a little more
2645 paranoid. Problem reported by Martin Tscholak.
2647 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2648 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2649 NB: ClamAV planning to remove STREAM in "middle of 2010".
2650 CL also introduces -bmalware, various -d+acl logging additions and
2651 more caution in buffer sizes.
2653 PP/09 Implemented reverse_ip expansion operator.
2655 PP/10 Bugzilla 937: provide a "debug" ACL control.
2657 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2659 PP/12 Bugzilla 973: Implement --version.
2661 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2663 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2665 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2667 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2668 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2669 ignore trailing whitespace.
2671 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2673 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2676 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2677 Notification from Dr Andrew Aitchison.
2679 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2680 ExtendedDetectionInfo response format.
2681 Notification from John Horne.
2683 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2686 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2687 XSL and documented dependency on system catalogs, with examples of how
2690 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2693 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2694 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2695 configuration file which is writeable by the Exim user or group.
2697 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2698 of configuration files to cover files specified with the -C option if
2699 they are going to be used with root privileges, not just the default
2702 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2703 option (effectively making it always true).
2705 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2706 files to be used while preserving root privileges.
2708 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2709 that rogue child processes cannot use them.
2711 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2712 run-time user, instead of root.
2714 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2715 Exim run-time user without dropping privileges.
2717 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2718 result string, instead of calling string_vformat() twice with the same
2721 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2722 for other users. Others should always drop root privileges if they use
2723 -C on the command line, even for a whitelisted configure file.
2725 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2727 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2733 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2734 $data_path, and $header_path variables; fixed documentation bugs and
2737 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2738 exipick to access non-standard spools, including the "frozen" queue
2741 NM/01 Bugzilla 965: Support mysql stored procedures.
2742 Patch from Alain Williams
2744 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2746 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2747 Patch from Andreas Metzler
2749 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2750 Patch from Kirill Miazine
2752 NM/05 Bugzilla 671: Added umask to procmail example.
2754 JJ/03 installed exipick 20100323.0, fixing doc bug
2756 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2757 directory. Notification and patch from Dan Rosenberg.
2759 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2761 TK/02 Improve log output when DKIM signing operation fails.
2763 MH/01 Treat the transport option dkim_domain as a colon separated
2764 list, not as a single string, and sign the message with each element,
2765 omitting multiple occurences of the same signer.
2767 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2768 Bugzilla 985, 986. Patch by Simon Arlott
2770 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2771 Patch by Simon Arlott
2773 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2774 MBX locking. Notification from Dan Rosenberg.
2780 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2782 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2784 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2786 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2788 NM/04 Bugzilla 915: Flag broken perl installation during build.
2794 TK/01 Added patch by Johannes Berg that expands the main option
2795 "spamd_address" if it starts with a dollar sign.
2797 TK/02 Write list of recipients to X-Envelope-Sender header when building
2798 the mbox-format spool file for content scanning (suggested by Jakob
2801 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2802 (http://www.dcc-servers.net/) support via dccifd. Activated by
2803 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2805 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2806 by Mark Daniel Reidel <mr@df.eu>.
2808 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2809 When building exim an external PCRE library is now needed -
2810 PCRE is a system library on the majority of modern systems.
2811 See entry on PCRE_LIBS in EDITME file.
2813 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2814 conversation. Added nologin parameter to request.
2815 Patch contributed by Kirill Miazine.
2817 TF/01 Do not log submission mode rewrites if they do not change the address.
2819 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2821 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2822 log files in place. Contributed by Roberto Lima.
2824 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2826 TF/03 Bugzilla 615: When checking the local_parts router precondition
2827 after a local_part_suffix or local_part_prefix option, Exim now
2828 does not use the address's named list lookup cache, since this
2829 contains cached lookups for the whole local part.
2831 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2832 Robert Millan. Documentation is in experimental-spec.txt.
2834 TF/04 Bugzilla 668: Fix parallel build (make -j).
2836 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2838 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2839 Patch provided by Jan Srzednicki.
2841 TF/05 Leading white space used to be stripped from $spam_report which
2842 wrecked the formatting. Now it is preserved.
2844 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2845 that they are available at delivery time.
2847 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2849 TF/08 TLS error reporting now respects the incoming_interface and
2850 incoming_port log selectors.
2852 TF/09 Produce a more useful error message if an SMTP transport's hosts
2853 setting expands to an empty string.
2855 NM/06 Bugzilla 744: EXPN did not work under TLS.
2856 Patch provided by Phil Pennock.
2858 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2859 Patch provided by Richard Godbee.
2861 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2862 acl_smtp_notquit, added index entry.
2864 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2865 Patch provided by Eugene Bujak.
2867 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2868 accept(). Patch provided by Maxim Dounin.
2870 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2871 Patch provided by Phil Pennock.
2873 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2875 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2876 Patch provided by Brad "anomie" Jorsch.
2878 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2879 Patch provided by Dean Brooks.
2881 TK/05 Add native DKIM support (does not depend on external libraries).
2883 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2884 Patch provided by Graeme Fowler.
2886 NM/16 Bugzilla 851: Documentation example syntax fix.
2888 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2890 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2893 NM/19 Bugzilla 745: TLS version reporting.
2894 Patch provided by Phil Pennock.
2896 NM/20 Bugzilla 167: bool: condition support.
2897 Patch provided by Phil Pennock.
2899 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2900 clients. Patch provided by Phil Pennock.
2902 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2903 Patch provided by Brad "anomie" Jorsch.
2905 NM/23 Bugzilla 687: Fix misparses in eximstats.
2906 Patch provided by Heiko Schlittermann.
2908 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2909 Patch provided by Heiko Schlittermann.
2911 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2912 plus update to original patch.
2914 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2916 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2917 Patch provided by David Brownlee.
2919 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2921 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2923 NM/30 Bugzilla 888: TLS documentation bugfixes.
2925 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2927 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2928 Unlike the original bugzilla I have changed all shell scripts in src tree.
2930 NM/33 Bugzilla 898: Transport filter timeout fix.
2931 Patch by Todd Rinaldo.
2933 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2934 Patch by Serge Demonchaux.
2936 NM/35 Bugzilla 39: Base64 decode bug fixes.
2937 Patch by Jakob Hirsch.
2939 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2941 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2943 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2945 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2951 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2952 ALT-N's libdkim that I have put here:
2953 http://duncanthrax.net/exim-experimental/
2955 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2956 were called 'true' and it seems that C99 defines that as a reserved
2957 keyword to be used with 'bool' variable types. That means you could
2958 not include C99-style headers which use bools without triggering
2959 build errors in sieve.c.
2961 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2962 as mailq or other aliases. Changed the --help handling significantly
2963 to do whats expected. exim_usage() emits usage/help information.
2965 SC/01 Added the -bylocaldomain option to eximstats.
2967 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2969 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2971 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2977 PH/01 Another patch from the Sieve maintainer.
2979 PH/02 When an IPv6 address is converted to a string for single-key lookup
2980 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2981 dots are used instead of colons so that keys in lsearch files need not
2982 contain colons. This was done some time before quoting was made available
2983 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2984 (notated using the quote facility) so as to distinguish them from IPv4
2985 keys. This meant that lookups for IP addresses in host lists did not work
2986 for iplsearch lookups.
2988 This has been fixed by arranging for IPv6 addresses to be expressed with
2989 colons if the lookup type is iplsearch. This is not incompatible, because
2990 previously such lookups could never work.
2992 The situation is now rather anomalous, since one *can* have colons in
2993 ordinary lsearch keys. However, making the change in all cases is
2994 incompatible and would probably break a number of configurations.
2996 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2999 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3000 conversion specification without a maximum field width, thereby enabling
3001 a rogue spamd server to cause a buffer overflow. While nobody in their
3002 right mind would setup Exim to query an untrusted spamd server, an
3003 attacker that gains access to a server running spamd could potentially
3004 exploit this vulnerability to run arbitrary code as the Exim user.
3006 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3007 $primary_hostname instead of what libspf2 thinks the hosts name is.
3009 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3010 a directory entry by the name of the lookup key. Previously, if a
3011 symlink pointed to a non-existing file or a file in a directory that
3012 Exim lacked permissions to read, a lookup for a key matching that
3013 symlink would fail. Now it is enough that a matching directory entry
3014 exists, symlink or not. (Bugzilla 503.)
3016 PH/03 The body_linecount and body_zerocount variables are now exported in the
3019 PH/04 Added the $dnslist_matched variable.
3021 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3022 This means they are set thereafter only if the connection becomes
3025 PH/06 Added the client_condition to authenticators so that some can be skipped
3026 by clients under certain conditions.
3028 PH/07 The error message for a badly-placed control=no_multiline_responses left
3029 "_responses" off the end of the name.
3031 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3033 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3034 (without spaces) instead of just copying the configuration text.
3036 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3038 PH/11 Added $max_received_linelength.
3040 PH/12 Added +ignore_defer and +include_defer to host lists.
3042 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3043 way in which PCRE > 7.0 is built.
3045 PH/14 Implemented queue_only_load_latch.
3047 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3048 MAIL command. The effect was to mangle the value on 64-bit systems.
3050 PH/16 Another patch from the Sieve maintainer.
3052 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3054 PH/18 If a system quota error occurred while trying to create the file for
3055 a maildir delivery, the message "Mailbox is full" was not appended to the
3056 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3057 only to a quota excession during the actual writing of the file.
3059 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3060 characters?) which causes problems in log lines. The DN values are now
3061 passed through string_printing() before being added to log lines.
3063 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3064 and InterBase are left for another time.)
3066 PH/21 Added message_body_newlines option.
3068 PH/22 Guard against possible overflow in moan_check_errorcopy().
3070 PH/23 POSIX allows open() to be a macro; guard against that.
3072 PH/24 If the recipient of an error message contained an @ in the local part
3073 (suitably quoted, of course), incorrect values were put in $domain and
3074 $local_part during the evaluation of errors_copy.
3080 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3081 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3084 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3085 issue a MAIL command.
3087 PH/02 In an ACL statement such as
3089 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3091 if a client was not listed at all, or was listed with a value other than
3092 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3093 the condition was not true (as it should be), so access was not denied.
3094 The bug was that the ! inversion was incorrectly passed on to the second
3095 item. This has been fixed.
3097 PH/03 Added additional dnslists conditions == and =& which are different from
3098 = and & when the dns lookup returns more than one IP address.
3100 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3101 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3103 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3104 FSYNC, which compiles an option called disable_fsync that allows for
3105 bypassing fsync(). The documentation is heavily laced with warnings.
3107 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3109 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3110 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3111 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3112 including adding "make clean"; (3) Added -fPIC when compiling the test
3113 dynamically loaded module, to get rid of a warning.
3115 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3116 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3117 = 0s. The bug is otherwise harmless.
3119 PH/07 There was a bug in the dovecot authenticator such that the value of
3120 $auth1 could be overwritten, and so not correctly preserved, after a
3121 successful authentication. This usually meant that the value preserved by
3122 the server_setid option was incorrect.
3124 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3126 PH/09 Installed PCRE release 7.0.
3128 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3129 run for batched SMTP input. It is now run at the start of every message
3130 in the batch. While fixing this I discovered that the process information
3131 (output by running exiwhat) was not always getting set for -bs and -bS
3132 input. This is fixed, and it now also says "batched" for BSMTP.
3134 PH/11 Added control=no_pipelining.
3136 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3137 patch, slightly modified), and move the expansion of helo_data till after
3138 the connection is made in the smtp transport (so it can use these
3141 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3143 PH/14 Added log_selector = +pid.
3145 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3147 PH/16 Add ${if forany and ${if forall.
3149 PH/17 Added dsn_from option to vary the From: line in DSNs.
3151 PH/18 Flush SMTP output before performing a callout, unless control =
3152 no_callout_flush is set.
3154 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3155 was true (the default) a successful delivery failed to delete the retry
3156 item, thus causing premature timeout of the address. The bug is now
3159 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3161 PH/21 Long custom messages for fakedefer and fakereject are now split up
3162 into multiline responses in the same way that messages for "deny" and
3163 other ACL rejections are.
3165 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3166 with slight modification.
3168 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3169 draft, changing the syntax and factoring some duplicate code".
3171 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3172 for deliveries of the second and subsequent messages over the same SMTP
3175 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3176 ${reduce, with only minor "tidies".
3178 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3180 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3181 expansion side effects.
3183 PH/27 When a message times out after an over-quota error from an Exim-imposed
3184 quota, the bounce message says "mailbox is full". This message was not
3185 being given when it was a system quota that was exceeded. It now should
3188 MH/03 Made $recipients available in local_scan(). local_scan() already has
3189 better access to the recipient list through recipients_list[], but
3190 $recipients can be useful in postmaster-provided expansion strings.
3192 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3193 in the case of a MAIL command with additional options following the
3194 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3195 were accidentally chopped off.
3197 PH/29 SMTP synchronization checks are implemented when a command is read -
3198 there is a check that no more input is waiting when there shouldn't be
3199 any. However, for some commands, a delay in an ACL can mean that it is
3200 some time before the response is written. In this time, more input might
3201 arrive, invalidly. So now there are extra checks after an ACL has run for
3202 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3203 pipelining has not been advertised.
3205 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3207 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3208 returned in the SMTP response when smtp_return_error_details was false.
3209 This has been fixed.
3211 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3212 instead of the C I/O that was originally supplied, because problems were
3213 reported on Solaris.
3215 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3216 Exim which did not show up earlier: it was assuming that a call to
3217 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3218 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3219 was a macro that became an assignment, so it seemed to work. This has
3220 changed to a proper function call with a void return, hence the compile
3221 error. Exim's code has been fixed.
3223 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3226 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3228 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3230 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3231 because it turns out to be too big or there is a timeout) there is no
3232 "Completed" line in the log. When some messages of this type were
3233 selected by exigrep, they were listed as "not completed". Others were
3234 picked up by some special patterns. I have improved the selection
3235 criteria to be more general.
3237 PH/38 The host_find_failed option in the manualroute router can now be set
3238 to "ignore", to completely ignore a host whose IP address cannot be
3239 found. If all hosts are ignored, the behaviour is controlled by the new
3240 host_all_ignored option.
3242 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3243 homing or because of multiple MX records with /mx) generated more than
3244 one IP address, and the following item turned out to be the local host,
3245 all the secondary addresses of the first item were incorrectly removed
3246 from the list, along with the local host and any following hosts (which
3247 is what is supposed to happen).
3249 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3250 whoever called Exim into the -H file. In the case of the daemon it was
3251 behaving confusingly. When first started, it used values for whoever
3252 started the daemon, but after a SIGHUP it used the Exim user (because it
3253 calls itself on a restart). I have changed the code so that it now always
3256 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3257 message are rejected with the same error (e.g. no authentication or bad
3258 sender address), and a DATA command is nevertheless sent (as can happen
3259 with PIPELINING or a stupid MUA), the error message that was given to the
3260 RCPT commands is included in the rejection of the DATA command. This is
3261 intended to be helpful for MUAs that show only the final error to their
3264 PH/42 Another patch from the Sieve maintainer.
3266 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3267 Eximstats - Fixed some broken HTML links and added missing column headers
3269 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3270 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3272 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3274 PH/43 Yet another patch from the Sieve maintainer.
3276 PH/44 I found a way to check for a TCP/IP connection going away before sending
3277 the response to the final '.' that terminates a message, but only in the
3278 case where the client has not sent further data following the '.'
3279 (unfortunately, this is allowed). However, in many cases there won't be
3280 any further data because there won't be any more messages to send. A call
3281 to select() can be used: if it shows that the input is "ready", there is
3282 either input waiting, or the socket has been closed. An attempt to read
3283 the next input character can distinguish the two cases. Previously, Exim
3284 would have sent an OK response which the client would never have see.
3285 This could lead to message repetition. This fix should cure that, at
3286 least in a lot of common cases.
3288 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3289 advertised in response to EHLO.
3295 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3296 fixed by 4.65/MH/01 (is this a record?) are fixed:
3298 (i) An empty string was always treated as zero by the numeric comparison
3299 operators. This behaviour has been restored.
3301 (ii) It is documented that the numeric comparison operators always treat
3302 their arguments as decimal numbers. This was broken in that numbers
3303 starting with 0 were being interpreted as octal.
3305 While fixing these problems I realized that there was another issue that
3306 hadn't been noticed. Values of message_size_limit (both the global option
3307 and the transport option) were treated as octal if they started with 0.
3308 The documentation was vague. These values are now always treated as
3309 decimal, and I will make that clear in the documentation.
3315 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3316 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3319 MH/01 Don't check that the operands of numeric comparison operators are
3320 integers when their expansion is in "skipping" mode (fixes bug
3321 introduced by 4.64-PH/07).
3323 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3324 child addresses, Exim now panics and dies. Previously, because the count
3325 is held in a short int, deliveries were likely to be lost. As such a
3326 large number of recipients for a single message is ridiculous
3327 (performance will be very, very poor), I have chosen to impose a limit
3328 rather than extend the field.
3334 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3335 leftover -K file (the existence of which was triggered by #402).
3336 While we were at it, introduced process PID as part of the -K
3337 filename. This should rule out race conditions when creating
3340 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3341 processing considerably. Previous code took too long for large mails,
3342 triggering a timeout which in turn triggers #401.
3344 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3345 in the DK code in transports.c. sendfile() is not really portable,
3346 hence the _LINUX specificness.
3348 TF/01 In the add_headers option to the mail command in an Exim filter,
3349 there was a bug that Exim would claim a syntax error in any
3350 header after the first one which had an odd number of characters
3353 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3354 callout verification, Exim cached a "reject" for the entire domain. This
3355 is correct for most verifications, but it is not correct for a recipient
3356 verification with use_sender or use_postmaster set, because in that case
3357 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3358 case of MAIL FROM:<> rejection from other early rejections (e.g.
3359 rejection of HELO). When verifying a recipient using a non-null MAIL
3360 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3361 Whatever the result of the callout, the value of the domain cache is
3362 left unchanged (for any other kind of callout, getting as far as trying
3363 RCPT means that the domain itself is ok).
3365 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3368 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3369 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3370 session, the daemon ignores ECONNECT errors and logs others; it now
3371 ignores EPIPE as well.
3373 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3374 (quoted-printable decoding).
3376 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3377 later the small subsequent patch to fix an introduced bug.
3379 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3381 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3383 PH/08 An error is now given if message_size_limit is specified negative.
3385 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3386 to be given (somewhat) arbitrary names.
3388 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3391 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3392 miscellaneous code fixes
3394 PH/10 Added the log_reject_target ACL modifier to specify where to log
3397 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3398 hostname. This is wrong, because it relates to the incoming message (and
3399 probably the interface on which it is arriving) and not to the outgoing
3400 callout (which could be using a different interface). This has been
3401 changed to use the value of the helo_data option from the smtp transport
3402 instead - this is what is used when a message is actually being sent. If
3403 there is no remote transport (possible with a router that sets up host
3404 addresses), $smtp_active_hostname is used.
3406 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3407 tweaks were necessary in order to get it to work (see also 21 below):
3408 (a) The code assumed that strncpy() returns a negative number on buffer
3409 overflow, which isn't the case. Replaced with Exim's string_format()
3411 (b) There were several signed/unsigned issues. I just did the minimum
3412 hacking in of casts. There is scope for a larger refactoring.
3413 (c) The code used strcasecmp() which is not a standard C function.
3414 Replaced with Exim's strcmpic() function.
3415 (d) The code set only $1; it now sets $auth1 as well.
3416 (e) A simple test gave the error "authentication client didn't specify
3417 service in request". It would seem that Dovecot has changed its
3418 interface. Fortunately there's a specification; I followed it and
3419 changed what the client sends and it appears to be working now.
3421 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3424 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3425 address A is aliased to B and C, where B exists and C does not. Without
3426 -v the output is "A verified" because verification stops after a
3427 successful redirection if more than one address is generated. However,
3428 with -v the child addresses are also verified. Exim was outputting "A
3429 failed to verify" and then showing the successful verification for C,
3430 with its parentage. It now outputs "B failed to verify", showing B's
3431 parentage before showing the successful verification of C.
3433 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3434 look up a TXT record in a specific list after matching in a combined
3437 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3438 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3439 they consult the DNS. I had assumed they would set it the way they
3440 wanted; and indeed my experiments on Linux seem to show that in some
3441 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3442 To be on the safe side, however, I have now made the interface to
3443 host_find_byname() similar to host_find_bydns(), with an argument
3444 containing the DNS resolver options. The host_find_byname() function now
3445 sets these options at its start, just as host_find_bydns() does. The smtp
3446 transport options dns_qualify_single and dns_search_parents are passed to
3447 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3448 of host_find_byname() use the default settings of RES_DEFNAMES
3449 (qualify_single) but not RES_DNSRCH (search_parents).
3451 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3452 spool_read_header() do less string testing, by means of a preliminary
3453 switch on the second character of optional "-foo" lines. (This is
3454 overdue, caused by the large number of possibilities that now exist.
3455 Originally there were few.) While I was there, I also converted the
3456 str(n)cmp tests so they don't re-test the leading "-" and the first
3457 character, in the hope this might squeeze out yet more improvement.
3459 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3460 flag allowing group syntax was set by the header_syntax check but not
3461 turned off, possible causing trouble later; (2) The flag was not being
3462 set at all for the header_verify test, causing "group"-style headers to
3463 be rejected. I have now set it in this case, and also caused header_
3464 verify to ignore an empty address taken from a group. While doing this, I
3465 came across some other cases where the code for allowing group syntax
3466 while scanning a header line wasn't quite right (mostly, not resetting
3467 the flag correctly in the right place). These bugs could have caused
3468 trouble for malformed header lines. I hope it is now all correct.
3470 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3471 with the "reply" argument non-NULL. The code, however (which originally
3472 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3473 but it didn't always do it. This confused somebody who was copying the
3474 code for some other use. I have removed all the tests.
3476 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3477 feature that was used to support insecure browsers during the U.S. crypto
3478 embargo. It requires special client support, and Exim is probably the
3479 only MTA that supported it -- and would never use it because real RSA is
3480 always available. This code has been removed, because it had the bad
3481 effect of slowing Exim down by computing (never used) parameters for the
3482 RSA_EXPORT functionality.
3484 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3485 authenticator to fail if there's a tab character in the incoming data
3486 (there should never be unless someone is messing about, as it's supposed
3487 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3488 if the connection is using TLS or if the remote IP is the same as the
3489 local IP, and the "valid-client-cert option" if a client certificate has
3492 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3493 authenticators. This can be used for authorization after authentication
3494 succeeds. (In the case of plaintext, it servers for both authentication
3497 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3498 if any retry times were supplied.
3500 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3501 connection, where there is no client IP address to check. In this
3502 situation, the verify now always succeeds.
3504 PH/25 Applied John Jetmore's -Mset patch.
3506 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3508 PH/27 In a string expansion for a processed (not raw) header when multiple
3509 headers of the same name were present, leading whitespace was being
3510 removed from all of them, but trailing whitespace was being removed only
3511 from the last one. Now trailing whitespace is removed from each header
3512 before concatenation. Completely empty headers in a concatenation (as
3513 before) are ignored.
3515 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3516 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3518 PH/29 [Removed. This was a change that I later backed out, and forgot to
3519 correct the ChangeLog entry (that I had efficiently created) before
3520 committing the later change.]
3522 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3523 address errors (4xx response to RCPT) over the same connection as other
3524 messages routed to the same hosts. Such deliveries are always "forced",
3525 so retry times are not inspected. This resulted in far too many retries
3526 for the affected addresses. The effect occurred only when there were more
3527 hosts than the hosts_max_try setting in the smtp transport when it had
3528 the 4xx errors. Those hosts that it had tried were not added to the list
3529 of hosts for which the message was waiting, so if all were tried, there
3530 was no problem. Two fixes have been applied:
3532 (i) If there are any address or message errors in an SMTP delivery, none
3533 of the hosts (tried or untried) are now added to the list of hosts
3534 for which the message is waiting, so the message should not be a
3535 candidate for sending over the same connection that was used for a
3536 successful delivery of some other message. This seems entirely
3537 reasonable: after all the message is NOT "waiting for some host".
3538 This is so "obvious" that I'm not sure why it wasn't done
3539 previously. Hope I haven't missed anything, but it can't do any
3540 harm, as the worst effect is to miss an optimization.
3542 (ii) If, despite (i), such a delivery is accidentally attempted, the
3543 routing retry time is respected, so at least it doesn't keep
3544 hammering the server.
3546 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3547 in ${readsocket because some servers need this prod.
3549 PH/32 Added some extra debug output when updating a wait-xxx database.
3551 PH/33 The hint "could be header name not terminated by colon", which has been
3552 given for certain expansion errors for a long time, was not being given
3553 for the ${if def:h_colon_omitted{... case.
3555 PH/34 The spec says: "With one important exception, whenever a domain list is
3556 being scanned, $domain contains the subject domain." There was at least
3557 one case where this was not true.
3559 PH/35 The error "getsockname() failed: connection reset by peer" was being
3560 written to the panic log as well as the main log, but it isn't really
3561 panic-worthy as it just means the connection died rather early on. I have
3562 removed the panic log writing for the ECONNRESET error when getsockname()
3565 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3566 runs only) independently of the message's sender address. This meant
3567 that, if the 4xx error was in fact related to the sender, a different
3568 message to the same recipient with a different sender could confuse
3569 things. In particular, this can happen when sending to a greylisting
3570 server, but other circumstances could also provoke similar problems.
3571 I have changed the default so that the retry time for these errors is now
3572 based a combination of the sender and recipient addresses. This change
3573 can be overridden by setting address_retry_include_sender=false in the
3576 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3577 remote server are returned as part of bounce messages. This was not
3578 happening for LMTP over a pipe (the lmtp transport), but now it is the
3579 same for both kinds of LMTP.
3581 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3582 in header lines that were in fact CNAMEs. This is no longer the case.
3584 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3585 and queue runs started by the daemon processed all messages. This has
3586 been fixed so that -R and -S can now usefully be given with -q<time>.
3588 PH/40 Import PCRE release 6.7 (fixes some bugs).
3590 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3592 PH/42 Give an error if -q is specified more than once.
3594 PH/43 Renamed the variables $interface_address and $interface_port as
3595 $received_ip_address and $received_port, to make it clear that these
3596 values apply to message reception, and not to the outgoing interface when
3597 a message is delivered. (The old names remain recognized, of course.)
3599 PH/44 There was no timeout on the connect() call when using a Unix domain
3600 socket in the ${readsocket expansion. There now is.
3602 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3603 be meaningful with "accept".
3605 SC/01 Eximstats V1.43
3606 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3608 SC/02 Eximstats V1.44
3609 Use a glob alias rather than an array ref in the generated
3610 parser. This improves both readability and performance.
3612 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3613 Collect SpamAssassin and rejection statistics.
3614 Don't display local sender or destination tables unless
3615 there is data to show.
3616 Added average volumes into the top table text output.
3618 SC/04 Eximstats V1.46
3619 Collect data on the number of addresses (recipients)
3620 as well as the number of messages.
3622 SC/05 Eximstats V1.47
3623 Added 'Message too big' to the list of mail rejection
3624 reasons (thanks to Marco Gaiarin).
3626 SC/06 Eximstats V1.48
3627 Mainlog lines which have GMT offsets and are too short to
3628 have a flag are now skipped.
3630 SC/07 Eximstats V1.49 (Alain Williams)
3631 Added the -emptyok flag.
3633 SC/08 Eximstats V1.50
3634 Fixes for obtaining the IP address from reject messages.
3636 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3637 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3638 whitespace changes from 4.64-PH/27
3640 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3643 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3644 are found, allow negative numbers in numeric criteria)
3646 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3648 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3649 to match changes made in 4.64-PH/43
3651 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3653 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3654 "domain" is actually an IP address, and force a failure. This locks out
3655 those revolvers/nameservers that support "A-for-A" lookups, in
3656 contravention of the specifications.
3658 PH/48 When a host name was looked up from an IP address, and the subsequent
3659 forward lookup of the name timed out, the host name was left in
3660 $sender_host_name, contrary to the specification.
3662 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3663 restricted to single-key lookups, Exim was not diagnosing an error if
3664 * or *@ was used with a query-style lookup.
3666 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3668 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3669 long ago, but noone interested enough thought of it. Let's just say that
3670 the "1.1" means that there are some new functions that weren't there at
3671 some point in the past.
3673 PH/51 Error processing for expansion failure of helo_data from an smtp
3674 transport during callout processing was broken.
3676 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3677 tested/used via the -bh/-bhc/-bs options.
3679 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3680 bug, fixed in subsequent PCRE releases).
3682 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3683 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3685 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3691 SC/01 Use a glob alias rather than an array ref in eximstats generated
3692 parser. This improves both readability and performance.
3694 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3695 Don't display local sender or destination tables in eximstats unless
3696 there is data to show.
3697 Added average volumes into the eximstats top table text output.
3699 SC/03 Collect data on the number of addresses (recipients) as well
3700 as the number of messages in eximstats.
3702 TF/01 Correct an error in the documentation for the redirect router. Exim
3703 does (usually) call initgroups() when daemonizing.
3705 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3706 with consistent privilege compared to when running as a daemon.
3708 TF/03 Note in the spec that $authenticated_id is not set for local
3709 submissions from trusted users.
3711 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3712 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3714 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3715 by adding some example configuration directives to the default
3716 configuration file. A little bit of work is required to uncomment the
3717 directives and define how usernames and passwords are checked, but
3718 there is now a framework to start from.
3720 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3721 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3722 without this. I don't know how relevant this is to other LDAP libraries.
3724 PH/02 Add the verb name to the "unknown ACL verb" error.
3726 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3728 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3730 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3731 directory not expanded when it should be if an expanded home directory
3732 was set for the address (which is overridden by the transport).
3734 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3737 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3738 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3739 because it is too late at that time, and has no effect.
3741 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3742 security issue with \' (bugzilla #107). I could not use the
3743 PQescapeStringConn() function, because it needs a PGconn value as one of
3746 PH/08 When testing addresses using -bt, indicate those final addresses that
3747 are duplicates that would not cause an additional delivery. At least one
3748 person was confused, thinking that -bt output corresponded to deliveries.
3749 (Suppressing duplicates isn't a good idea as you lose the information
3750 about possibly different redirections that led to the duplicates.)
3752 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3753 systems where poll() doesn't work, in particular OS X.
3755 PH/10 Added more information to debugging output for retry time not reached.
3757 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3758 operations in malware.c.
3760 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3763 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3764 syslog_duplication set false, log lines that would normally be written
3765 both the the main log and to the reject log were not written to syslog at
3768 PH/14 In the default configuration, change the use of "message" in ACL warn
3769 statements to "add_header".
3771 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3772 not followed by a command (e.g. "seen endif").
3774 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3775 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3778 PH/17 Added extra conditions to the default value of delay_warning_condition
3782 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3783 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3784 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3787 The Auto-Submitted: and various List- headers are standardised, whereas I
3788 don't think Precedence: ever was.
3790 PH/18 Refactored debugging code in route_finduser() to show more information,
3791 in particular, the error code if getpwnam() issues one.
3793 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3794 This is apparently needed in addition to the PH/07 change above to avoid
3795 any possible encoding problems.
3797 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3798 but not after initializing Perl.
3800 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3801 output them only if debugging. By default they are written stderr,
3802 apparently, which is not desirable.
3804 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3807 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3810 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3812 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3813 authenticated or an ident call has been made. Suppress the default
3814 values for $authenticated_id and $authenticated_sender (but permit -oMai
3815 and -oMas) when testing with -bh.
3817 PH/24 Re-jigged the order of the tests in the default configuration so that the
3818 tests for valid domains and recipients precede the DNS black list and CSA
3819 tests, on the grounds that those ones are more expensive.
3821 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3822 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3823 This bug exists in every version of Exim that I still have, right back to
3826 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3827 However, an attempt to turn on case-sensitivity in a regex key by
3828 including (?-i) didn't work because the subject string was already
3829 lowercased, and the effects were non-intuitive. It turns out that a
3830 one-line patch can be used to allow (?-i) to work as expected.
3836 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3837 other effects) broke the use of negated acl sub-conditions.
3839 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3842 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3843 "Deny" causes Exim to reject the incoming connection with a 554 error.
3844 Unfortunately, if there is a major crisis, such as a disk failure,
3845 tcp-wrappers gives "deny", whereas what one would like would be some
3846 kind of temporary error. A kludge has been added to help with this.
3847 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3848 554 error is used if errno is still zero or contains ENOENT (which occurs
3849 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3852 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3854 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3855 errors. Otherwise a message that provokes a temporary error (when other
3856 messages do not) can cause a whole host to time out.
3858 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3859 addresses were routed directly to files or pipes from a redirect router.
3860 File deliveries just didn't batch; pipe deliveries might have suffered
3863 PH/06 A failure to get a lock for a hints database would erroneously always say
3864 "Failed to get write lock", even when it was really a read lock.
3866 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3867 of 0600. This has been changed to use the value of the lockfile_mode
3868 option (which defaults to 0600).
3870 PH/08 Applied small patch from the Sieve maintainer.
3872 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3873 folder from quota calculations, a direct delivery into this folder messed
3874 up the contents of the maildirsize file. This was because the regex was
3875 used only to exclude .Trash (or whatever) when the size of the mailbox
3876 was calculated. There was no check that a delivery was happening into an
3877 excluded directory. This bug has been fixed by ignoring all quota
3878 processing for deliveries into excluded directories.
3880 PH/10 Added the maildirfolder_create_regex option to appendfile.
3886 PH/01 The code for finding all the local interface addresses on a FreeBSD
3887 system running IPv6 was broken. This may well have applied to all BSD
3888 systems, as well as to others that have similar system calls. The broken
3889 code found IPv4 interfaces correctly, but gave incorrect values for the
3890 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3891 that it would not match correctly against @[] and not recognize the IPv6
3894 PH/02 The ipliteral router was not recognizing addresses of the form user@
3895 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3897 PH/03 Added disable_ipv6.
3899 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3900 decoded form, because it is most often used to construct To: headers
3901 lines in autoreplies, and the decoded form may well be syntactically
3902 invalid. However, $reply_address has leading white space removed, and all
3903 newlines turned into spaces so that the autoreply transport does not
3906 PH/05 If group was specified without a user on a router, and no group or user
3907 was specified on a transport, the group from the router was ignored.
3909 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3910 for visible compile-time settings that can be used to change these
3911 numbers, for those that want even more. Backwards compatibility with old
3912 spool files has been maintained. However, going back to a previous Exim
3913 release will lost any variables that are in spool files.
3915 PH/07 Two small changes when running in the test harness: increase delay when
3916 passing a TCP/IP connection to a new process, in case the original
3917 process has to generate a bounce, and remove special handling of
3918 127.0.0.2 (sic), which is no longer necessary.
3920 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3921 be the same on different OS.
3923 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3926 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3927 whether --show-vars was specified or not
3929 JJ/02 exipick: Added support for new ACL variable spool format introduced
3932 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3933 syntactically invalid From: or Reply-to: line, and a filter used this to
3934 generate an autoreply, and therefore failed to obtain an address for the
3935 autoreply, Exim could try to deliver to a non-existent relative file
3936 name, causing unrelated and misleading errors. What now happens is that
3937 it logs this as a hard delivery error, but does not attempt to create a
3940 PH/11 The exinext utility has a -C option for testing purposes, but although
3941 the given file was scanned by exinext itself; it wasn't being passed on
3942 when Exim was called.
3944 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3945 an end-of-file indication when reading a command response.
3947 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3948 compiled. In many other places in Exim, IPv6 addresses are always
3949 recognized, so I have changed this. It also means that IPv4 domain
3950 literals of the form [IPV4:n.n.n.n] are now always recognized.
3952 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3953 used if the router is not running as root, for example, when verifying at
3954 ACL time, or when using -bh. The debugging output from this situation was
3955 non-existent - all you got was a failure to exec. I have made two
3958 (a) Failures to set uid/gid, the current directory, or a process leader
3959 in a subprocess such as that created by queryprogram now generate
3960 suitable debugging output when -d is set.
3962 (b) The queryprogram router detects when it is not running as root,
3963 outputs suitable debugging information if -d is set, and then runs
3964 the subprocess without attempting to change uid/gid.
3966 PH/15 Minor change to Makefile for building test_host (undocumented testing
3969 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3970 additional section of a DNS packet that returns MX or SRV records.
3971 Instead, it always explicitly searches for A/AAAA records. This avoids
3972 major problems that occur when a DNS server includes only records of one
3973 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3974 fixed another bug: if SRV records were looked up and the corresponding
3975 address records were *not* found in the additional section, the port
3976 values from the SRV records were lost.
3978 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3979 using the correct key (the original address) when searching the retry
3980 rules in order to find which one to use for generating the retry hint.
3982 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3983 adding the default one. Similarly, if it contains a Reply-To: header, the
3984 errors_reply_to option, if set, is not used.
3986 PH/19 When calculating a retry time, Exim used to measure the "time since
3987 failure" by looking at the "first failed" field in the retry record. Now
3988 it does not use this if it is later than than the arrival time of the
3989 message. Instead it uses the arrival time. This makes for better
3990 behaviour in cases where some deliveries succeed, thus re-setting the
3991 "first failed" field. An example is a quota failure for a huge message
3992 when small messages continue to be delivered. Without this change, the
3993 "time since failure" will always be short, possible causing more frequent
3994 delivery attempts for the huge message than are intended.
3995 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3997 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3998 $1, $2, $3) because the numerical variables can be reset during some
3999 expansion items (e.g. "match"), thereby losing the authentication data.
4001 PH/21 Make -bV show the size of off_t variables so that the test suite can
4002 decide whether to run tests for quotas > 2G.
4004 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4005 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4006 filecount value is greater than 2G or if a quota value is greater than 2G
4007 on a system where the size of off_t is not greater than 4, a panic error
4010 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4011 never match. The debug and -bh output now contains an explicit error
4012 message indicating a malformed IPv4 address or mask.
4014 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4015 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4016 PH/23 above applies.
4018 PH/25 Do not write to syslog when running in the test harness. The only
4019 occasion when this arises is a failure to open the main or panic logs
4020 (for which there is an explicit test).
4022 PH/26 Added the /no_tell option to "control=freeze".
4024 PH/27 If a host name lookup failed very early in a connection, for example, if
4025 the IP address matched host_lookup and the reverse lookup yielded a name
4026 that did not have a forward lookup, an error message of the form "no IP
4027 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4028 could be logged. Now it outputs the IP address instead of "NULL".
4030 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4031 allows the sender and the authenticated sender to be set when
4032 submitting a message from within Exim. Since child_open_exim() is
4033 documented for local_scan(), the new function should be too.
4035 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4036 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4037 results in an empty string is now treated as unset.
4039 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4041 PH/31 Added sender_verify_fail logging option.
4043 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4044 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4045 tidied the source and removed it altogether.
4047 PH/33 When a queue run was abandoned because the load average was too high, a
4048 log line was always written; now it is written only if the queue_run log
4049 selector is set. In addition, the log line for abandonment now contains
4050 information about the queue run such as the pid. This is always present
4051 in "start" and "stop" lines but was omitted from the "abandon" line.
4053 PH/34 Omit spaces between a header name and the colon in the error message that
4054 is given when verify = headers_syntax fails (if there are lots of them,
4055 the message gets confusing).
4057 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4058 names, as there are now some PTR records that contain slashes. This check
4059 is only to protect against broken name servers that fall over on strange
4060 characters, so the fact that it applies to all lookups doesn't matter.
4062 PH/36 Now that the new test suite is complete, we can remove some of the
4063 special code in Exim that was needed for the old test suite. For example,
4064 sorting DNS records because real resolvers return them in an arbitrary
4065 order. The new test suite's fake resolver always returns records in the
4068 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4069 bounces) except when queue_only is set, to avoid logging races between
4070 the different processes.
4072 PH/38 Panic-die if .include specifies a non-absolute path.
4074 PH/39 A tweak to the "H" retry rule from its user.
4076 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4077 a label. They prevented compilation on older perls.
4079 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4080 a warning to be raised on newish perls.
4082 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4083 on queue. Changes to match documented behaviour of showing count of
4084 messages matching specified criteria.
4086 PH/40 Changed the default ident timeout from 30s to 5s.
4088 PH/41 Added support for the use of login_cap features, on those BSD systems
4089 that have them, for controlling the resources used by pipe deliveries.
4091 PH/42 The content-scanning code uses fopen() to create files in which to put
4092 message data. Previously it was not paying any attention to the mode of
4093 the files. Exim runs with umask(0) because the rest of the code creates
4094 files with open(), and sets the required mode explicitly. Thus, these
4095 files were ending up world-writeable. This was not a big issue, because,
4096 being within the spool directory, they were not world-accessible. I have
4097 created a function called modefopen, which takes an additional mode
4098 argument. It sets umask(777), creates the file, chmods it to the required
4099 mode, then resets the umask. All the relevant calls to fopen() in the
4100 content scanning code have been changed to use this function.
4102 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4103 to 24 hours. This avoids potential overflow problems when processing G
4104 and H retry rules. I suspect nobody ever tinkers with this value.
4106 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4108 PH/45 When the plaintext authenticator is running as a client, the server's
4109 challenges are checked to ensure they are valid base64 strings. By
4110 default, the authentication attempt is cancelled if an invalid string is
4111 received. Setting client_ignore_invalid_base64 true ignores these errors.
4112 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4113 they are received. Thus, the responses can be made to depend on the
4114 challenges. If an invalid string is ignored, an empty string is placed in
4117 PH/46 Messages that are created by the autoreply transport now contains a
4118 References: header, in accordance with RFCs 2822 and 3834.
4120 PH/47 Added authenticated_sender_force to the smtp transport.
4122 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4124 PH/49 Installed latest patch from the Sieve maintainer.
4126 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4127 was also set, the appendfile transport was unnecessarily scanning a
4128 directory of message files (e.g. for maildir delivery) to find the count
4129 of files (along with the size), even though it did not need this
4130 information. It now does the scan only if it needs to find either the
4131 size of the count of files.
4133 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4135 PH/52 Two bugs concerned with error handling when the smtp transport is
4138 (i) Exim was not creating retry information for temporary errors given
4139 for individual recipients after the DATA command when the smtp transport
4140 was used in LMTP mode. This meant that they could be retried too
4141 frequently, and not timed out correctly.
4143 (ii) Exim was setting the flag that allows error details to be returned
4144 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4145 recipients that were returned after the DATA command.
4147 PH/53 This is related to PH/52, but is more general: for any failing address,
4148 when detailed error information was permitted to be returned to the
4149 sender, but the error was temporary, then after the final timeout, only
4150 "retry timeout exceeded" was returned. Now it returns the full error as
4151 well as "retry timeout exceeded".
4153 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4154 do this, and (what is worse) MTAs that accept it.
4156 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4157 will now be deprecated.
4159 PH/56 New os.c-cygwin from the Cygwin maintainer.
4161 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4162 formats (previously only available in exim formats via -bpr, -bpru,
4163 and -bpra. Now also available in native and exiqgrep formats)
4165 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4166 with very large, slow to parse queues
4168 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4170 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4172 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4173 responses to authentication challenges, though it was showing the
4174 challenges; (ii) I've removed the CR characters from the debug output for
4177 PH/58 Allow for the insertion of a newline as well as a space when a string
4178 is turned into more than one encoded-word during RFC 2047 encoding. The
4179 Sieve code now uses this.
4181 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4182 data_4xx, lost_connection, tls_required.
4184 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4185 message was being sent as an SMTP response.
4187 PH/61 Add -l and -k options to exicyclog.
4189 PH/62 When verifying, if an address was redirected to one new address, so that
4190 verification continued, and the new address failed or deferred after
4191 having set something in $address_data, the value of $address_data was not
4192 passed back to the ACL. This was different to the case when no
4193 redirection occurred. The value is now passed back in both cases.
4195 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4196 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4197 use login_cap.h, so on its own it isn't the distinguishing feature. The
4198 new name refers directly to the setclassresources() function.
4200 PH/65 Added configuration files for NetBSD3.
4202 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4204 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4205 is preferred over IPv4.
4207 PH/68 The bounce_return_message and bounce_return_body options were not being
4208 honoured for bounces generated during the reception of non-SMTP messages.
4209 In particular, this applied to messages rejected by the ACL. This bug has
4210 been fixed. However, if bounce_return_message is true and bounce_return_
4211 body is false, the headers that are returned for a non-SMTP message
4212 include only those that have been read before the error was detected.
4213 (In the case of an ACL rejection, they have all been read.)
4215 PH/69 The HTML version of the specification is now built in a directory called
4216 spec_html instead of spec.html, because the latter looks like a path with
4217 a MIME-type, and this confuses some software.
4219 PH/70 Catch two compiler warnings in sieve.c.
4221 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4222 function verify_get_ident() calls ip_connect() to connect a socket, but
4223 if the "connect()" function timed out, ip_connect() used to close the
4224 socket. However, verify_get_ident() also closes the socket later, and in
4225 between Exim writes to the log, which may get opened at this point. When
4226 the socket was closed in ip_connect(), the log could get the same file
4227 descriptor number as the socket. This naturally causes chaos. The fix is
4228 not to close the socket in ip_connect(); the socket should be closed by
4229 the function that creates it. There was only one place in the code where
4230 this was missing, in the iplookup router, which I don't think anybody now
4231 uses, but I've fixed it anyway.
4233 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4234 well as to direct DNS lookups. Otherwise the handling of names in host
4235 lists is inconsistent and therefore confusing.
4241 PH/01 Two changes to the default runtime configuration:
4243 (1) Move the checks for relay_from_hosts and authenticated clients from
4244 after to before the (commented out) DNS black list checks.
4246 (2) Add control=submission to the relay_from_hosts and authenticated
4247 clients checks, on the grounds that messages accepted by these
4248 statements are most likely to be submissions.
4250 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4252 (1) Generate an error if the third argument for the ${prvs expansion is
4255 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4258 (3) Reset the variables that are obtained from the first argument of
4259 ${prvscheck and used in the second argument before leaving the code,
4260 because their memory is reclaimed, so using them afterwards may do
4263 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4264 one (it's much easier than Tom thought :-).
4266 (5) Because of (4), we can now allow for the use of $prvscheck_result
4267 inside the third argument.
4269 PH/03 For some reason, the default setting of PATH when running a command from
4270 a pipe transport was just "/usr/bin". I have changed it to
4273 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4274 anything to be listed in the output from -bV.
4276 PH/05 When a filter generated an autoreply, the entire To: header line was
4277 quoted in the delivery log line, like this:
4279 => >A.N.Other <ano@some.domain> <original@ddress> ...
4281 This has been changed so that it extracts the operative address. There
4282 may be more than one such address. If so, they are comma-separated, like
4285 => >ano@some.domain,ona@other.domain <original@ddress> ...
4287 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4288 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4289 not being looked up in the rDNS to get a host name, Exim was showing the
4290 IP address twice in Received: lines, even though the IP addresses were
4291 identical. For example:
4293 Received: from [1.2.3.4] (helo=[1.2.3.4])
4295 However, if the real host name was known, it was omitting the HELO data
4296 if it matched the actual IP address. This has been tidied up so that it
4297 doesn't show the same IP address twice.
4299 PH/07 When both +timestamp and +memory debugging was on, the value given by
4300 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4301 called by the debug printing, thereby overwriting the timestamp buffer.
4302 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4304 PH/08 When the original message was included in an autoreply transport, it
4305 always said "this is a copy of the message, including all the headers",
4306 even if body_only or headers_only was set. It now gives an appropriate
4309 PH/09 Applied a patch from the Sieve maintainer which:
4311 o fixes some comments
4312 o adds the (disabled) notify extension core
4313 o adds some debug output for the result of if/elsif tests
4314 o points to the current vacation draft in the documentation
4315 and documents the missing references header update
4319 o fixes a bug in processing the envelope test (when testing
4320 multiple envelope elements, the last element determined the
4323 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4324 Electronic Mail") by including:
4326 Auto-submitted: auto-generated
4328 in the messages that it generates (bounce messages and others, such as
4329 warnings). In the case of bounce messages for non-SMTP messages, there was
4330 also a typo: it was using "Auto_submitted" (underscore instead of
4331 hyphen). Since every message generated by Exim is necessarily in response
4332 to another message, thes have all been changed to:
4334 Auto-Submitted: auto-replied
4336 in accordance with these statements in the RFC:
4338 The auto-replied keyword:
4340 - SHOULD be used on messages sent in direct response to another
4341 message by an automatic process,
4343 - MUST NOT be used on manually-generated messages,
4345 - MAY be used on Delivery Status Notifications (DSNs) and Message
4346 Disposition Notifications (MDNs),
4348 - MUST NOT be used on messages generated by automatic or periodic
4349 processes, except for messages which are automatic responses to
4352 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4353 to the default Received: header definition.
4355 PH/12 Added log selector acl_warn_skipped (default on).
4357 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4358 variables because (a) they are in the wrong storage pool and (b) even if
4359 they were copied, it wouldn't work properly because of the caching.
4361 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4362 checking when decoding. Apparently there are clients that generate
4363 overlong encoded strings. Why am I not surprised?
4365 PH/15 If the first argument of "${if match_address" was not empty, but did not
4366 contain an "@" character, Exim crashed. Now it writes a panic log message
4367 and treats the condition as false.
4369 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4371 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4372 extension "envelope-auth". The code is finished and in agreement with
4373 other implementations, but there is no documentation so far and in fact,
4374 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4375 not changing the active code.
4377 Print executed "if" and "elsif" statements when debugging is used. This
4378 helps a great deal to understand what a filter does.
4380 Document more things not specified clearly in RFC3028. I had all this
4381 sorted out, when out of a sudden new issues came to my mind. Oops."
4383 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4386 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4387 canonical form (as documented). However, after a host name lookup from
4388 the IP address, check_host() was doing a simple string comparison with
4389 addresses acquired from the DNS when checking that the found name did
4390 have the original IP as one of its addresses. Since any found IPv6
4391 addresses are likely to be in abbreviated form, the comparison could
4392 fail. Luckily, there already exists a function for doing the comparison
4393 by converting both addresses to binary, so now that is used instead of
4394 the text comparison.
4396 PH/20 There was another similar case to PH/19, when a complete host name was
4397 given in a host list; looking up its IP address could give an abbreviated
4398 form, whereas the current host's name might or might not be abbreviated.
4399 The same fix has been applied.
4405 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4406 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4409 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4410 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4412 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4414 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4415 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4416 Also, for "Auto-Submitted", treat anything other than "no" as
4417 non-personal, in accordance with RFC 3834. (Previously it treated
4418 anything starting "auto-" as non-personal.)
4420 TF/01 The control=submission/name=... option had a problem with syntax
4421 errors if the name included a slash character. The /name= option
4422 now slurps the rest of the string, so it can include any characters
4423 but it must come last in the list of options (after /sender_retain
4426 PH/05 Some modifications to the interface to the fake nameserver for the new
4434 TK/01 Added the "success_on_redirect" address verification option. See
4435 NewStuff for rationale and an example.
4437 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4439 PH/02 Patch to exigrep to allow it to work on syslog lines.
4441 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4442 fread() to skip over the body file's header line, because in Cygwin the
4443 header line is locked and is inaccessible.
4445 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4446 co-exist for some time) to make it clear that it is the Exim ID that is
4447 referenced, not the Message-ID: header line.
4449 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4450 string_format() function, because snprintf() does not exist on all
4453 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4454 ${stat: expansion item.
4456 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4457 protocol synchronization error", to keep the pedants happy.
4459 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4460 well as for IRIX systems, when gcc is being used. See the host.c source
4463 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4465 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4468 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4469 transport and to the smtp transport in LMTP mode.
4471 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4473 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4474 run a filter in a subprocess. This could lead to confusion in subsequent
4475 lookups in the parent process. There should also be a search_tidyup() at
4476 the end of the subprocess.
4478 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4479 only if the host matched helo_try_verify_hosts, which caused the
4480 verification to occur when the EHLO/HELO command was issued. The ACL just
4481 tested the remembered result. Now, if a previous verification attempt has
4482 not happened, "verify = helo" does it there and then.
4484 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4486 TK/03 Fix log output including CR from clamd.
4488 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4489 exist provoked a memory error which could cause a segfault.
4491 PH/15 Installed PCRE 6.2
4493 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4495 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4496 of the problem. Specifically, suggested +O2 rather than +O1 for the
4499 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4501 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4502 record, the port was not correctly logged when the outgoing_port log
4503 selector was set (it logged the transort's default port).
4505 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4506 fallback_hosts, and "hosts" in the smtp transport.
4508 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4509 host errors such as "Connection refused".
4511 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4512 authentication with radiusclient 0.4.9:
4514 - Error returned from rc_read_config was caught wrongly
4515 - Username/password not passed on to radius server due to wrong length.
4517 The presumption is that some radiusclient API changes for 4.51/PH/17
4518 were not taken care of correctly. The code is still untested by me (my
4519 Linux distribution still has 0.3.2 of radiusclient), but it was
4520 contributed by a Radius user.
4522 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4523 expanding the "port" option of the smtp transport.
4525 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4526 while reading a MIME header. Thanks to Tom Hughes for a patch.
4528 PH/24 Include config.h inside local_scan.h so that configuration settings are
4531 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4532 that in an ACL for RCPT (for example), you can examine exactly what was
4535 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4536 commands, but it was not correctly comparing the address with the actual
4537 client host address. Thus, it would show the EHLO address in Received:
4538 header lines when this was not necessary.
4540 PH/27 Added the % operator to ${eval:}.
4542 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4543 it should be ignoring failures (because with -C, for example, it has lost
4544 privilege). It wasn't ignoring creation failures other than "already
4547 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4550 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4551 because an input line was too long, either on its own, or by virtue of
4552 too many continuations, the temporary file was not being removed, and the
4553 return code was incorrect.
4555 PH/31 Missing "BOOL" in function definition in filtertest.c.
4557 PH/32 Applied Sieve patches from the maintainer.
4559 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4561 PH/33 Added "verify = not_blind".
4563 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4564 Local/Makefile (with some defaults set). These are used in built scripts
4565 such as exicyclog, but they have never been used in the exim_install
4566 script (though there are many overriding facilities there). I have
4567 arranged that the exim_install script now takes note of these two
4570 PH/35 Installed configuration files for Dragonfly.
4572 PH/36 When a locally submitted message by a trusted user did not contain a
4573 From: header, and the sender address was obtained from -f or from an SMTP
4574 MAIL command, and the trusted user did not use -F to supply a sender
4575 name, $originator_name was incorrectly used when constructing a From:
4576 header. Furthermore, $originator_name was used for submission mode
4577 messages from external hosts without From: headers in a similar way,
4578 which is clearly wrong.
4580 PH/37 Added control=suppress_local_fixups.
4582 PH/38 When log_selector = +received_sender was set, and the addition of the
4583 sender made the log line's construction buffer exactly full, or one byte
4584 less than full, an overflow happened when the terminating "\n" was
4587 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4588 when the result of a list match is failure because a DNS lookup failed.
4590 PH/40 RM_COMMAND is now used in the building process.
4592 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4593 the "build-* directories that it finds.
4595 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4596 address in a domain literal was a prefix of an interface address.
4598 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4599 when verifying a sender address, unless rewrite_headers is false.
4601 PH/44 Wrote a long comment about why errors_to addresses are verified as
4602 recipients, not senders.
4604 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4605 the ratelimit ACL was added.
4607 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4609 PH/46 Added extra information about PostgreSQL errors to the error string.
4611 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4612 avoiding the need to install special zones in a real server. This is
4613 backwards compatible; if it can't find the fake resolver, it drops back.
4614 Thus, both old and new test suites can be run.
4616 TF/02 Added util/ratelimit.pl
4618 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4619 clock is set back in time.
4621 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4622 Candler <B.Candler@pobox.com>.
4624 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4625 OK for addresses that are the result of redirections.
4627 PH/48 A number of further additions for the benefit of the new test suite,
4628 including a fake gethostbyname() that interfaces to the fake DNS resolver
4631 TF/06 The fix for widen_domains has also been applied to qualify_single and
4632 search_parents which are the other dnslookup options that can cause
4635 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4638 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4640 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4641 service" if it encountered an MX record with an empty target hostname.
4642 The message is now "an MX or SRV record indicated no SMTP service".
4644 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4645 if the DNS of the sending site is misconfigured. This is quite a
4646 common situation. This change restores the behaviour of treating a
4647 helo verification defer as a failure.
4649 PH/51 If self=fail was set on a router, the bounce message did not include the
4650 actual error message.
4656 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4658 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4659 command itself ended in error, the underlying message about the transport
4660 filter timeout was being overwritten with the pipe command error. Now the
4661 underlying error message should be appended to the second error message.
4663 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4665 PH/02 Reduce the amount of output that "make" produces by default. Full output
4666 can still be requested.
4668 PH/03 The warning log line about a condition test deferring for a "warn" verb
4669 was being output only once per connection, rather than after each
4670 occurrence (because it was using the same function as for successful
4671 "warn" verbs). This seems wrong, so I have changed it.
4673 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4674 it should not have, which might have caused a crash in the right
4675 circumstances, but probably never did.
4677 PH/04 Installed a modified version of Tony Finch's patch to make submission
4678 mode fix the return path as well as the Sender: header line, and to
4679 add a /name= option so that you can make the user's friendly name appear
4682 TF/03 Added the control = fakedefer ACL modifier.
4684 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4685 Mark Lowes for thorough testing.
4687 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4689 TK/03 Merged latest SRS patch from Miles Wilton.
4691 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4692 with the definition in sysexits.h (which is #included earlier).
4693 Fortunately, Exim does not actually use EX_OK. The code used to try to
4694 preserve the sysexits.h value, by assuming that macro definitions were
4695 scanned for macro replacements. I have been disabused of this notion,
4696 so now the code just undefines EX_OK before #including unistd.h.
4698 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4699 in the smtp transport. When a block could not be written in a single
4700 write() function, the timeout was being re-applied to each part-write.
4701 This seems wrong - if the receiver was accepting one byte at a time it
4702 would take for ever. The timeout is now adjusted when this happens. It
4703 doesn't have to be particularly precise.
4705 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4706 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4708 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4709 without a domain if the check to <postmaster@domain> fails.
4711 SC/01 Eximstats: added -xls and the ability to specify output files
4712 (patch written by Frank Heydlauf).
4714 SC/02 Eximstats: use FileHandles for outputting results.
4716 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4718 SC/04 Eximstats: fixed display of large numbers with -nvr option
4720 SC/05 Eximstats: fixed merging of reports with empty tables.
4722 SC/06 Eximstats: added the -include_original_destination flag
4724 SC/07 Eximstats: removed tabs and trailing whitespace.
4726 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4728 TK/06 MBOX spool code: Add real "From " MBOX separator line
4729 so the .eml file is really in mbox format (even though
4730 most programs do not really care). Patch from Alex Miller.
4732 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4733 The latter is generated from $received_to and is only set if the
4734 message has one envelope recipient. SA can use these headers,
4735 obviously out-of-the-box. Patch from Alex Miller.
4737 PH/08 The ${def test on a variable was returning false if the variable's
4738 value was "0", contrary to what the specification has always said!
4739 The result should be true unless the variable is empty.
4741 PH/09 The syntax error of a character other than { following "${if
4742 def:variable_name" (after optional whitespace) was not being diagnosed.
4743 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4744 accidental colon was present, for example, could give incorrect results.
4746 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4747 result is used (not including appendfile, where other changes are about
4750 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4751 This involved changing a lot of size variables from int to off_t. It
4752 should work with maildirs and everything.
4754 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4755 spamd dying while we are connected to it.
4757 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4760 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4761 function for BDB 4.3.
4763 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4765 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4766 the sources for PCRE have been re-arranged and more files are now
4769 PH/15 The code I had for printing potentially long long variables in PH/11
4770 above was not the best (it lost precision). The length of off_t variables
4771 is now inspected at build time, and an appropriate printing format (%ld
4772 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4773 to be "long long int" or "long int". This is needed for the internal
4774 formatting function string_vformat().
4776 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4777 the configuration file to be ":syslog", then the script "guesses" where
4778 the logs files are, rather than using the compiled in default. In our
4779 case the guess is not the same as the compiled default, so the script
4780 suddenly stopped working when I started to use syslog. The patch checks
4781 to see if log_file_path is "". If so, it attempts to read it from exim
4782 with no configuration file to get the compiled in version, before it
4783 falls back to the previous guessing code."
4785 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4786 implementing BATV in an Exim configuration. See NewStuff for the gory
4789 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4790 Makefile that are specific to HP-UX.
4792 PH/18 If the "use_postmaster" option was set for a recipient callout together
4793 with the "random" option, the postmaster address was used as the MAIL
4794 FROM address for the random test, but not for the subsequent recipient
4795 test. It is now used for both.
4797 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4798 patch removes a few documentation additions to RFC 3028, because the
4799 latest draft now contains them. It adds the new en;ascii-case comparator
4800 and a new error check for 8bit text in MIME parts. Comparator and
4801 require names are now matched exactly. I enabled the subaddress
4802 extension, but it is not well tested yet (read: it works for me)."
4804 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4805 rework some of the code of TK/09 above to avoid the hardwired use of
4806 "%lld" and "long long". Replaced the call to snprintf() with a call to
4809 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4810 records point to non-existent hosts", "retry timeout exceeded", and
4811 "retry time not reached for any host after a long failure period".
4813 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4814 experimental DomainKeys support:
4816 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4817 (2) On an error such as an illegally used "control", the wrong name for
4818 the control was given.
4820 These problems did NOT occur unless DomainKeys support was compiled.
4822 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4824 PH/24 Added ${if match_ip condition.
4826 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4827 databases so that it will be absolutely obvious if a crash occurs in the
4828 DB library. This is a regular occurrence (often caused by mis-matched
4831 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4832 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4833 picked up on a user's system that detects such things. There doesn't seem
4834 to be a gcc warning option for this - only an attribute that has to be
4835 put on the function's prototype. It seems that in Fedora Core 4 they have
4836 set this on a number of new functions. No doubt there will be more in due
4839 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4840 specify a transport. However, if an address that was verified by such a
4841 router was the subject of a callout, Exim crashed because it tried to
4842 read the rcpt_include_affixes from the non-existent transport. Now it
4843 just assumes that the setting of that option is false. This bug was
4844 introduced by 4.51/PH/31.
4846 PH/28 Changed -d+all to exclude +memory, because that information is very
4847 rarely of interest, but it makes the output a lot bigger. People tend to
4848 do -d+all out of habit.
4850 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4851 code in os-type was giving problems when libc.so lives in lib64, like on
4854 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4855 aren't the modern standard, and it seems that some systems' include files
4856 don't always have them. Exim was already checking for some of the newer
4857 ones like T_AAAA, and defining it itself. I've added checks for all the
4858 record types that Exim uses.
4860 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4861 not automatically generating a new one, as it is supposed to. This
4862 prevented TLS from working. If the file did exist, but contained invalid
4863 data, a new version was generated, as expected. It was only the case of a
4864 non-existent file that was broken.
4866 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4867 with a change in libdomainkeys > 0.64.
4869 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4870 from DNS. If the selector record carries the flag, it now has
4871 precedence over the domain-wide flag.
4873 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4875 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4876 the use of a port name that isn't defined in /etc/services) occurred, the
4877 message was deferred as in a normal delivery, and thus remained on the
4878 spool, instead of being failed because of the mua_wrapper setting. This
4879 is now fixed, and I tidied up some of the mua_wrapper messages at the
4882 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4883 the messages in a hash of arrays rather than using individual hashes.
4884 This is a bit cleaner and results in dramatic memory savings, albeit
4885 at a slight CPU cost.
4887 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4888 as requested by Marc Sherman.
4890 SC/10 Eximstats: added histograms for user specified patterns as requested
4893 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4895 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4896 fopen() in the content-scanning modules that did not already have it.
4902 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4903 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4905 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4907 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4909 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4910 to test Sieve filters that use "vacation".
4912 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4913 that changes the way the GnuTLS parameters are stored in the cache file.
4914 The new format can be generated externally. For backward compatibility,
4915 if the data in the cache doesn't make sense, Exim assumes it has read an
4916 old-format file, and it generates new data and writes a new file. This
4917 means that you can't go back to an older release without removing the
4920 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4921 work if there are any delivery delays because "one_time" forces the
4922 parent to be marked "delivered", so its unseen clone is never tried
4923 again. For this reason, Exim now forbids the simultaneous setting of
4926 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4927 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4928 entry if you want to know the details. The fix, however, neglected to
4929 consider the case when local delivery batching is involved. The test for
4930 "previously delivered" was not happening when checking to see if an
4931 address could be batched with a previous (undelivered) one; under
4932 certain circumstances this could lead to multiple deliveries to the same
4935 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4936 in its include files, and this causes problems building Exim.
4938 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4939 header_syntax) but Exim was just ignoring anything given after a slash.
4940 In particular, this caused confusion with an attempt to use "verify =
4941 reverse_host_lookup/defer_ok". An error is now given when options are
4942 supplied for verify items that do not have them. (Maybe reverse_host_
4943 lookup should have a defer_ok option, but that's a different point.)
4945 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4946 defined by RFC 821) to 2048, because there were problems with some AUTH
4947 commands, and RFC 1869 says the size should be increased for extended
4948 SMTP commands that take arguments.
4950 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4953 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4954 "unknown" error; now it says that the functionality isn't in the binary.
4956 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4957 an address' error message when a string expansion fails (syntax or
4958 whatever). Otherwise the password may appear in the log. Following change
4959 PH/42 below, there is no longer a chance of it appearing in a bounce
4962 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4964 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4965 of its addresses was ever tried. (Bugzilla bug #2.)
4967 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4968 the result incorrectly in the debug output. (It correctly added a newline
4969 to what was transported.)
4971 TF/01 Added $received_time.
4973 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4974 commented out examples of how to interface to a virus scanner and to
4975 SpamAssassin. Also added commented examples of av_scanner and
4976 spamd_address settings.
4978 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4979 and controls are allowed in which ACLs. There were a couple of minor
4980 errors. Some of the entries in the conditions table (which is a table of
4981 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4982 negation of where the condition IS allowed.
4984 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4986 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4987 header file does not have a version number, so I've had to invent a new
4988 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4989 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4990 radiusclient), but it was contributed by a Radius user.
4992 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4993 files or directories, for OpenSSL.
4995 PH/19 When an Exim process that is running as root has to create an Exim log
4996 file, it does so in a subprocess that runs as exim:exim so as to get the
4997 ownership right at creation (otherwise, other Exim processes might see
4998 the file with the wrong ownership). There was no test for failure of this
4999 fork() call, which would lead to the process getting stuck as it waited
5000 for a non-existent subprocess. Forks do occasionally fail when resources
5001 run out. I reviewed all the other calls to fork(); they all seem to check
5004 PH/20 When checking for unexpected SMTP input at connect time (before writing
5005 the banner), Exim was not dealing correctly with a non-positive return
5006 from the read() function. If the client had disconnected by this time,
5007 the result was a log entry for a synchronization error with an empty
5008 string after "input=" when read() returned zero. If read() returned -1
5009 (an event I could not check), uninitialized data bytes were printed.
5010 There were reports of junk text (parts of files, etc) appearing after
5013 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5015 PH/22 Added support for macro redefinition, and (re)definition in between
5016 driver and ACL definitions.
5018 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5019 forgetting to use the resulting value; it was using the unexpanded value.
5021 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5022 hadn't been configured. The fix is from Juergen Kreileder, who
5023 understands it better than I do:
5025 "Here's what I see happening with three configured cyrus_sasl
5026 authenticators configured (plain, login, cram-md5):
5028 On startup auth_cyrus_sasl_init() gets called for each of these.
5029 This means three calls to sasl_listmech() without a specified mech_list.
5030 => SASL tests which mechs of all available mechs actually work
5031 => three warnings about OTP not working
5032 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5034 With the patch, sasl_listmech() also gets called three times. But now
5035 SASL's mech_list option is set to the server_mech specified in the the
5036 authenticator. Or in other words, the answer from sasl_listmech()
5037 gets limited to just the mech you're testing for (which is different
5039 => the return list contains just 'plain' or 'login', 'cram-md5' or
5040 nothing depending on the value of ob->server_mech.
5042 I've just tested the patch: Authentication still works fine,
5043 unavailable mechs specified in the exim configuration are still
5044 caught, and the auth.log warnings about OTP are gone."
5046 PH/25 When debugging is enabled, the contents of the command line are added
5047 to the debugging output, even when log_selector=+arguments is not
5050 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5051 answer is "GNU", and only if the return is "GNU/something" is the answer
5054 PH/27 $acl_verify_message is now set immediately after the failure of a
5055 verification in an ACL, and so is available in subsequent modifiers. In
5056 particular, the message can be preserved by coding like this:
5058 warn !verify = sender
5059 set acl_m0 = $acl_verify_message
5061 Previously, $acl_verify_message was set only while expanding "message"
5062 and "log_message" when a very denied access.
5064 PH/28 Modified OS/os.c-Linux with
5066 -#ifndef OS_LOAD_AVERAGE
5067 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5069 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5070 nomenclature these days.)
5072 PH/29 Installed patch from the Sieve maintainer that adds the options
5073 sieve_useraddress and sieve_subaddress to the redirect router.
5075 PH/30 In these circumstances:
5076 . Two addresses routed to the same list of hosts;
5077 . First host does not offer TLS;
5078 . First host accepts first address;
5079 . First host gives temporary error to second address;
5080 . Second host offers TLS and a TLS session is established;
5081 . Second host accepts second address.
5082 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5083 and peerdn, if requested) that were in fact used only for the second
5086 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5087 attention to any local part prefix or suffix that was matched by the
5088 router that accepted the address. It now behaves in the same way as it
5089 does for delivery: the affixes are removed from the local part unless
5090 rcpt_include_affixes is set on the transport.
5092 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5093 timeout during the DATA phase of an incoming message.
5095 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5096 applied a patch sanctioned by the Sieve maintainer.
5098 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5099 the uid or gid is negative. A case of a negative gid caused this to be
5100 noticed. The fix allows for either to be negative.
5102 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5103 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5106 PH/36 The change PH/12 above was broken. Fixed it.
5108 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5109 the grounds that routing the same address twice would always produce the
5110 same answer. This might have been true once, but it is certainly no
5111 longer true now. Routing a child address may depend on the previous
5112 routing that produced that child. Some complicated redirection strategies
5113 went wrong when messages had multiple recipients, and made Exim's
5114 behaviour dependent on the order in which the addresses were given.
5116 I have moved the duplicate checking until after the routing is complete.
5117 Exim scans the addresses that are assigned to local and remote
5118 transports, and removes any duplicates. This means that more work will be
5119 done, as duplicates will always all be routed, but duplicates are
5120 presumably rare, so I don't expect this is of any significance.
5122 For deliveries to pipes, files, and autoreplies, the duplicate checking
5123 still happens during the routing process, since they are not going to be
5126 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5127 It corrects a timeout issue with spamd. This is Ian's comment: "The
5128 background is that sometimes spamd either never reads data from a
5129 connection it has accepted, or it never writes response data. The exiscan
5130 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5131 blindly assumes that writes won't block so it may never time out."
5133 PH/39 Allow G after quota size as well as K and M.
5135 PH/40 The value set for $authenticated_id in an authenticator may not contain
5136 binary zeroes or newlines because the value is written to log lines and
5137 to spool files. There was no check on this. Now the value is run through
5138 the string_printing() function so that such characters are converted to
5139 printable escape sequences.
5141 PH/41 $message_linecount is a new variable that contains the total number of
5142 lines in the message. Compare $body_linecount, which is the count for the
5145 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5146 bounce and delay warning messages, except in certain special cases, which
5149 (a) An SMTP error message from a remote host;
5150 (b) A message specified in a :fail: redirection;
5151 (c) A message specified in a "fail" command in a system filter;
5152 (d) A message specified in a FAIL return from the queryprogram router;
5153 (e) A message specified by the cannot_route_message router option.
5155 In these cases only, Exim does include the error details in bounce and
5156 warning messages. There are also a few cases where bland messages such
5157 as "unrouteable address" or "local delivery error" are given.
5159 PH/43 $value is now also set for the "else" part of a ${run expansion.
5161 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5162 being worked on, but at least Exim now implements the latest version to
5165 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5166 process to complete was treated as a delivery failure, a timeout while
5167 writing the message to the pipe was logged, but erroneously treated as a
5168 successful delivery. Such timeouts include transport filter timeouts. For
5169 consistency with the overall process timeout, these timeouts are now
5170 treated as errors, giving rise to delivery failures by default. However,
5171 there is now a new Boolean option for the pipe transport called
5172 timeout_defer, which, if set TRUE, converts the failures into defers for
5173 both kinds of timeout. A transport filter timeout is now identified in
5176 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5177 systems where "make" and "gmake" are different, calling "gmake" at top
5178 level broke things. I've arranged for the value of $(MAKE) to be passed
5179 from the Makefile to this script so that it can call the same version of
5183 A note about Exim versions 4.44 and 4.50
5184 ----------------------------------------
5186 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5187 changes of various kinds. As a consequence, a big documentation update was
5188 needed. This delayed the release for rather longer than seemed good, especially
5189 in the light of a couple of (minor) security issues. Therefore, the changes
5190 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5191 release. So 4.44 and 4.50 are in effect two different branches that both start
5194 I have left the 4.50 change log unchanged; it contains all the changes since
5195 4.43. The change log for 4.44 is below; many of its items are identical to
5196 those for 4.50. This seems to be the most sensible way to preserve the
5197 historical information.
5203 1. Minor wording change to the doc/README.SIEVE file.
5205 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5206 computation of the current number of files was incorrect.
5208 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5209 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5212 4. Give more explanation in the error message when the command for a transport
5213 filter fails to execute.
5215 5. There are several places where Exim runs a non-Exim command in a
5216 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5217 was being done only for the command run by the queryprogram router. It is
5218 now done for all such subprocesses. The other cases are: ${run, transport
5219 filters, and the commands run by the lmtp and pipe transports.
5221 6. Added CONFIGURE_GROUP build-time option.
5223 7. Some older OS have a limit of 256 on the maximum number of file
5224 descriptors. Exim was using setrlimit() to set 1000 as a large value
5225 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5226 systems. I've change it so that if it can't get 1000, it tries for 256.
5228 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5229 was an oversight, and furthermore, ever since the addition of extra
5230 controls (e.g. 4.43/32), the checks on when to allow different forms of
5231 "control" were broken. There should now be diagnostics for all cases when a
5232 control that does not make sense is encountered.
5234 9. Added the /retain_sender option to "control=submission".
5236 10. $recipients is now available in the predata ACL (oversight).
5238 11. Tidy the search cache before the fork to do a delivery from a message
5239 received from the command line. Otherwise the child will trigger a lookup
5240 failure and thereby defer the delivery if it tries to use (for example) a
5241 cached ldap connection that the parent has called unbind on.
5243 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5244 of $address_data from the recipient verification was clobbered by the
5245 sender verification.
5247 13. The value of address_data from a sender verification is now available in
5248 $sender_address_data in subsequent conditions in the ACL statement.
5250 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5252 15. Added a new option "connect=<time>" to callout options, to set a different
5255 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5256 was its contents. (It was OK if the option was not defined at all.)
5258 17. A "Completed" log line is now written for messages that are removed from
5259 the spool by the -Mrm option.
5261 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5262 information about exactly what failed.
5264 19. Added -dd to debug only the daemon process.
5266 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5267 handles timeouts, both on the server side and network timeouts. Renamed the
5268 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5270 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5271 It is now set to "smtps".
5273 22. $host_address is now set to the target address during the checking of
5274 ignore_target_hosts.
5276 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5277 being passed; this would have caused $sender_host_name to have been used if
5278 matching the list had actually called for a host name (not very likely,
5279 since this list is usually IP addresses). A host name is now passed as
5282 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5283 code that specifies a non-restarting handler (typically sigaction() in
5284 modern systems) in an attempt to fix a rare and obscure crash bug.
5286 25. Narrowed the window for a race in the daemon that could cause it to ignore
5287 SIGCHLD signals. This is not a major problem, because they are used only to
5288 wake it up if nothing else does.
5290 26. A malformed maildirsize file could cause Exim to calculate negative values
5291 for the mailbox size or file count. Odd effects could occur as a result.
5292 The maildirsize information is now recalculated if the size or filecount
5295 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5296 support for a long time. Removed HAVE_SYS_VFS_H.
5298 28. Installed the latest version of exipick from John Jetmore.
5300 29. In an address list, if the pattern was not a regular expression, an empty
5301 subject address (from a bounce message) matched only if the pattern was an
5302 empty string. Non-empty patterns were not even tested. This was the wrong
5303 because it is perfectly reasonable to use an empty address as part of a
5304 database query. An empty address is now tested by patterns that are
5305 lookups. However, all the other forms of pattern expect the subject to
5306 contain a local part and a domain, and therefore, for them, an empty
5307 address still always fails if the pattern is not itself empty.
5309 30. Exim went into a mad DNS loop when attempting to do a callout where the
5310 host was specified on an smtp transport, and looking it up yielded more
5311 than one IP address.
5313 31. Re-factored the code for checking spool and log partition space into a
5314 function that finds that data and another that does the check. The former
5315 is then used to implement four new variables: $spool_space, $log_space,
5316 $spool_inodes, and $log_inodes.
5318 32. The RFC2047 encoding function was originally intended for short strings
5319 such as real names; it was not keeping to the 75-character limit for
5320 encoded words that the RFC imposes. It now respects the limit, and
5321 generates multiple encoded words if necessary. To be on the safe side, I
5322 have increased the buffer size for the ${rfc2047: expansion operator from
5325 33. It is now permitted to omit both strings after an "if" condition; if the
5326 condition is true, the result is "true". As before, when the second string
5327 is omitted, a false condition yields an empty string. This makes it less
5328 cumbersome to write custom ACL and router conditions.
5330 34. Failure to deliver a bounce message always caused it to be frozen, even if
5331 there was an errors_to setting on the router. The errors_to setting is now
5334 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5335 canonical form (fully expanded) before being placed in
5336 $sender_host_address.
5338 36. The table in the code that translates DNS record types into text (T_A to
5339 "A" for instance) was missing entries for NS and CNAME. It is just possible
5340 that this could have caused confusion if both these types were looked up
5341 for the same domain, because the text type is used as part of Exim's
5342 per-process caching. But the chance of anyone hitting this buglet seems
5345 37. The dnsdb lookup has been extended in a number of ways.
5347 (1) There is a new type, "zns", which walks up the domain tree until it
5348 finds some nameserver records. It should be used with care.
5350 (2) There is a new type, "mxh", which is like "mx" except that it returns
5351 just the host names, not the priorities.
5353 (3) It is now possible to give a list of domains (or IP addresses) to be
5354 looked up. The behaviour when one of the lookups defers can be
5355 controlled by a keyword.
5357 (4) It is now possible to specify the separator character for use when
5358 multiple records are returned.
5360 38. The dnslists ACL condition has been extended: it is now possible to supply
5361 a list of IP addresses and/or domains to be looked up in a particular DNS
5364 39. Added log_selector=+queue_time_overall.
5366 40. When running the queue in the test harness, wait just a tad after forking a
5367 delivery process, to get repeatability of debugging output.
5369 41. Include certificate and key file names in error message when GnuTLS fails
5370 to set them up, because the GnuTLS error message doesn't include the name
5371 of the failing file when there is a problem reading it.
5373 42. Allow both -bf and -bF in the same test run.
5375 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5377 44. The "Exiscan patch" is now merged into the mainline Exim source.
5379 45. Sometimes the final signoff response after QUIT could fail to get
5380 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5381 < 0 before doing a fflush(). This bug looks as though it goes back to the
5382 introduction of TLS in release 3.20, but "sometimes" must have been rare
5383 because the tests only now provoked it.
5385 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5386 (this can affect the format of dates).
5388 47. exim_tidydb, when checking for the continued existence of a message for
5389 which it has found a message-specific retry record, was not finding
5390 messages that were in split spool directories. Consequently, it was
5391 deleting retry records that should have stayed in existence.
5393 48. Steve fixed some bugs in eximstats.
5395 49. The SPA authentication driver was not abandoning authentication and moving
5396 on to the next authenticator when an expansion was forced to fail,
5397 contradicting the general specification for all authenticators. Instead it
5398 was generating a temporary error. It now behaves as specified.
5400 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5401 (the order specifies the preference for clients). The order is now AES256,
5402 AES128, 3DES, ARCFOUR128.
5404 51. Small patch to Sieve code - explicitly set From: when generating an
5407 52. Exim crashed if a remote delivery caused a very long error message to be
5408 recorded - for instance if somebody sent an entire SpamAssassin report back
5409 as a large number of 550 error lines. This bug was coincidentally fixed by
5410 increasing the size of one of Exim's internal buffers (big_buffer) that
5411 happened as part of the Exiscan merge. However, to be on the safe side, I
5412 have made the code more robust (and fixed the comments that describe what
5415 53. Now that there can be additional text after "Completed" in log lines (if
5416 the queue_time_overall log selector is set), a one-byte patch to exigrep
5417 was needed to allow it to recognize "Completed" as not the last thing in
5420 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5421 patch that reportedly fixes this has been added. I am not expert enough to
5422 create a test for it. This is what the patch creator wrote:
5424 "I found a little strange behaviour of ldap code when working with
5425 Windows 2003 AD Domain, where users was placed in more than one
5426 Organization Units. When I tried to give exim partial DN, the exit code
5427 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5428 But simultaneously result of request was absolutely normal ldap result,
5429 so I produce this patch..."
5431 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5432 REFERENCE, so I have modified the code to exclude the patch when that macro
5435 55. Some experimental protocols are using DNS PTR records for new purposes. The
5436 keys for these records are domain names, not reversed IP addresses. The
5437 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5438 leaves it alone. Component reversal etc. now happens only for IP addresses.
5441 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5443 57. Double the size of the debug message buffer (to 2048) so that more of very
5444 long debug lines gets shown.
5446 58. The exicyclog utility now does better if the number of log files to keep
5447 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5449 59. Two changes related to the smtp_active_hostname option:
5451 (1) $smtp_active_hostname is now available as a variable.
5452 (2) The default for smtp_banner uses $smtp_active_hostname instead
5453 of $primary_hostname.
5455 60. The host_aton() function is supposed to be passed a string that is known
5456 to be a valid IP address. However, in the case of IPv6 addresses, it was
5457 not checking this. This is a hostage to fortune. Exim now panics and dies
5458 if the condition is not met. A case was found where this could be provoked
5459 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5460 components; fortuitously, this particular loophole had already been fixed
5461 by change 4.50/55 above.
5463 If there are any other similar loopholes, the new check in host_aton()
5464 itself should stop them being exploited. The report I received stated that
5465 data on the command line could provoke the exploit when Exim was running as
5466 exim, but did not say which command line option was involved. All I could
5467 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5468 running as the user.
5471 61. There was a buffer overflow vulnerability in the SPA authentication code
5472 (which came originally from the Samba project). I have added a test to the
5473 spa_base64_to_bits() function which I hope fixes it.
5476 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5477 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5479 63. The daemon start-up calls getloadavg() while still root for those OS that
5480 need the first call to be done as root, but it missed one case: when
5481 deliver_queue_load_max is set with deliver_drop_privilege. This is
5482 necessary for the benefit of the queue runner, because there is no re-exec
5483 when deliver_drop_privilege is set.
5485 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5486 This has been fixed.
5488 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5489 was in use, was not putting the data itself into the right store pool;
5490 consequently, it could be overwritten for a subsequent message in the same
5491 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5494 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5496 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5497 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5498 the function were treating the return as a boolean value, which happened to
5499 work because 0=false and not-0=true, but is not correct code.
5501 68. The host_aton() function was not handling scoped IPv6 addresses (those
5502 with, for example, "%eth0" on the end) correctly.
5504 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5505 negated items (that is, ~something) in unsigned ints. Some compilers
5506 apparently mutter when there is no cast.
5508 70. If an address verification called from an ACL failed, and did not produce a
5509 user-specific message (i.e. there was only a "system" message), nothing was
5510 put in $acl_verify_message. In this situation, it now puts the system
5513 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5514 session; change 4.31/43 added the unwanted input to the log line - except
5515 that it did not do this in the start of session case. It now does.
5517 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5518 This is wrong and can cause the other end to generate a synchronization
5519 error if it is another Exim or anything else that does the synchronization
5520 check. A QUIT command is no longer sent after a timeout.
5522 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5523 during host lookups.
5525 74. The defer_ok option of callout verification was not working if it was used
5526 when verifying addresses in header lines, that is, for this case:
5528 verify = header_sender/callout=defer_ok
5530 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5531 those file descriptors could be used for SMTP connections. If anything
5532 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5533 could be sent to the SMTP client, causing chaos. The daemon now opens
5534 stdin, stdout, and stderr to /dev/null when it puts itself into the
5537 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5538 log by default. The user can override this with suitable Perl magic.
5540 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5541 the log message when discard triggers, was not working for the DATA ACL or
5542 for the non-SMTP ACL.
5544 78. Error message wording change in sieve.c.
5546 79. If smtp_accept_max_per_host was set, the number of connections could be
5547 restricted to fewer than expected, because the daemon was trying to set up
5548 a new connection before checking whether the processes handling previous
5549 connections had finished. The check for completed processes is now done
5550 earlier. On busy systems, this bug wouldn't be noticed because something
5551 else would have woken the daemon, and it would have reaped the completed
5554 80. If a message was submitted locally by a user whose login name contained one
5555 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5556 It caused a spool format error. I have fixed the spool reading code. A
5557 related problem was that the "from" clause in the Received: line became
5558 illegal because of the space(s). It is now covered by ${quote_local_part.
5560 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5563 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5565 83. Patch from Sieve maintainer to fix unterminated string problem in
5566 "vacation" handling.
5568 84. Some minor changes to the Linux configuration files to help with other
5569 OS variants using glibc.
5571 85. One more patch for Sieve to update vacation handling to latest spec.
5574 ----------------------------------------------------
5575 See the note above about the 4.44 and 4.50 releases.
5576 ----------------------------------------------------
5582 1. Change 4.43/35 introduced a bug that caused file counts to be
5583 incorrectly computed when quota_filecount was set in an appendfile
5586 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5587 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5590 3. Give more explanation in the error message when the command for a transport
5591 filter fails to execute.
5593 4. There are several places where Exim runs a non-Exim command in a
5594 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5595 was being done only for the command run by the queryprogram router. It is
5596 now done for all such subprocesses. The other cases are: ${run, transport
5597 filters, and the commands run by the lmtp and pipe transports.
5599 5. Some older OS have a limit of 256 on the maximum number of file
5600 descriptors. Exim was using setrlimit() to set 1000 as a large value
5601 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5602 systems. I've change it so that if it can't get 1000, it tries for 256.
5604 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5605 was an oversight, and furthermore, ever since the addition of extra
5606 controls (e.g. 4.43/32), the checks on when to allow different forms of
5607 "control" were broken. There should now be diagnostics for all cases when a
5608 control that does not make sense is encountered.
5610 7. $recipients is now available in the predata ACL (oversight).
5612 8. Tidy the search cache before the fork to do a delivery from a message
5613 received from the command line. Otherwise the child will trigger a lookup
5614 failure and thereby defer the delivery if it tries to use (for example) a
5615 cached ldap connection that the parent has called unbind on.
5617 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5618 of $address_data from the recipient verification was clobbered by the
5619 sender verification.
5621 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5622 was its contents. (It was OK if the option was not defined at all.)
5624 11. A "Completed" log line is now written for messages that are removed from
5625 the spool by the -Mrm option.
5627 12. $host_address is now set to the target address during the checking of
5628 ignore_target_hosts.
5630 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5631 being passed; this would have caused $sender_host_name to have been used if
5632 matching the list had actually called for a host name (not very likely,
5633 since this list is usually IP addresses). A host name is now passed as
5636 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5637 code that specifies a non-restarting handler (typically sigaction() in
5638 modern systems) in an attempt to fix a rare and obscure crash bug.
5640 15. Narrowed the window for a race in the daemon that could cause it to ignore
5641 SIGCHLD signals. This is not a major problem, because they are used only to
5642 wake it up if nothing else does.
5644 16. A malformed maildirsize file could cause Exim to calculate negative values
5645 for the mailbox size or file count. Odd effects could occur as a result.
5646 The maildirsize information is now recalculated if the size or filecount
5649 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5650 support for a long time. Removed HAVE_SYS_VFS_H.
5652 18. Updated exipick to current release from John Jetmore.
5654 19. Allow an empty sender to be matched against a lookup in an address list.
5655 Previously the only cases considered were a regular expression, or an
5658 20. Exim went into a mad DNS lookup loop when doing a callout where the
5659 host was specified on the transport, if the DNS lookup yielded more than
5662 21. The RFC2047 encoding function was originally intended for short strings
5663 such as real names; it was not keeping to the 75-character limit for
5664 encoded words that the RFC imposes. It now respects the limit, and
5665 generates multiple encoded words if necessary. To be on the safe side, I
5666 have increased the buffer size for the ${rfc2047: expansion operator from
5669 22. Failure to deliver a bounce message always caused it to be frozen, even if
5670 there was an errors_to setting on the router. The errors_to setting is now
5673 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5674 canonical form (fully expanded) before being placed in
5675 $sender_host_address.
5677 24. Updated eximstats to version 1.33
5679 25. Include certificate and key file names in error message when GnuTLS fails
5680 to set them up, because the GnuTLS error message doesn't include the name
5681 of the failing file when there is a problem reading it.
5683 26. Expand error message when OpenSSL has problems setting up cert/key files.
5686 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5687 (this can affect the format of dates).
5689 28. exim_tidydb, when checking for the continued existence of a message for
5690 which it has found a message-specific retry record, was not finding
5691 messages that were in split spool directories. Consequently, it was
5692 deleting retry records that should have stayed in existence.
5694 29. eximstats updated to version 1.35
5695 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5696 1.35 - bugfix such that pie charts by volume are generated correctly
5698 30. The SPA authentication driver was not abandoning authentication and moving
5699 on to the next authenticator when an expansion was forced to fail,
5700 contradicting the general specification for all authenticators. Instead it
5701 was generating a temporary error. It now behaves as specified.
5703 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5704 (the order specifies the preference for clients). The order is now AES256,
5705 AES128, 3DES, ARCFOUR128.
5707 31. Small patch to Sieve code - explicitly set From: when generating an
5710 32. Exim crashed if a remote delivery caused a very long error message to be
5711 recorded - for instance if somebody sent an entire SpamAssassin report back
5712 as a large number of 550 error lines. This bug was coincidentally fixed by
5713 increasing the size of one of Exim's internal buffers (big_buffer) that
5714 happened as part of the Exiscan merge. However, to be on the safe side, I
5715 have made the code more robust (and fixed the comments that describe what
5718 33. Some experimental protocols are using DNS PTR records for new purposes. The
5719 keys for these records are domain names, not reversed IP addresses. The
5720 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5721 leaves it alone. Component reversal etc. now happens only for IP addresses.
5724 34. The host_aton() function is supposed to be passed a string that is known
5725 to be a valid IP address. However, in the case of IPv6 addresses, it was
5726 not checking this. This is a hostage to fortune. Exim now panics and dies
5727 if the condition is not met. A case was found where this could be provoked
5728 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5729 components; fortuitously, this particular loophole had already been fixed
5730 by change 4.50/55 or 4.44/33 above.
5732 If there are any other similar loopholes, the new check in host_aton()
5733 itself should stop them being exploited. The report I received stated that
5734 data on the command line could provoke the exploit when Exim was running as
5735 exim, but did not say which command line option was involved. All I could
5736 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5737 running as the user.
5740 35. There was a buffer overflow vulnerability in the SPA authentication code
5741 (which came originally from the Samba project). I have added a test to the
5742 spa_base64_to_bits() function which I hope fixes it.
5745 36. The daemon start-up calls getloadavg() while still root for those OS that
5746 need the first call to be done as root, but it missed one case: when
5747 deliver_queue_load_max is set with deliver_drop_privilege. This is
5748 necessary for the benefit of the queue runner, because there is no re-exec
5749 when deliver_drop_privilege is set.
5751 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5752 was in use, was not putting the data itself into the right store pool;
5753 consequently, it could be overwritten for a subsequent message in the same
5754 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5757 38. Sometimes the final signoff response after QUIT could fail to get
5758 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5759 < 0 before doing a fflush(). This bug looks as though it goes back to the
5760 introduction of TLS in release 3.20, but "sometimes" must have been rare
5761 because the tests only now provoked it.
5767 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5768 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5769 Now it returns an integer. A number of calls were still expecting a T/F
5770 return. Fortuitously, in all cases, the tests worked in OK situations,
5771 which is the norm. However, things would have gone wrong on any write
5772 failures on the smtp file descriptor. This function is used when sending
5773 messages over SMTP and also when doing verify callouts.
5775 2. When Exim is called to do synchronous delivery of a locally submitted
5776 message (the -odf or -odi options), it no longer closes stderr before doing
5779 3. Implemented the mua_wrapper option.
5781 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5783 5. Implemented the functions header_remove(), header_testname(),
5784 header_add_at_position(), and receive_remove_recipient(), and exported them
5787 6. If an ACL "warn" statement specified the addition of headers, Exim already
5788 inserted X-ACL-Warn: at the start if there was no header name. However, it
5789 was not making this test for the second and subsequent header lines if
5790 there were newlines in the string. This meant that an invalid header could
5791 be inserted if Exim was badly configured.
5793 7. Allow an ACL "warn" statement to add header lines at the start or after all
5794 the Received: headers, as well as at the end.
5796 8. Added the rcpt_4xx retry error code.
5798 9. Added postmaster_mailfrom=xxx to callout verification option.
5800 10. Added mailfrom=xxxx to the callout verification option, for verify=
5803 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5804 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5806 12. Inserted some casts to stop certain compilers warning when using pointer
5807 differences as field lengths or precisions in printf-type calls (mostly
5808 affecting debugging statements).
5810 13. Added optional readline() support for -be (dynamically loaded).
5812 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5813 same clock tick as a message's arrival, so that its received time was the
5814 same as the "first fail" time on the retry record, and that message
5815 remained on the queue past the ultimate address timeout, every queue runner
5816 would try a delivery (because it was past the ultimate address timeout) but
5817 after another failure, the ultimate address timeout, which should have then
5818 bounced the address, did not kick in. This was a "< instead of <=" error;
5819 in most cases the first failure would have been in the next clock tick
5820 after the received time, and all would be well.
5822 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5823 being recognized when the domain list was tested by the match_domain
5824 condition in an expansion string.
5826 16. Added the ${str2b64: operator.
5828 17. Exim was always calling setrlimit() to set a large limit for the number of
5829 processes, without checking whether the existing limit was already
5830 adequate. (It did check for the limit on file descriptors.) Furthermore,
5831 errors from getrlimit() and setrlimit() were being ignored. Now they are
5832 logged to the main and panic logs, but Exim does carry on, to try to do its
5833 job under whatever limits there are.
5835 18. Imported PCRE 5.0.
5837 19. Trivial typo in log message " temporarily refused connection" (the leading
5840 20. If the log selector return_path_on_delivery was set and an address was
5841 redirected to /dev/null, the delivery process crashed because it assumed
5842 that a return path would always be set for a "successful" delivery. In this
5843 case, the whole delivery is bypassed as an optimization, and therefore no
5846 21. Internal re-arrangement: the function for sending a challenge and reading
5847 a response while authentication was assuming a zero-terminated challenge
5848 string. It's now changed to take a pointer and a length, to allow for
5849 binary data in such strings.
5851 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5853 23. Exim was not respecting finduser_retries when seeking the login of the
5854 uid under which it was called; it was always trying 10 times. (The default
5855 setting of finduser_retries is zero.) Also, it was sleeping after the final
5856 failure, which is pointless.
5858 24. Implemented tls_on_connect_ports.
5860 25. Implemented acl_smtp_predata.
5862 26. If the domain in control=submission is set empty, Exim assumes that the
5863 authenticated id is a complete email address when it generates From: or
5864 Sender: header lines.
5866 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5867 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5868 chown and chgrp in /bin and hostname in /usr/bin.
5870 28. Exim was keeping the "process log" file open after each use, just as it
5871 does for the main log. This opens the possibility of it remaining open for
5872 long periods when the USR1 signal hits a daemon. Occasional processlog
5873 errors were reported, that could have been caused by this. Anyway, it seems
5874 much more sensible not to leave this file open at all, so that is what now
5877 29. The long-running daemon process does not normally write to the log once it
5878 has entered its main loop, and it closes the log before doing so. This is
5879 so that log files can straightforwardly be renamed and moved. However,
5880 there are a couple of unusual error situations where the daemon does write
5881 log entries, and I had neglected to close the log afterwards.
5883 30. The text of an SMTP error response that was received during a remote
5884 delivery was being truncated at 512 bytes. This is too short for some of
5885 the long messages that one sometimes sees. I've increased the limit to
5888 31. It is now possible to make retry rules that apply only when a message has a
5889 specific sender, in particular, an empty sender.
5891 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5892 it possible to be selective about when SMTP synchronization is enforced.
5894 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5896 32. Implemented hosts_connection_nolog.
5898 33. Added an ACL for QUIT.
5900 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5903 35. Added mailbox_size and mailbox_filecount to appendfile.
5905 36. Added control = no_multiline_responses to ACLs.
5907 37. There was a bug in the logic of the code that waits for the clock to tick
5908 in the case where the clock went backwards by a substantial amount such
5909 that the microsecond fraction of "now" was more than the microsecond
5910 fraction of "then" (but the whole seconds number was less).
5912 38. Added support for the libradius Radius client library this is found on
5913 FreeBSD (previously only the radiusclient library was supported).
5919 1. When certain lookups returned multiple values in the form name=value, the
5920 quoting of the values was not always being done properly. Specifically:
5921 (a) If the value started with a double quote, but contained no whitespace,
5923 (b) If the value contained whitespace other than a space character (i.e.
5924 tabs or newlines or carriage returns) it was not quoted.
5925 This fix has been applied to the mysql and pgsql lookups by writing a
5926 separate quoting function and calling it from the lookup code. The fix
5927 should probably also be applied to nisplus, ibase and oracle lookups, but
5928 since I cannot test any of those, I have not disturbed their existing code.
5930 2. A hit in the callout cache for a specific address caused a log line with no
5931 reason for rejecting RCPT. Now it says "Previous (cached) callout
5932 verification failure".
5934 3. There was an off-by-one bug in the queryprogram router. An over-long
5935 return line was truncated at 256 instead of 255 characters, thereby
5936 overflowing its buffer with the terminating zero. As well as fixing this, I
5937 have increased the buffer size to 1024 (and made a note to document this).
5939 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5940 when Exim is waiting for an SMTP response from a remote server, Exim
5941 restarts its select() call on the socket, thereby resetting its timeout.
5942 This is not a problem when such interrupts are rare. Somebody set up a cron
5943 job to run exiwhat every 2 minutes, which is less than the normal select()
5944 timeout (5 or 10 minutes). This meant that the select() timeout never
5945 kicked in because it was always reset. I have fixed this by comparing the
5946 time when an interrupt arrives with the time at the start of the first call
5947 to select(). If more time than the timeout has elapsed, the interrupt is
5948 treated as a timeout.
5950 5. Some internal re-factoring in preparation for the addition of Sieve
5951 extensions (by MH). In particular, the "personal" test is moved to a
5952 separate function, and given an option for scanning Cc: and Bcc: (which is
5953 not set for Exim filters).
5955 6. When Exim created an email address using the login of the caller as the
5956 local part (e.g. when creating a From: or Sender: header line), it was not
5957 quoting the local part when it contained special characters such as @.
5959 7. Installed new OpenBSD configuration files.
5961 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5962 try to make them clearer.
5964 9. Callout options, other than the timeout value, were being ignored when
5965 verifying sender addresses in header lines. For example, when using
5967 verify = header_sender/callout=no_cache
5969 the cache was (incorrectly) being used.
5971 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5972 only the Cygwin environment.
5974 11. When return_path_on_delivery was set as a log selector, if different remote
5975 addresses in the same message used different return paths and parallel
5976 remote delivery occurred, the wrong values would sometimes be logged.
5977 (Whenever a remote delivery process finished, the return path value from
5978 the most recently started remote delivery process was logged.)
5980 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5981 lines when AUTH and/or TLS are in use. This is the "received protocol"
5982 field. Exim used to use "asmtp" for authenticated SMTP, without any
5983 indication (in the protocol name) for TLS use. Now it follows the RFC and
5984 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5985 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5986 names appear in log lines as well as in Received: header lines.
5988 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5989 extensions, and comparison tests, and to fix some bugs.
5991 14. Changes to the "personal" filter test:
5993 (1) The test was buggy in that it was just doing the equivalent of
5994 "contains" tests on header lines. For example, if a user's address was
5995 anne@some.where, the "personal" test would incorrectly be true for
5997 To: susanne@some.where
5999 This test is now done by extracting each address from the header in turn,
6000 and checking the entire address. Other tests that are part of "personal"
6001 are now done using regular expressions (for example, to check local parts
6002 of addresses in From: header lines).
6004 (2) The list of non-personal local parts in From: addresses has been
6005 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6006 taken from the Sieve specification recommendations.
6008 (3) If the message contains any header line starting with "List-" it is
6009 treated as non-personal.
6011 (4) The test for "circular" in the Subject: header line has been removed
6012 because it now seems ill-conceived.
6014 15. Minor typos in src/EDITME comments corrected.
6016 16. Installed latest exipick from John Jetmore.
6018 17. If headers_add on a router specified a text string that was too long for
6019 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6020 of string_sprintf() is now avoided.
6022 18. $message_body_size was not set (it was always zero) when running the DATA
6023 ACL and the local_scan() function.
6025 19. For the "mail" command in an Exim filter, no default was being set for
6026 the once_repeat time, causing a random time value to be used if "once" was
6027 specified. (If the value happened to be <= 0, no repeat happened.) The
6028 default is now 0s, meaning "never repeat". The "vacation" command was OK
6029 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6030 (I found it when inspecting the code).
6032 20. There is now an overall timeout for performing a callout verification. It
6033 defaults to 4 times the callout timeout, which applies to individual SMTP
6034 commands during the callout. The overall timeout applies when there is more
6035 than one host that can be tried. The timeout is checked before trying the
6036 next host. This prevents very long delays if there are a large number of
6037 hosts and all are timing out (e.g. when the network connections are timing
6038 out). The value of the overall timeout can be changed by specifying an
6039 additional sub-option for "callout", called "maxwait". For example:
6041 verify = sender/callout=5s,maxwait=20s
6043 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6044 to the end before writing, but this should make it even safer).
6046 22. Exim was forgetting that it had advertised PIPELINING for the second and
6047 subsequent messages on an SMTP connection. It was also not resetting its
6048 memory on STARTTLS and an internal HELO.
6050 23. When Exim logs an SMTP synchronization error within a session, it now
6051 records whether PIPELINING has been advertised or not.
6053 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6054 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6055 rather than long int.
6057 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6059 26. Added the never_mail option to autoreply.
6065 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6066 crash if the getsockname() call failed; this can happen if a connection is
6067 closed very soon after it is established. The problem was simply in the
6068 order in which certain operations were done, causing Exim to try to write
6069 to the SMTP stream before it had set up the file descriptor. The bug has
6070 been fixed by making things happen in the correct order.
6076 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6077 before the connection was closed, thus losing the rejection response.
6079 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6080 some early Solaris releases, but causes trouble in current releases where
6081 socklen_t is defined.
6083 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6086 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6089 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6090 that releasing the top of it at the end releases what was used for sub-
6091 expansions (unless the block got too big). However, discard this block if
6092 the first thing is a variable or header, so that we can use its block when
6093 it is dynamic (useful for very large $message_headers, for example).
6095 6. Lookups now cache *every* query, not just the most recent. A new, separate
6096 store pool is used for this. It can be recovered when all lookup caches are
6097 flushed. Lookups now release memory at the end of their result strings.
6098 This has involved some general refactoring of the lookup sources.
6100 7. Some code has been added to the store_xxx() functions to reduce the amount
6101 of flapping under certain conditions.
6103 8. log_incoming_interface used to affect only the <= reception log lines. Now
6104 it causes the local interface and port to be added to several more SMTP log
6105 lines, for example "SMTP connection from", and rejection lines.
6107 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6109 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6111 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6112 could be overwritten at the end of the current message (or the start of a
6113 new message if it was set in a HELO ACL). The value is now preserved for
6114 the duration of the SMTP connection.
6116 12. If a transport had a headers_rewrite setting, and a matching header line
6117 contained an unqualified address, that address was qualified, even if it
6118 did not match any rewriting rules. The underlying bug was that the values
6119 of the flags that permit the existence of unqualified sender and recipient
6120 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6121 non-local messages, and by -bnq for local messages) were not being
6122 preserved with the message after it was received.
6124 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6125 "next input=" as part of the text comprising the host identity instead of
6126 the correct text. The code was using the same buffer for two different
6127 strings. However, depending on which order the printing function evaluated
6128 its arguments, the bug did not always show up. Under Linux, for example, my
6129 test suite worked just fine.
6131 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6132 doesn't work with some older versions of Perl. It has been changed to "my",
6133 which in any case is probably the better facility to use.
6135 15. A really picky compiler found some instances of statements for creating
6136 error messages that either had too many or two few arguments for the format
6139 16. The size of the buffer for calls to the DNS resolver has been increased
6140 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6141 for addresses that have a lot of PTR records. This alleviates a problem; it
6142 does not fully solve it.
6144 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6145 buffer now truncates the list and logs the incident, which is the same
6146 action as happens when Exim is looking up a host name and its aliases.
6147 Previously in this situation something unpredictable would happen;
6148 sometimes it was "internal error: store_reset failed".
6150 18. If a server dropped the connection unexpectedly when an Exim client was
6151 using GnuTLS and trying to read a response, the client delivery process
6152 crashed while trying to generate an error log message.
6154 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6155 string, for example:
6157 warn message = H1: something\nH2: something
6159 the text was added as a single header line from Exim's point of view
6160 though it ended up OK in the delivered message. However, searching for the
6161 second and subsequent header lines using $h_h2: did not work. This has been
6162 fixed. Similarly, if a system filter added multiple headers in this way,
6163 the routers could not see them.
6165 20. Expanded the error message when iplsearch is called with an invalid key to
6166 suggest using net-iplsearch in a host list.
6168 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6169 ACLs are no longer actually imposed (and a message to that effect is
6172 22. If a "gecos" field in a passwd entry contained escaped characters, in
6173 particular, if it contained a \" sequence, Exim got it wrong when building
6174 a From: or a Sender: header from that name. A second bug also caused
6175 incorrect handling when an unquoted " was present following a character
6176 that needed quoting.
6178 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6179 was not being matched caselessly.
6181 24. Arranged for all hyphens in the exim.8 source to be escaped with
6184 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6185 properly. Recipient callout cache records were still being keyed to include
6186 the sender, even when use_sender was set false. This led to far more
6187 callouts that were necessary. The sender is no longer included in the key
6188 when use_sender is false.
6190 26. Added "control = submission" modifier to ACLs.
6192 27. Added the ${base62d: operator to decode base 62 numbers.
6194 28. dnsdb lookups can now access SRV records.
6196 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6197 the configuration file.
6199 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6200 (-v) mode. This makes the output for a verbose queue run more intelligible.
6202 31. Added a use_postmaster feature to recipient callouts.
6204 32. Added the $body_zerocount variable, containing the number of binary zero
6205 bytes in the message body.
6207 33. The time of last modification of the "new" subdirectory is now used as the
6208 "mailbox time last read" when there is a quota error for a maildir
6211 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6213 35. Added +ignore_unknown as a special item in host lists.
6215 36. Code for decoding IPv6 addresses in host lists is now included, even if
6216 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6217 address was recognized as an IP address, but was then not correctly decoded
6218 into binary, causing unexpected and incorrect effects when compared with
6225 1. Very minor rewording of debugging text in manualroute to say "list of
6226 hosts" instead of "hostlist".
6228 2. If verify=header_syntax was set, and a header line with an unqualified
6229 address (no domain) and a large number of spaces between the end of the
6230 name and the colon was received, the reception process suffered a buffer
6231 overflow, and (when I tested it) crashed. This was caused by some obsolete
6232 code that should have been removed. The fix is to remove it!
6234 3. When running in the test harness, delay a bit after writing a bounce
6235 message to get a bit more predictability in the log output.
6237 4. Added a call to search_tidyup() just before forking a reception process. In
6238 theory, someone could use a lookup in the expansion of smtp_accept_max_
6239 per_host which, without the tidyup, could leave open a database connection.
6241 5. Added the variables $recipient_data and $sender_data which get set from a
6242 lookup success in an ACL "recipients" or "senders" condition, or a router
6243 "senders" option, similar to $domain_data and $local_part_data.
6245 6. Moved the writing of debug_print from before to after the "senders" test
6248 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6249 problems for message scanning, either using a data ACL, or using
6250 local_scan() because the Received: header was not generated till after they
6251 were called (in order to set the time as the time of reception completion).
6252 I have revised the way this works. The header is now generated after the
6253 body is received, but before the ACL or local_scan() are called. After they
6254 are run, the timestamp in the header is updated.
6260 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6261 before starting a queue runner without re-exec. This happened only when
6262 deliver_drop_privilege was set or when the Exim user was set to root. The
6263 effect of the bug was that timeouts during subsequent deliveries caused
6264 crashes instead of being properly handled. The handler is now left at its
6265 default (and expected) setting.
6267 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6268 message, again when deliver_drop_privilege is set or Exim is run as root.
6269 The bug described in (1) was not present in this case, but the tidying up
6270 of the other signals was missing. I have made the two cases consistent.
6272 3. The ignore_target_hosts setting on a manualroute router was being ignored
6273 for hosts that were looked up using the /MX notation.
6275 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6278 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6279 operated on the sender address. After changing the $sender_address to <>
6280 for the sender address verify, Exim was re-instated it as the original
6281 (before rewriting) address, but remembering that it had rewritten it, so it
6282 wasn't rewriting it again. This bug also had the effect of breaking the
6283 sender address verification caching when the sender address was rewritten.
6285 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6286 This has been changed so that if the ip literal address matches
6287 ignore_target_hosts, the router declines.
6289 7. Added expansion conditions match_domain, match_address, and match_local_
6290 part (NOT match_host).
6292 8. The placeholder for the Received: header didn't have a length field set.
6294 9. Added code to Exim itself and to exim_lock to test for a specific race
6295 condition that could lead to file corruption when using MBX delivery. The
6296 issue is with the lockfile that is created in /tmp. If this file is removed
6297 after a process has opened it but before that process has acquired a lock,
6298 there is the potential for a second process to recreate the file and also
6299 acquire a lock. This could lead to two Exim processes writing to the file
6300 at the same time. The added code performs the same test as UW imapd; it
6301 checks after acquiring the lock that its file descriptor still refers to
6302 the same named file.
6304 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6305 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6308 11. Added the smtp_active_hostname option. If used, this will typically be made
6309 to depend on the incoming interface address. Because $interface_address is
6310 not set up until the daemon has forked a reception process, error responses
6311 that can happen earlier (such as "too many connections") no longer contain
6314 12. If an expansion in a condition on a "warn" statement fails because a lookup
6315 defers, the "warn" statement is abandoned, and the next ACL statement is
6316 processed. Previously this caused the whole ACL to be aborted.
6318 13. Added the iplsearch lookup type.
6320 14. Added ident_timeout as a log selector.
6322 15. Added tls_certificate_verified as a log selector.
6324 16. Added a global option tls_require_ciphers (compare the smtp transport
6325 option of the same name). This controls incoming TLS connections.
6327 17. I finally figured out how to make tls_require_ciphers do a similar thing
6328 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6329 before starting the TLS session.
6331 18. Tabs are now shown as \t in -bP output.
6333 19. If the log selector return_path_on_delivery was set, Exim crashed when
6334 bouncing a message because it had too many Received: header lines.
6336 20. If two routers both had headers_remove settings, and the first one included
6337 a superfluous trailing colon, the final name in the first list and the
6338 first name in the second list were incorrectly joined into one item (with a
6339 colon in the middle).
6345 1. Added -C and -D options to the exinext utility, mainly to make it easier
6346 to include in the automated testing, but these could be helpful when
6347 multiple configurations are in use.
6349 2. The exinext utility was not formatting the output nicely when there was
6350 an alternate port involved in the retry record key, nor when there was a
6351 message id as well (for retries that were specific to a specific message
6352 and a specific host). It was also confused by IPv6 addresses, because of
6353 the additional colons they contain. I have fixed the IPv4 problem, and
6354 patched it up to do a reasonable job for IPv6.
6356 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6357 delivery, the log line now contains "pipelined" if PIPELINING was used.
6359 4. An SMTP transport process used to panic and die if the bind() call to set
6360 an explicit outgoing interface failed. This has been changed; it is now
6361 treated in the same way as a connect() failure.
6363 5. A reference to $sender_host_name in the part of a conditional expansion
6364 that was being skipped was still causing a DNS lookup. This no longer
6367 6. The def: expansion condition was not recognizing references to header lines
6368 that used bh_ and bheader_.
6370 7. Added the _cache feature to named lists.
6372 8. The code for checking quota_filecount in the appendfile transport was
6373 allowing one more file than it should have been.
6375 9. For compatibility with Sendmail, the command line option
6383 and sets the incoming protocol and host name (for trusted callers). The
6384 host name and its colon can be omitted when only the protocol is to be set.
6385 Note the Exim already has two private options, -pd and -ps, that refer to
6386 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6387 "s", but I don't think that's a major issue.
6389 10. A number of refactoring changes to the code, none of which should affect
6392 (a) The number of logging options was getting close to filling up the
6393 32-bit word that was used as a bit map. I have split them into two classes:
6394 those that are passed in the argument to log_write(), and those that are
6395 only ever tested independently outside of that function. These are now in
6396 separate 32-bit words, so there is plenty of room for expansion again.
6397 There is no change in the user interface or the logging behaviour.
6399 (b) When building, for example, log lines, the code previously used a
6400 macro that called string_cat() twice, in order to add two strings. This is
6401 not really sufficiently general. Furthermore, there was one instance where
6402 it was actually wrong because one of the argument was used twice, and in
6403 one call a function was used. (As it happened, calling the function twice
6404 did not affect the overall behaviour.) The macro has been replaced by a
6405 function that can join an arbitrary number of extra strings onto a growing
6408 (c) The code for expansion conditions now uses a table and a binary chop
6409 instead of a serial search (which was left over from when there were very
6410 few conditions). Also, it now recognizes conditions like "pam" even when
6411 the relevant support is not compiled in: a suitably worded error message is
6412 given if an attempt is made to use such a condition.
6414 11. Added ${time_interval:xxxxx}.
6416 12. A bug was causing one of the ddress fields not to be passed back correctly
6417 from remote delivery subprocesses. The field in question was not being
6418 subsequently used, so this caused to problems in practice.
6420 13. Added new log selectors queue_time and deliver_time.
6422 14. Might have fixed a bug in maildirsizefile handling that threw up
6423 "unexpected character" debug warnings, and recalculated the data
6424 unnecessarily. In any case, I expanded the warning message to give more
6427 15. Added the message "Restricted characters in address" to the statements in
6428 the default ACL that block characters like @ and % in local parts.
6430 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6431 Three changes have been made:
6433 (a) There was a serious bug; a negative response to MAIL caused the whole
6434 recipient domain to be cached as invalid, thereby blocking all messages
6435 to all local parts at the same domain, from all senders. This bug has
6436 been fixed. The domain is no longer cached after a negative response to
6437 MAIL if the sender used is not empty.
6439 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6442 (c) A new callout option, "use_sender" has been added for people who want
6443 the modified behaviour.
6449 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6452 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6453 indeed breaks things for older releases.
6455 3. Added additional logging to the case where there is a problem reading data
6456 from a filter that is running in a subprocess using a pipe, in order to
6457 try to track down a specific problem.
6459 4. Testing facility fudge: when running in the test harness and attempting
6460 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6461 getting "No route to host". Convert this to a timeout.
6463 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6466 6. Some OS don't have socklen_t but use size_t instead. This affects the
6467 fifth argument of getsockopt() amongst other things. This is now
6468 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6469 can be set for individual OS. I have set it for SunOS5, OSF1, and
6470 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6471 some earlier ones do not.
6473 7. Change 4.30/15 was not doing the test caselessly.
6475 8. The standard form for an IPv6 address literal was being rejected by address
6476 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6477 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6478 this, as well as the form without the "IPv6" on the front (but only when
6479 address literals are enabled, of course).
6481 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6483 10. Exim crashed if a message with an empty sender address specified by -f
6484 encountered a router with an errors_to setting. This could be provoked only
6485 by a command such as
6489 where an empty string was supplied; "<>" did not hit this bug.
6491 11. Installed PCRE release 4.5.
6493 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6494 remained set. It is now erased.
6496 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6497 times from message ids (which are base 36 rather than the normal 62).
6499 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6500 were being counted as actual protocol errors, and logged if the log
6501 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6502 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6503 and DATA following a set of rejected RCPTs do not count as protocol errors.
6504 In other words, Exim assumes they were pipelined, though this may not
6505 actually be the case. Of course, in all cases the client gets an
6506 appropriate error code.
6508 15. If a lookup fails in an ACL condition, a message about the failure may
6509 be available; it is used if testing the ACL cannot continue, because most
6510 such messages specify what the cause of the deferral is. However, some
6511 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6512 that caused an old message to be retained and used if a later statement
6513 caused a defer, replacing the real cause of the deferral.
6515 16. If an IP address had so many PTR records that the DNS lookup buffer
6516 was not large enough to hold them, Exim could crash while trying to process
6517 the truncated data. It now detects and logs this case.
6519 17. Further to 4.21/58, another change has been made: if (and only if) the
6520 first line of a message (the first header line) ends with CRLF, a bare LF
6521 in a subsequent header line has a space inserted after it, so as not to
6522 terminate the header.
6524 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6525 unnecessarily, used atoi() instead of strtol(), and didn't check the
6526 termination when getting file sizes from file names by regex.
6528 19. Completely re-implemented the support for maildirsize files, in the light
6529 of a number of problems with the previous contributed implementation
6530 (4.30/29). In particular:
6532 . If the quota is zero, the maildirsize file is maintained, but no quota is
6535 . If the maildir directory does not exist, it is created before any attempt
6536 to write a maildirsize file.
6538 . The quota value in the file is just a cache; if the quota is changed in
6539 the transport, the new value overrides.
6541 . A regular expression is available for excluding directories from the
6544 20. The autoreply transport checks the characters in options that define the
6545 message's headers; it allows continued headers, but it was checking with
6546 isspace() after an embedded newline instead of explicitly looking for a
6549 21. If all the "regular" hosts to which an address was routed had passed their
6550 expiry times, and had not reached their retry times, the address was
6551 bounced, even if fallback hosts were defined. Now Exim should go on to try
6554 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6555 equivalent code in the SMTP transport. Some hosts send humungous responses
6556 to HELO/EHLO, more than 1024 it seems.
6558 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6559 gives compiler warnings in some environments. I've now done it "properly",
6562 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6563 (because of problems with the built-in one) was declared to return uschar *
6564 instead of char *, causing compiler failure.
6566 25. Fixed a file descriptor leak when processing alias/forward files.
6568 26. Fixed a minor format string issue in dbfn.c.
6570 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6572 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6573 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6576 29. When a sender address is verified, it is cached, to save repeating the test
6577 when there is more than one recipient in a message. However, when the
6578 verification involves a callout, it is possible for different callout
6579 options to be set for different recipients. It is too complicated to keep
6580 track of this in the cache, so now Exim always runs a verification when a
6581 callout is required, relying on the callout cache for the optimization.
6582 The overhead is duplication of the address routing, but this should not be
6585 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6586 to be verified with callout=postmaster, and the main callout worked but the
6587 postmaster check failed, the verification correctly failed. However, if a
6588 subsequent RCPT command asked for sender verification *without* the
6589 postmaster check, incorrect caching caused this verification also to fail,
6592 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6593 it was not caching the DNS options (qualify_single, search_parents) that
6594 were used when the lookup failed. A subsequent lookup with different
6595 options therefore always gave the same answer, though there were cases
6596 where it should not have. (Example: a "domains = !$mx_any" option on a
6597 dnslookup router: the "domains" option is always processed without any
6598 widening, but the router might have qualify_single set.) Now Exim uses the
6599 cached value only when the same options are set.
6601 32. Added John Jetmore's "exipick" utility to the distribution.
6603 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6604 than a timeout (e.g. a certificate is required, and is not provided), an
6605 Exim server now closes the connection immediately. Previously it waited for
6606 the client to close - but if the client is SSL, it seems that they each
6607 wait for each other, leading to a delay before one of them times out.
6609 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6610 maintained 0.8.x compatibility because I don't think many are using it, and
6611 it is clearly obsolete.
6613 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6616 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6617 client certificate was expired. A simple patch fixes this, though I don't
6618 understand the full logic of why the verify callback is called multiple
6621 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6622 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6623 which causes problems with some clients (such as the Certicom SSL Plus
6624 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6625 disables the coutermeasure allowing Eudora to connect."
6627 38. Exim was not checking that a write() to a log file succeeded. This could
6628 lead to Bad Things if a log got too big, in particular if it hit a file
6629 size limit. Exim now panics and dies if it cannot write to a log file, just
6630 as it does if it cannot open a log file.
6632 39. Modified OS/Makefile-Linux so that it now contains
6634 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6636 The two -D definitions ensure that Exim is compiled with large file
6637 support, which makes it possible to handle log files that are bigger than
6640 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6641 instance) a domain was checked against a named list that involved a lookup,
6642 causing $domain_data to be set, then another domain was checked against the
6643 same list, then the first domain was re-checked, the value of $domain_data
6644 after the final check could be wrong. In particular, if the second check
6645 failed, it could be set empty. This bug probably also applied to
6648 41. The strip_trailing_dot option was not being applied to the address given
6649 with the -f command-line option.
6651 42. The code for reading a message's header from the spool was incrementing
6652 $received_count, but never initializing it. This meant that the value was
6653 incorrect (doubled) while delivering a message in the same process in which
6654 it was received. In the most common configuration of Exim, this never
6655 happens - a fresh exec is done - but it can happen when
6656 deliver_drop_privilege is set.
6658 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6659 it now includes up to 150 characters of the unexpected data in the log
6662 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6663 and building data strings. The size of both of these buffers was 10 000
6664 bytes - far larger than anybody would *ever* want, thought I. Needless to
6665 say, somebody hit the limit. I have increased the maximum line length to
6666 20 000 and the maximum data length of concatenated lines to 100 000. I have
6667 also fixed two bugs, because there was no checking on these buffers. Tsk,
6668 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6669 buffer is too small.
6671 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6672 lsearch lookups. Now it does.
6674 46. When parsing a route_list item in a manualroute router, a fixed-length
6675 buffer was used for the list of hosts. I made this 1024 bytes long,
6676 thinking that nobody would ever have a list of hosts that long. Wrong.
6677 Somebody had a whole pile of complicated expansion conditions, and the
6678 string was silently truncated, leading to an expansion error. It turns out
6679 that it is easier to change to an unlimited length (owing to other changes
6680 that have happened since this code was originally written) than to build
6681 structure for giving a limitation error. The length of the item that
6682 expands into the list of hosts is now unlimited.
6684 47. The lsearch lookup could not handle data where the length of text line was
6685 more than 4095 characters. Such lines were truncated, leading to shortened
6686 data being returned. It should now handle lines of any length.
6688 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6689 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6692 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6693 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6694 no longer happens in comment lines. A list of replacements is now placed
6695 at the head of all of the source files, except those whose only change is
6696 to replace PERL_COMMAND in the very first #! line.
6698 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6699 messages on the queue, with a bottom-up merge sort, using code contributed
6700 by Michael Haardt. This should make operations like -bp somewhat faster on
6701 large queues. It won't affect queue runners, except when queue_run_in_order
6704 51. Installed eximstats 1.31 in the distribution.
6706 52. Added support for SRV lookups to the dnslookup router.
6708 53. If an ACL referred to $message_body or $message_body_end, the value was not
6709 reset for any messages that followed in the same SMTP session.
6711 54. The store-handling optimization for building very long strings was not
6712 differentiating between the different store pools. I don't think this
6713 actually made any difference in practice, but I've tidied it.
6715 55. While running the routers to verify a sender address, $sender_address
6716 was still set to the sender address. This is wrong, because when routing to
6717 send a bounce to the sender, it would be empty. Therefore, I have changed
6718 it so that, while verifying a sender address, $sender_address is set to <>.
6719 (There is no change to what happens when verifying a recipient address.)
6721 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6722 target A or AAAA records (if not already returned) without resetting the
6723 qualify_single or search_parents options of the DNS resolver. These are
6724 inappropriate in this case because the targets of MX and SRV records must
6725 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6726 target that, when qualified, matched something in the local domain. These
6727 two options are now turned off when doing these lookups.
6729 57. It seems that at least some releases of Reiserfs (which does not have the
6730 concept of a fixed number of inodes) returns zero and not -1 for the
6731 number of available inodes. This interacted badly with check_spool_inodes,
6732 which assumed that -1 was the "no such thing" setting. What I have done is
6733 to check that the total number of inodes is greater than zero before doing
6734 the test of how many are available.
6736 58. When a "warn" ACL statement has a log_message modifier, the message is
6737 remembered, and not repeated. This is to avoid a lot of repetition when a
6738 message has many recipients that cause the same warning to be written.
6739 However, Exim was preserving the list of already written lines for an
6740 entire SMTP session, which doesn't seem right. The memory is now reset if a
6741 new message is started.
6743 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6744 debugging output unless log_rewrite was also set.
6746 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6747 of (int)(handle) when we know that handle contains (void *)(-1).
6749 61. The Exim daemon panic-logs an error return when it closes the incoming
6750 connection. However "connection reset by peer" seems to be common, and
6751 isn't really an error worthy of noting specially, so that particular error
6754 62. When Exim is trying to find all the local interfaces, it used to panic and
6755 die if the ioctl to get the interface flags failed. However, it seems that
6756 on at least one OS (Solaris 9) it is possible to have an interface that is
6757 included in the list of interfaces, but for which you get a failure error
6758 for this call. This happens when the interface is not "plumbed" into a
6759 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6760 failure of the "get flags" call assumes that the interface is down.
6762 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6763 makes life easier for people who are doing arithmetic on fields extracted
6764 from dates, where you often get leading zeros that should not be
6765 interpreted as octal.
6767 64. Added qualify_domain to the redirect router, to override the global
6770 65. If a pathologically long header line contained very many addresses (the
6771 report of this problem mentioned 10 000) and each of them was rewritten,
6772 Exim could use up a very large amount of memory. (It kept on making new
6773 copies of the header line as it rewrote, and never released the old ones.)
6774 At the expense of a bit more processing, the header rewriting function has
6775 been changed so that it no longer eats memory in this way.
6777 66. The generation of the Received: header has been moved from the time that a
6778 message starts to be received, to the time that it finishes. The timestamp
6779 in the Received: header should now be very close to that of the <= log
6780 line. There are two side-effects of this change:
6782 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6783 logged header lines no longer include the local Received: line, because
6784 it has not yet been created. The same applies to a copy of the message
6785 that is returned to a non-SMTP sender when a message is rejected.
6787 (b) When a filter file is tested using -bf, no additional Received: header
6788 is added to the test message. After some thought, I decided that this
6791 This change does not affect the value of $received_for. It is still set
6792 after address rewriting, but before local_scan() is called.
6794 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6796 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6797 gave an unhelpful panic error message, and a defer error. I have managed to
6798 change this behaviour so that it now rejects any supplied certificate,
6799 which seems right, as the list of acceptable certificates is empty.
6801 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6802 gave an unhelpful defer error. I have not managed to make this reject any
6803 supplied certificates, but the error message it gives is "no certificate
6804 supplied", which is not helpful.
6806 70. exigrep's output now also includes lines that are not associated with any
6807 message, but which match the given pattern. Implemented by a patch from
6808 Martin Sluka, which also tidied up the Perl a bit.
6810 71. Recipient callout verification, like sender verification, was using <> in
6811 the MAIL FROM command. This isn't really the right thing, since the actual
6812 sender may affect whether the remote host accepts the recipient or not. I
6813 have changed it to use the actual sender in the callout; this means that
6814 the cache record is now keyed on a recipient/sender pair, not just the
6815 recipient address. There doesn't seem to be a real danger of callout loops,
6816 since a callout by the remote host to check the sender would use <>.
6817 [SEE ABOVE: changed after hitting problems.]
6819 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6820 temporary errors. However, in the case of such a code being given after
6821 the end of a data transmission (i.e. after ".") Exim was failing to write
6822 a retry record for the message. (Yes, there was some broken host that was
6823 actually sending 8xx at this point.)
6825 73. An unknown lookup type in a host list could cause Exim to panic-die when
6826 the list was checked. (An example that provoked this was putting <; in the
6827 middle of a list instead of at the start.) If this happened during a DATA
6828 ACL check, a -D file could be left lying around. This kind of configuration
6829 error no longer causes Exim to die; instead it causes a defer error. The
6830 incident is still logged to the main and panic logs.
6832 74. Buglet left over from Exim 3 conversion. The message "too many messages
6833 in one connection" was written to the rejectlog but not the mainlog, except
6834 when address rewriting (yes!) was being logged.
6836 75. Added write_rejectlog option.
6838 76. When a system filter was run not as root (that is, when system_filter_user
6839 was set), the values of the $n variables were not being returned to the
6840 main process; thus, they were not subsequently available in the $sn
6843 77. Added +return_path_on_delivery log selector.
6845 78. A connection timeout was being treated differently from recipients deferred
6846 when testing hosts_max_try with a message that was older than the host's
6847 retry timeout. (The host should not be counted, thus allowing all hosts to
6848 be tried at least once before bouncing.) This may have been the cause of an
6849 occasionally reported bug whereby a message would remain on the queue
6850 longer than the retry timeout, but would be bounced if a delivery was
6851 forced. I say "may" because I never totally pinned down the problem;
6852 setting up timeout/retry tests is difficult. See also the next item.
6854 79. The ultimate address timeout was not being applied to errors that involved
6855 a combination of host plus message (for example, a timeout on a MAIL
6856 command). When an address resolved to a number of possible hosts, and they
6857 were not all tried for each delivery (e.g. because of hosts_max_try), a
6858 message could remain on the queue longer than the retry timeout.
6860 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6863 81. Fixed an obscure SMTP outgoing bug which required at least the following
6864 conditions: (a) there was another message waiting for the same server;
6865 (b) the server returned 5xx to all RCPT commands in the first message so
6866 that the message was not completed; (c) the server dropped the connection
6867 or gave a negative response to the RSET that Exim sends to abort the
6868 transaction. The observed case was a dropped connection after DATA that had
6869 been sent in pipelining mode. That is, the server had advertised PIPELINING
6870 but was not implementing it correctly. The effect of the bug was incorrect
6871 behaviour, such as trying another host, and this could lead to a crash.
6877 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6878 and daemon.c were passed as pointers to ints; they should have been
6879 pointers to socklen_t variables (which are typically unsigned ints).
6881 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6884 3. Fixed a really odd bug that affected only the testing scheme; patching a
6885 certain fixed string in the binary changed the value of another string that
6886 happened to be identical to the end of the original first string.
6888 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6889 name", it returns that address as the IP address. On some operating
6890 systems (e.g. Solaris), it also passes back the IP address string as the
6891 "host name". However, on others (e.g. Linux), it passes back an empty
6892 string. Exim wasn't checking for this, and was changing the host name to an
6893 empty string, assuming it had been canonicalized.
6895 5. Although rare, it is permitted to have more than one PTR record for a given
6896 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6897 all the names associated with an address, because they do in Solaris.
6898 However, it seems that they do not in Linux for data that comes from the
6899 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6900 I found this out when I moved to a new Linux workstation and tried to run
6901 the Exim test suite.
6903 To get round this problem I have changed the code so that it now does its
6904 own call to the DNS to look up PTR records when searching for a host name.
6905 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6906 addresses that are only in /etc/hosts are still found.
6908 This behaviour is, however, controlled by an option called host_lookup_
6909 order, which defaults to "bydns:byaddr". If people want to use the other
6910 order, or indeed, just use one or the other means of lookup, they can
6911 specify it in this variable.
6913 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6914 some operating systems, this comes back from gethostbyaddr() as an empty
6915 string, and this is what Exim used to test for. However, it seems that in
6916 other systems, "." is yielded. Exim now tests for this case too.
6918 7. The values of check_spool_space and check_log_space are now held internally
6919 as a number of kilobytes instead of an absolute number of bytes. If a
6920 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6921 kilobyte. This means that much larger values can be stored.
6923 8. Exim monitor: an attempt to get the action menu when not actually pointing
6924 at a message produces an empty menu entitled "No message selected". This
6925 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6926 no entries in it ("Shell widget menu has zero width and/or height"). So I
6927 have added a single, blank menu entry in this case.
6929 9. Added ${quote_local_part.
6931 10. MIME decoding is now applied to the contents of Subject: header lines when
6934 11. Now that a reference to $sender_host_address automatically causes a reverse
6935 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6936 host lookup before query-style lookups in lists that might use this
6937 variable. This has therefore been abolished, and the "net-" prefix is no
6938 longer necessary for query-style lookups.
6940 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6941 have been a typo for LFLAGS, so it has been changed.
6943 13. The install script calls Exim with "-C /dev/null" in order to find the
6944 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6945 to be output. However, since Exim outputs its version number before the
6946 error, it didn't break the script. It just looked ugly. I fixed this by
6947 always allowing "-C /dev/null" if the caller is root.
6949 14. Ignore overlarge ACL variable number when reading spool file - insurance
6950 against a later release with more variables having written the file.
6952 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6953 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6954 this, as well as the form without the "IPv6" on the front.
6956 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6957 OS/Makefile-Darwin file.
6959 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6960 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6962 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6963 variable. However, some OS use ints for inodes. I've added cast to long int
6964 to get rid of the compiler warning.
6966 19. I had forgotten to lock out "/../" in configuration file names when
6967 ALT_CONFIG_PREFIX was set.
6969 20. Routers used for verification do not need to specify transports. However,
6970 if such a router generated a host list, and callout was configured, Exim
6971 crashed, because it could not find a port number from the (non-existent)
6972 transport. It now assumes port 25 in this circumstance.
6974 21. Added the -t option to exigrep.
6976 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6977 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6978 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6979 them is set, LOOKUP_LSEARCH is forced.
6981 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6983 24. Added sender and host information to the "rejected by local_scan()" log
6984 line; previously there was no indication of these.
6986 25. Added .include_if_exists.
6988 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6989 Linux. It turns out that not all file systems support this. Apparently some
6990 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6991 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6992 not supported on the file descriptor, is now ignored when Exim is trying to
6993 sync a directory. This applies only to Linux.
6995 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6997 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6998 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7001 29. Added support for maildirsize files from supplied patch (modified a bit).
7003 30. The use of :fail: followed by an empty string could lead Exim to respond to
7004 sender verification failures with (e.g.):
7006 550 Verification failed for <xxx>
7007 550 Sender verify failed
7009 where the first response line was missing the '-' that indicates it is not
7010 the final line of the response.
7012 31. The loop for finding the name of the user that called Exim had a hardwired
7013 limit of 10; it now uses the value of finduser_retries, which is used for
7014 all other user lookups.
7016 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7019 33. Exim was neglecting to zero errno before one call of strtol() when
7020 expanding a string and expecting an integer value. On some systems this
7021 resulted in spurious "integer overflow" errors. Also, it was casting the
7022 result into an int without checking.
7024 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7025 did not work. The code looks as if it has *never* worked, though it appears
7026 to have been documented since at least release 1.62. I have made it work.
7028 35. The "timeout_DNS" error in retry rules, also documented since at least
7029 1.62, also never worked. As it isn't clear exactly what this means, and
7030 clearly it isn't a major issue, I have abolished the feature by treating it
7031 as "timeout", and writing a warning to the main and panic logs.
7033 36. The display of retry rules for -brt wasn't always showing the error code
7036 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7037 timeout_connect_A, timeout_connect_MX.
7039 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7040 to the empty sender.
7042 39. The daemon was not analysing the content of -oX till after it had closed
7043 stderr and disconnected from the controlling terminal. This meant that any
7044 syntax errors were only noted on the panic log, and the return code from
7045 the command was 0. By re-arranging the code a little, I've made the
7046 decoding happen first, so such errors now appear on stderr, and the return
7047 code is 1. However, the actual setting up of the sockets still happens in
7048 the disconnected process, so errors there are still only recorded on the
7051 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7052 connections (as happens on some IP stacks) was logged at start up time as
7053 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7054 it from "IPv6 and IPv4", which means that two separate sockets are being
7057 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7058 says whether AF_INET or AF_INET6 was passed as an argument.
7060 42. Exiwhat output was messed up when time zones were included in log
7063 43. Exiwhat now gives more information about the daemon's listening ports,
7064 and whether -tls-on-connect was used.
7066 44. The "port" option of the smtp transport is now expanded.
7068 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7069 silently ignored. Now an error message is written to the main and panic
7072 46. There's a new ACL modifier called "logwrite" which writes to a log file
7073 as soon as it is encountered.
7075 47. Added $local_user_uid and $local_user_gid at routing time.
7077 48. Exim crashed when trying to verify a sender address that was being
7080 49. Exim was recognizing only a space character after ".include". It now also
7081 recognizes a tab character.
7083 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7084 extracting the relevant information from the specification. The man page no
7085 longer contains scrambled data for the -d option, and I've added a section
7086 at the front about calling Exim under different names.
7088 51. Added "extra_headers" argument to the "mail" command in filter files.
7090 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7093 53. Installed eximstats 1.29.
7095 54. Added transport_filter_timeout as a generic transport option.
7097 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7098 Cc: header lines. This was required by RFC 822, but it not required by RFC
7101 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7102 incoming messages that did not have them. Now it does so only if the
7103 message originates locally, that is, if there is no associated remote host
7104 address. When Resent- header lines are present, this applies to the Resent-
7105 lines rather than the non-Resent- lines.
7107 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7108 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7110 58. Messages for configuration errors now include the name of the main
7111 configuration file - useful now that there may be more than one file in a
7112 list (.included file names were always shown).
7114 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7115 for those rare installations that do not start the daemon as root or run it
7116 setuid root. I've cut out the call to initgroups() if the daemon is not
7119 60. The Exim user and group can now be bound into the binary as text strings
7120 that are looked up at the start of Exim's processing.
7122 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7124 62. Added $mailstore_basename variable.
7126 63. Installed patch to sieve.c from Michael Haardt.
7128 64. When Exim failed to open the panic log after failing to open the main log,
7129 the original message it was trying to log was written to stderr and debug
7130 output, but if they were not available (the usual case in production), it
7131 was lost. Now it is written to syslog before the two lines that record the
7132 failures to open the logs.
7134 65. Users' Exim filters run in subprocesses under the user's uid. It is
7135 possible for a "deliver" command or an alias in a "personal" command to
7136 provoke an address rewrite. If logging of address rewriting is configured,
7137 this fails because the process is not running as root or exim. There may be
7138 a better way of dealing with this, but for the moment (because 4.30 needs
7139 to be released), I have disabled address rewrite logging when running a
7140 filter in a non-root, non-exim process.
7146 1. The buildconfig auxiliary program wasn't quoting the value set for
7147 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7148 not defined. This bug was masked in 4.22 by the effect that was fixed in
7151 2. Some messages that were rejected after a message id was allocated were
7152 shown as "incomplete" by exigrep. It no longer does this for messages that
7153 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7155 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7156 have allow_domain_literals set, the ID did not get logged in the <= line.
7157 Domain literals are now always recognized in Message-ID: header lines.
7159 4. The first argument for a ${extract expansion item is the key name or field
7160 number. Leading and trailing spaces in this item were not being ignored,
7161 causing some misleading effects.
7163 5. When deliver_drop_privilege was set, single queue runner processes started
7164 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7165 same command in the process it spins off) were not dropping privilege.
7167 6. When the daemon running as "exim" started a queue runner, it always
7168 re-executed Exim in the spun-off process. This is a waste of effort when
7169 deliver_drop_privilege is set. The new process now just calls the
7170 queue-runner function directly.
7176 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7179 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7180 ignored. Though the use of -oP was forcing the writing of a pid file, it
7181 was always written to the default place.
7183 3. If the message "no IP address found for host xxxx" is generated during
7184 incoming verification, it is now followed by identification of the incoming
7185 connection (so you can more easily find what provoked it).
7187 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7189 5. Added some features to "harden" Exim a bit more against certain attacks:
7191 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7192 be put in Local/Makefile. This is like the never_users runtime option,
7193 but it cannot be overridden. The default setting is "root".
7195 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7196 prefix string with which any file named in a -C command line option
7199 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7200 is retained for -C and -D only if the caller of Exim is root. Without
7201 it, the exim user may also use -C and -D and retain privilege.
7203 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7204 command line option is disabled.
7206 6. Macro names set by the -D option must start with an upper case letter, just
7207 like macro names defined in the configuration file.
7209 7. Added "dereference=" facility to LDAP.
7211 8. Two instances of the typo "uknown" in the source files are fixed.
7213 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7214 the Configure-Makefile script screwed up while processing it.
7216 10. Incorporated PCRE 4.4.
7218 11. The SMTP synchronization check was not operating right at the start of an
7219 SMTP session. For example, it could not catch a HELO sent before the client
7220 waited for the greeting. There is now a check for outstanding input at the
7221 point when the greeting is written. Because of the duplex, asynchronous
7222 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7223 way, but not yet received, when the check is performed.
7225 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7226 on TCP/IP sockets, because this apparently causes some broken clients to
7229 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7230 unchanged) from the Cygwin maintainer.
7232 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7233 was supported instead of testing for mbx. Effectively a typo.
7235 15. The spa authenticator server code was not checking that the input it
7236 received was valid base64.
7238 16. The debug output line for the "set" modifier in ACLs was not showing the
7239 name of the variable that was being set.
7241 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7243 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7244 lookup on its own. Something else was needed to trigger the lookup. For
7245 example, a match in host_lookup or the need for a host name in a host list.
7246 Now, if $sender_host_name is referenced and the host name has not yet been
7247 looked up, a lookup is performed. If the lookup fails, the variable remains
7248 empty, and $host_lookup_failed is set to "1".
7250 19. Added "eqi" as a case-independent comparison operator.
7252 20. The saslauthd authentication condition could segfault if neither service
7253 nor realm was specified.
7255 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7256 error message that was logged was misleading, and incoming SMTP
7257 connections were dropped. The message is now more accurate, and temporary
7258 errors are given to SMTP connections.
7260 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7261 (except RSET) with a 421 error, until QUIT is received. However, it was
7262 failing to send a response to QUIT.
7264 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7265 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7266 ACL is now run after the helo_try_verify_hosts code.
7268 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7269 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7270 case-independent, so other case variants are also recognized). Apparently
7271 some systems use these upper case variants.
7273 25. If more than two messages were waiting for the same host, and a transport
7274 filter was specified for the transport, Exim sent two messages over the
7275 same TCP/IP connection, and then failed with "socket operation on non-
7276 socket" when it tried to send the third.
7278 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7280 27. The extern definition of crypt16() in expand.c was not being excluded when
7281 the OS had its own crypt16() function.
7283 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7284 as a preferred synonym for return_size_limit, both as an option and as an
7287 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7289 30. Changed the default configuration ACL to relax the local part checking rule
7290 for addresses that are not in any local domains. For these addresses,
7291 slashes and pipe symbols are allowed within local parts, but the sequence
7292 /../ is explicitly forbidden.
7294 31. SPA server authentication was not clearing the challenge buffer before
7297 32. log_message in a "warn" ACL statement was writing to the reject log as
7298 well as to the main log, which contradicts the documentation and doesn't
7299 seem right (because no rejection is happening). So I have stopped it.
7301 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7302 However, I am unable to do any testing of this.
7304 34. Fixed an infelicity in the appendfile transport. When checking directories
7305 for a mailbox, to see if any needed to be created, it was accidentally
7306 using path names with one or more superfluous leading slashes; tracing
7307 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7309 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7310 contents are added to the log line that is written for every discarded
7311 recipient. (Previously a log_message setting was ignored.)
7313 36. The ${quote: operator now quotes the string if it is empty.
7315 37. The install script runs exim in order to find its version number. If for
7316 some reason other than non-existence or emptiness, which it checks, it
7317 could not run './exim', it was installing it with an empty version number,
7318 i.e. as "exim-". This error state is now caught, and the installation is
7321 38. An argument was missing from the function that creates an error message
7322 when Exim fails to connect to the socket for saslauthd authentication.
7323 This could cause Exim to crash, or give a corrupted message.
7325 39. Added isip, isip4, and isip6 to ${if conditions.
7327 40. The ACL variables $acl_xx are now saved with the message, and can be
7328 accessed later in routers, transports, and filters.
7330 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7331 strings in the file are not string-expanded.
7333 42. If a MAIL command specified a SIZE value that was too large to fit into an
7334 int variable, the check against message_size_limit failed. Such values are
7335 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7336 day this will have to be increased, but I don't think I want to be around
7337 when emails are that large.
7344 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7345 iconv() is not standard in FreeBSD.
7347 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7348 IPv6 enabled. The observed symptom was a segmentation fault on return from
7349 the function os_common_find_running_interfaces() in src/os.c.
7351 3. In the check_special_case() function in daemon.c I had used "errno" as an
7352 argument name, which causes warnings on some systems. This was basically a
7353 typo, since it was named "eno" in the comments!
7355 4. The code that waits for the clock to tick (at a resolution of some fraction
7356 of a second) so as to ensure message-id uniqueness was always waiting for
7357 at least one whole tick, when it could have waited for less. [This is
7358 almost certainly not relevant at current processor speeds, where it is
7359 unlikely to ever wait at all. But we try to future-proof.]
7361 5. The function that sleeps for a time interval that includes fractions of a
7362 second contained a race. It did not block SIGALRM between setting the
7363 timer, and suspending (a couple of lines later). If the interval was short
7364 and the sigsuspend() was delayed until after it had expired, the suspension
7365 never ended. On busy systems this could lead to processes getting stuck for
7368 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7369 process, before it forks any delivery processes. The open lookup caching
7370 mechanism meant that the open file or database connection was passed into
7371 the delivery process. The problem was that delivery processes always tidy
7372 up cached lookup data. This could cause a problem for the next delivery
7373 process started by the queue runner, because the external queue runner
7374 process does not know about the closure. So the next delivery process
7375 still has data in the lookup cache. In the case of a file lookup, there was
7376 no problem because closing a file descriptor in a subprocess doesn't affect
7377 the parent. However, if the lookup was caching a connection to a database,
7378 the connection was closed, and the second delivery process was likely to
7379 see errors such as "PGSQL: query failed: server closed the connection
7380 unexpectedly". The problem has been fixed by closing all cached lookups
7381 in a queue runner before running a delivery process.
7383 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7384 seem to have the "const" qualifier which it has on other OS. I've
7387 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7388 *of the same type* (client or server) with the same public name that an
7389 error should be diagnosed.
7391 9. When Exim looked up a host name for an IP address, but failed to find the
7392 original IP address when looking up the host name (a safety check), it
7393 output the message "<ip address> does not match any IP for NULL", which was
7394 confusing, to say the least. The bug was that the host name should have
7395 appeared instead of "NULL".
7397 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7398 user that is built into the binary, and the -C or -D options is used, root
7399 privilege is dropped before the configuration file is read. In addition,
7400 logging is switched to stderr instead of the normal log files. If the
7401 configuration then re-defines the Exim user, the unprivileged environment
7402 is probably not what is expected, so Exim logs a panic warning message (but
7405 However, if deliver_drop_privilege is set, the unprivileged state may well
7406 be exactly what is intended, so the warning has been cut out in that case,
7407 and Exim is allowed to try to write to its normal log files.
7413 1. smtp_return_error_details was not giving details for temporary sender
7414 or receiver verification errors.
7416 2. Diagnose a configuration error if two authenticators have the same public
7419 3. Exim used not to create the message log file for a message until the first
7420 delivery attempt. This could be confusing when incoming messages were held
7421 for policy or load reasons. The message log file is now created at the time
7422 the message is received, and an initial "Received" line is written to it.
7424 4. The automatically generated man page for command line options had a minor
7425 bug that caused no ill effects; however, a more serious problem was that
7426 the procedure for building the man page automatically didn't always
7427 operate. Consequently, release 4.20 contains an out-of-date version. This
7428 shouldn't happen again.
7430 5. When building Exim with embedded Perl support, the script that builds the
7431 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7432 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7434 6. The freeze_tell option was not being used for messages that were frozen on
7435 arrival, either by an ACL or by local_scan().
7437 7. Added the smtp_incomplete_transaction log selector.
7439 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7440 was accepting AUTH without a new EHLO.
7442 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7443 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7446 10. Logging of TCP/IP connections (when configured) now happens in the main
7447 daemon process instead of the child process, so that the TCP/IP connection
7448 count is more accurate (but it can never be perfect).
7450 11. The use of "drop" in a nested ACL was not being handled correctly in the
7451 outer ACL. Now, if condition failure induced by the nested "drop" causes
7452 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7453 or "require"), the connection is dropped.
7455 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7456 that yield "discard" can now be used with an "accept" or a "discard" verb,
7457 but an error is generated for any others (because I can't see a useful way
7458 to define what should happen).
7460 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7461 are now processed in the same way as lines in the Exim configuration file.
7462 In particular, continuation lines are supported.
7464 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7466 15. Added -ti meaning -t -i.
7468 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7469 domains, and warn by logging if others are found.
7471 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7472 returned from GIFCONF when Exim is trying to find the list of interfaces on
7473 a host. The code in os.c has been modified to copy each ifreq to an aligned
7474 structure in all cases.
7476 Also, in some cases, the returned ifreq's were being copied to a 'struct
7477 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7478 means the last couple of bytes of an IPv6 address could be chopped if the
7479 ifreq contained only a normal sockaddr (14 bytes storage).
7481 18. Named domain lists were not supported in the hosts_treat_as_local option.
7482 An entry such as +xxxx was not recognized, and was treated as a literal
7485 19. Ensure that header lines added by a DATA ACL are included in the reject log
7486 if the ACL subsequently rejects the message.
7488 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7489 MD5 (which is deprecated).
7491 21. When testing a filter file using -bf, Exim was writing a message when it
7492 took the sender from a "From " line in the message, but it was not doing so
7493 when it took $return_path from a Return-Path: header line. It now does.
7495 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7496 with a valid header line field name (a series of printing characters
7497 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7499 23. Changed "disc" in the source to "disk" to conform to the documentation and
7500 the book and for uniformity.
7502 24. Ignore Sendmail's -Ooption=value command line item.
7504 25. When execve() failed while trying to run a command in a pipe transport,
7505 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7506 could be confused with a return value of 69 from the command itself. This
7507 has been changed to 127, the value the shell returns if it is asked to run
7508 a non-existent command. The wording for the related log line suggests a
7509 non-existent command as the problem.
7511 26. If received_header_text expands to an empty string, do not add a Received:
7512 header line to the message. (Well, it adds a token one on the spool, but
7513 marks it "old" so that it doesn't get used or transmitted.)
7515 27. Installed eximstats 1.28 (addition of -nt option).
7517 28. There was no check for failure on the call to getsockname() in the daemon
7518 code. This can fail if there is a shortage of resources on the system, with
7519 ENOMEM, for example. A temporary error is now given on failure.
7521 29. Contrary to the C standard, it seems that in some environments, the
7522 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7523 program. Exim now does this explicitly; it affects the formatting of
7524 timestamps using strftime().
7526 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7527 complaints. I've now initialized all the variables, to avoid this.
7529 32. Header lines added by a system filter were not being "seen" during
7530 transport-time rewrites.
7532 33. The info_callback() function passed to OpenSSL is set up with type void
7533 (*)(SSL *, int, int), as described somewhere. However, when calling the
7534 function (actually a macro) that sets it up, the type void(*)() is
7535 expected. I've put in a cast to prevent warnings from picky compilers.
7537 34. If a DNS black list lookup found a CNAME record, but there were no A
7538 records associated with the domain it pointed at, Exim crashed.
7540 35. If a DNS black list lookup returned more than one A record, Exim ignored
7541 all but the first. It now scans all returned addresses if a particular IP
7542 value is being sought. In this situation, the contents of the
7543 $dnslist_value variable are a list of all the addresses, separated by a
7546 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7547 to accept a host name and all its aliases if the forward lookup for any of
7548 them yielded the IP address of the incoming connection. Now it accepts only
7549 those names whose forward lookup yields the correct IP address. Any other
7550 names are discarded. This closes a loophole whereby a rogue DNS
7551 administrator could create reverse DNS records to break through a
7552 wildcarded host restriction in an ACL.
7554 37. If a user filter or a system filter that ran in a subprocess used any of
7555 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7556 the wrong values were passed to the pipe command ($thisaddress had the
7557 value of $0, $0 had the value of $1, etc). This bug was introduced by
7558 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7560 38. Improved the line breaking for long SMTP error messages from ACLs.
7561 Previously, if there was no break point between 40 and 75 characters, Exim
7562 left the rest of the message alone. Two changes have been made: (a) I've
7563 reduced the minimum length to 35 characters; (b) if it can't find a break
7564 point between 35 and 75 characters, it looks ahead and uses the first one
7565 that it finds. This may give the occasional overlong line, but at least the
7566 remaining text gets split now.
7568 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7569 file descriptors might be low, and that setting 1000 would always raise it.
7570 It turns out that in some environments, the limit is already over 1000 and
7571 that lowering it causes trouble. So now Exim takes care not to decrease it.
7573 40. When delivering a message, the value of $return_path is set to $sender_
7574 address at the start of routing (routers may change the value). By an
7575 oversight, this default was not being set up when an address was tested by
7576 -bt or -bv, which affected the outcome if any router or filter referred to
7579 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7580 log only when "message" or "log_message" are set. However, if one of the
7581 conditions was an address verification, or a call to a nested ACL, the
7582 messages generated by the underlying test were being passed through. This
7583 no longer happens. The underlying message is available in $acl_verify_
7584 message for both "message" and "log_message" expansions, so it can be
7585 passed through if needed.
7587 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7588 new expansion $bh_ to give the encoded byte string without charset
7589 translation. Translation happens only if iconv() is available; HAVE_ICONV
7590 indicates this at build time. HEADERS_CHARSET gives the charset to
7591 translate to; headers_charset can change it in the configuration, and
7592 "headers charset" can change it in an individual filter file.
7594 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7595 that creates RFC 2047 encoded "words" labels them as that charset instead
7596 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7597 expansion operator; (ii) when Exim creates a From: line for a local
7598 message; (iii) when a header line is rewritten to include a "phrase" part.
7600 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7601 buggy, causing it to skip the first lines of messages whose message ID
7602 ended in 'D'. This would not have bitten before Exim release 4.14, because
7603 message IDs were unlikely to end in 'D' before then. The effect was to have
7604 incorrect size information for certain domains.
7606 45. #include "config.h" was missing at the start of the crypt16.c module. This
7607 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7610 46. If there was a timeout during a "random" callout check, Exim treated it as
7611 a failure of the random address, and carried on sending RSET and the real
7612 address. If the delay was just some slowness somewhere, the response to the
7613 original RCPT would be taken as a response to RSET and so on, causing
7614 mayhem of various kinds.
7616 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7617 when I implemented it. It didn't allow for the fact that some option values
7618 may legitimately be negative (e.g. size_addition), and it didn't even do
7619 the right test for positive values.
7621 48. Domain names in DNS records are case-independent. Exim always looks them up
7622 in lower case. Some resolvers return domain names in exactly the case they
7623 appear in the zone file, that is, they may contain uppercase letters. Not
7624 all resolvers do this - some return always lower case. Exim was treating a
7625 change of case by a resolver as a change of domain, similar to a widening
7626 of a domain abbreviation. This triggered its re-routing code and so it was
7627 trying to route what was effectively the same domain again. This normally
7628 caused routing to fail (because the router wouldn't handle the domain
7629 twice). Now Exim checks for this case specially, and just changes the
7630 casing of the domain that it ultimately uses when it transmits the message
7633 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7636 50. If a filter generated a file delivery with a non-absolute name (possible if
7637 no home directory exists for the router), the forbid_file option was not
7640 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7641 the existing equality matching.
7643 52. Exim was using ints instead of ino_t variables in some places where it was
7644 dealing with inode numbers.
7646 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7647 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7648 called TMPDIR, and if it finds it is different, it changes its value.
7650 54. The smtp_printf() function is now made available to local_scan() so
7651 additional output lines can be written before returning. There is also an
7652 smtp_fflush() function to enable the detection of a dropped connection.
7653 The variables smtp_input and smtp_batched_input are exported to
7656 55. Changed the default runtime configuration: the message "Unknown user"
7657 has been removed from the ACL, and instead placed on the localuser router,
7658 using the cannot_route_message feature. This means that any verification
7659 failures that generate their own messages won't get overridden. Similarly,
7660 the "Unrouteable address" message that was in the ACL for unverifiable
7661 relay addresses has also been removed.
7663 56. Added hosts_avoid_esmtp to the smtp transport.
7665 57. The exicyclog script was not checking for the esoteric option
7666 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7667 will work only if exicyclog is run under the appropriate euid.
7669 58. Following a discussion on the list, the rules by which Exim recognises line
7670 endings on incoming messages have been changed. The -dropcr and drop_cr
7671 options are now no-ops, retained only for backwards compatibility. The
7672 following line terminators are recognized: LF CRLF CR. However, special
7673 processing applies to CR:
7675 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7676 nor a local message in the state where . is a terminator.
7678 (ii) If a bare CR is encountered in a header line, an extra space is added
7679 after the line terminator so as not to end the header. The reasoning
7680 behind this is that bare CRs in header lines are most likely either
7681 to be mistakes, or people trying to play silly games.
7683 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7684 was being incorrectly given as 18 bytes larger than it should have been.
7685 This is a VOB (very old bug).
7687 60. This may never have affected anything current, but just in case it has:
7688 When the local host is found other than at the start of a list of hosts,
7689 the local host, those with the same MX, and any that follow, are discarded.
7690 When the list in question was part of a longer list of hosts, the following
7691 hosts (not currently being processed) were also being discarded. This no
7692 longer happens. I'm not sure if this situation could ever has previously
7695 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7698 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7699 header. This is something that is recommended in a new Internet Draft, and
7700 is something that is documented as being done by Sendmail. There are two
7701 possible values. For messages generated by the autoreply transport, Exim
7704 Auto-Submitted: auto-replied
7706 whereas for all other generated messages (e.g. bounces) it adds
7708 Auto-Submitted: auto-generated
7710 63. The "personal" condition in filters now includes a test for the
7711 Auto-Submitted: header. If it contains the string "auto-" the message it
7712 not considered personal.
7714 64. Added rcpt_include_affixes as a generic transport option.
7716 65. Added queue_only_override (default true).
7718 66. Added the syslog_duplication option.
7720 67. If what should have been the first header line of a message consisted of
7721 a space followed by a colon, Exim was mis-interpreting it as a header line.
7722 It isn't of course - it is syntactically invalid and should therefore be
7723 treated as the start of the message body. The misbehaviour could have
7724 caused a number of strange effects, including loss of data in subsequent
7725 header lines, and spool format errors.
7727 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7728 client host had authenticated. This control can now be exercised by an ACL
7729 for more flexibility.
7731 69. By default, callouts do not happen when testing with -bh. There is now a
7732 variant, -bhc, which does actually run the callout code, including
7733 consulting and updating the callout cache.
7735 70. Added support for saslauthd authentication, courtesy of Alexander
7738 71. If statvfs() failed on the spool or log directories while checking their
7739 size for availability, Exim confusingly gave the error "space shortage".
7740 Furthermore, in debugging mode it crashed with a floating point exception.
7741 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7742 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7743 serious problem, Exim now writes to the main and panic logs when this
7744 happens, with details of the failure. It then refuses to accept the
7745 incoming message, giving the message "spool directory problem" or "log
7746 directory problem" with a 421 code for SMTP messages.
7748 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7749 0, 1, and 2 exist, because some OS complain for execs without them (see
7750 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7751 descriptors. However, the code omitted to check that the open succeeded,
7752 causing mysterious errors if for some reason the permissions on /dev/null
7753 got screwed. Now Exim writes a message to the main and panic logs, and
7754 bombs out if it can't open /dev/null.
7756 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7757 interact so that it is all more flexible. It is supposed to remain
7758 backwards compatible. Also added extra_local_interfaces.
7760 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7761 to bomb out with an assertion failure - to the client this appears as a
7762 connection drop. This problem occurs in the part of the code that was taken
7763 from the Samba project. Fortunately, the assertion is in a very simple
7764 function, so I have fixed this by reproducing the function inline in the
7765 one place where it is called, and arranging for authentication to fail
7766 instead of killing the process with assert().
7768 75. The SPA client code was not working when the server requested OEM rather
7769 than Unicode encoding.
7771 76. Added code to make require_files with a specific uid setting more usable in
7772 the case where statting the file as root fails - usually a non-root-mounted
7773 NFS file system. When this happens and the failure is EACCES, Exim now
7774 forks a subprocess and does the per-uid checking as the relevant uid.
7776 77. Added process_log_path.
7778 78. If log_file_path was not explicitly set, a setting of check_log_space or
7779 check_log_inodes was ignored.
7781 79. If a space check for the spool or log partitions fails, the incident is now
7782 logged. Of course, in the latter case the data may get lost...
7784 80. Added the %p formatting code to string_format() so that it can be used to
7785 print addresses in debug_print(). Adjusted all the address printing in the
7786 debugging in store.c to use %p rather than %d.
7788 81. There was a concern that a line of code in smtp_in.c could overflow a
7789 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7790 initially expressed, the concern was not well-founded, because trailing
7791 spaces are removed early. However, if the trailing spaces were followed by
7792 a NULL, they did not get removed, so the overflow was possible. Two fixes
7795 (a) I re-wrote the offending code in a cleaner fashion.
7796 (b) If an incoming SMTP command contains a NULL character, it is rejected
7799 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7800 runs initgroups(), so that if the Exim user is in any additional groups,
7801 they will be used during message reception.
7807 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob