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.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
42 JH/01 Avoid costly startup code when not strictly needed. This reduces time
43 for some exim process initialisations. It does mean that the logging
44 of TLS configuration problems is only done for the daemon startup.
46 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
48 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
49 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
52 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
53 DKIM signing is being done. Previously a transport_filter would always
54 disable CHUNKING, falling back to traditional DATA.
56 JH/05 Regard command-line receipients as tainted.
58 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
60 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
61 PAM library frees one of the arguments given to it, despite the
62 documentation. Therefore a plain malloc must be used.
64 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
65 on-stack buffers were used, resulting in a taint trap when DSN information
66 copied from a received message was written into the buffer.
68 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
69 the ordering of its ARC headers. This caused a crash.
71 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
72 a new record was being constructed with information from the peer, a trap
75 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
76 installation would get error messages from DMARC verify, when it hit the
77 nonexistent file indicated by the default. Distros wanting DMARC enabled
78 should both provide the file and set the option.
79 Also enforce no DMARC verification for command-line sourced messages.
81 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
82 could, depending on the platform, hang at the STARTTLS response.
84 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
85 message on a connection. Previously if one message had ARC headers and
86 the following one did not, a crash could result when adding an
87 Authentication-Results: header.
89 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
90 Exim main code and Exim-related utities. The introduction of taint
91 tracking also did many adjustments to string handling. Since then, eximon
92 frequently terminated with an assert failure.
94 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
95 check for 452 responses. This slightly helps the inefficieny of doing
96 a large alias-expansion into a recipient-limited target. The max_rcpt
97 transport option still applies (and at the current default, will override
98 the new feature). The check is done for either cause of synch, and forces
99 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
100 connection. The new facility is not tunable at this time.
102 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
103 library live data was being used, so the results became garbage. Make
104 copies while it is still usable.
106 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
107 on delivery deferred (==) and failed (**) lines (if a delivery was
108 attemtped). Previously it was only on completion (=>) lines.
110 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
111 for the expansion of the server_scram_iter and server_scram_salt options.
113 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
114 are now specifically given a NO_DATA response without hitting the system
115 resolver. The library goes on to do the now-standard TXT lookup.
116 Use of dnsdb lookups is not affected.
118 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
119 only retrieve the errormessage once. Previously two calls to dlerror()
120 were used, and the second one (for mainlog/paniclog) retrieved null
123 JH/20 Taint checking: disallow use of tainted data for
124 - the appendfile transport file and directory options
125 - the pipe transport command
126 - the autoreply transport file, log and once options
127 - file names used by the redirect router (including filter files)
129 - paths used by single-key lookups
130 Previously this was permitted.
132 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
133 adjusted the size of a major service buffer; this failed because the
134 buffer was in use at the time. Change to a compile-time increase in the
135 buffer size, when this authenticator is compiled into exim.
137 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
138 previous fast-mode was untenable in the face of glibs using mmap to
139 support larger malloc requests.
141 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
142 New values supported, if defined on system where compiled:
143 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
144 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
146 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
147 running a limited number of proceses in parallel, a benefit is gained. The
148 amount varies with the platform hardware and load. The use of the option
149 queue_run_in_order means we cannot do this, as ordering becomes
152 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
153 had introduced a string-copy (for ensuring NUL-termination) which was not
154 appropriate for that case, which can include embedded NUL bytes in the
155 block of data. Investigation showed the copy to actually be needless, the
156 data being length-specified.
158 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
159 done during a receiving connection, and both used TLS, global info was
160 used rather than per-connection info for tracking the state of data
161 queued for transmission. This could result in a connection hang.
163 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
164 Previously, when delivering serveral messages down a single connection
165 only the first would provide a SIZE. This was due to the size information
166 not being properly tracked.
168 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
169 TAI (at 37 seconds currently), pretend to be in UTC for time-related
170 expansion and logging. Previously, spurious values such as a future
171 minute could be seen.
173 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
174 it could crash from a null-deref. This could also affect the
175 ${addresses: } operator and ${readsock } item.
177 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
178 message following a mime one, the variable was not reset.
180 JH/30 When an pipelined-connect fails at the first response, assume incorrect
181 cached capability (perhaps the peer reneged?) and immediately retry in
184 JH/31 Fix spurious detection of timeout while writing to transport filter.
186 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
187 an attempt to copy the string was made before checking it.
189 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
190 taint of the lookup key was maintained; we now regard the presence in the
191 filesystem as sufficient validation.
193 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
194 argument is supplied.
196 JH/35 The dsearch lookup now requires that the directory is an absolute path.
197 Previously this was not checked, and nonempty relative paths made an
198 access under Exim's current working directory.
200 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
201 Previously no event was raised.
203 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
204 parameter supplied by the sender MAIL FROM command. Previously it was
205 ignored, and only the check_spool_space option value for the required
208 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
209 the size of the signing public-key. Previously it was instead giving
210 the size of the signature hash.
212 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
213 the default. See the (new) dkim_verify_min_keysizes option.
215 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
216 and an ACL use a lookup for checking either the local_part or domain,
217 stale data could be accessed. Ensure that variable references are
218 dropped between messages.
220 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
221 by the client was not checked as pointing within response data before
222 being used. A malicious client could thus cause an out-of-bounds read and
223 possibly gain authentication. Fix by adding the check.
225 JH/42 Internationalisation: change the default for downconversion in the smtp
226 transport to be "if needed". Previously it was "as previously set" for
227 the message, which usually meant "if needed" for message-submission but
228 "no" for everything else. However, MTAs have been seen using SMTPUTF8
229 even when the envelope addresses did not need it, resulting in forwarding
230 failures to non-supporting MTAs. A downconvert in such cases will be
231 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
232 transport. The change does mean that addresses needing conversion will
233 be converted when previously a delivery failure would occur.
235 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
236 response was received it would be used unchecked in a fail-DSN, violating
237 standards on line-length limits. Truncate if needed.
239 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
240 doesn't work. (Additionally add a new main config option to configure the
247 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
248 side, to decode a TLS 1.3 packet capture.
250 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
251 Previously the default library behaviour applied, sending two, each in
254 JH/03 Debug output for ACL now gives the config file name and line number for
257 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
259 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
261 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
262 buffer overrun for (non-chunking) other transports.
264 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
265 TLS1.3, means that a server rejecting a client certificate is not visible
266 to the client until the first read of encrypted data (typically the
267 response to EHLO). Add detection for that case and treat it as a failed
268 TLS connection attempt, so that the normal retry-in-clear can work (if
269 suitably configured).
271 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
272 and/or domain. Found and fixed by Jason Betts.
274 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
275 configuration). If a CNAME target was not a wellformed name pattern, a
278 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
279 the OS reports them interleaved with other addresses.
281 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
282 used both for input and for a verify callout, both encrypted, SMTP
283 responses being sent by the server could be lost. This resulted in
284 dropped connections and sometimes bounces generated by a peer sending
287 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
288 string. Previously it was possible to cause undefined behaviour in a
289 library routine (usually a crash). Found by "zerons".
291 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
294 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
295 API was removed, so update to use the newer ones.
297 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
298 any timeout set, is taking a long time. Previously we would hang on to a
299 rotated logfile "forever" if the input was arriving with long gaps
300 (a previous attempt to fix addressed lack, for a long time, of initial
303 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
304 shared (NFS) environment. The length of the tempfile name is now
305 4 + 16 ("hdr.$message_exim_id") which might break on file
306 systems which restrict the file name length to lower values.
309 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
310 shared (NFS) environment.
312 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
313 did for all versions <4.90). Notably -M, -m, --invert, -I may be
316 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
317 on some platforms for bit 31.
319 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
320 to changes apparently associated with TLS1.3 handling some of the APIs
321 previously used were either nonfunctional or inappropriate. Strings
322 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
323 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
324 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
325 This affects log line X= elements, the $tls_{in,out}_cipher variables,
326 and the use of specific cipher names in the encrypted= ACL condition.
328 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
330 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
331 verification result was not updated unless hosts_require_ocsp applied.
333 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
334 queue_list_requires_admin set to false, non-admin users were denied the
337 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
338 directory-of-certs mode. Previously they were advertised despite the
341 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
342 A single TCP connection by a client will now hold a TLS connection open
343 for multiple message deliveries, by default. Previously the default was to
346 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
347 default. If built with the facility, DANE will be used. The facility
348 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
350 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
351 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
352 must be defined and you must still, unless you define DISABLE_TLS, manage
353 the the include-dir and library-file requirements that go with that
354 choice. Non-TLS builds are still supported.
356 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
359 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
360 default. If the platform supports and has the facility enabled, it will
361 be requested on all coneections.
363 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
364 controlled by the build-time option SUPPORT_PIPE_CONNECT.
366 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
368 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
369 success-DSN messages. Previously the From: header was always the default
370 one for these; the option was ignored.
372 JH/28 Fix the timeout on smtp response to apply to the whole response.
373 Previously it was reset for every read, so a teergrubing peer sending
374 single bytes within the time limit could extend the connection for a
375 long time. Credit to Qualsys Security Advisory Team for the discovery.
377 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
378 delivery address, which leaked information of the results of local
379 forwarding. Change to the original envelope recipient address, per
382 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
383 requested. Previously not bounce was generated and a log entry of
384 error ignored was made.
386 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
388 JH/32 Introduce a general tainting mechanism for values read from the input
389 channel, and values derived from them. Refuse to expand any tainted
390 values, to catch one form of exploit.
392 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
393 was unused and the unexpanded text used for the test. Found and
394 fixed by Ruben Jenster.
396 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
397 an attempt to use a TLS library read routine dereffed a nul pointer,
400 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
401 them in our smtp response.
403 JH/36 Have the general router option retry_use_local_part default to true when
404 any of the restrictive preconditions are set (to anything). Previously it
405 was only for check_local user. The change removes one item of manual
406 configuration which is required for proper retries when a remote router
407 handles a subset of addresses for a domain.
409 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
410 link count into consideration.
412 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
413 caused the extension of big_buffer, the following lines were ignored.
415 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
416 accordance with RFC 2308. Previously there was no expiry, so a longlived
417 receive process (eg. due to ACL delays) versus a short SOA value could
420 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
422 JH/39 Promote DMARC support to mainline.
424 JH/40 Bug 2452: Add a References: header to DSNs.
426 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
427 parameters. The relevant library call is documented as "Deprecated: This
428 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
429 3.6.0, DH parameters are negotiated following RFC7919."
431 HS/06 Change the default of dnssec_request_domains to "*"
433 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
434 carried on and emitted a BDAT command, even when PIPELINING was not
437 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
438 buffer was used for the filename, resulting in a trap when tainted
439 arguments (eg. $domain) were used.
441 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
442 recommended to avoid a possible server-load attack. The feature can be
443 re-enabled via the openssl_options main cofiguration option.
445 JH/45 local_scan API: documented the current smtp_printf() call. This changed
446 for version 4.90 - adding a "more data" boolean to the arguments.
447 Bumped the ABI version number also, this having been missed previously;
448 release versions 4.90 to 4.92.3 inclusive were effectively broken in
449 respect of usage of smtp_printf() by either local_scan code or libraries
450 accessed via the ${dlfunc } expansion item. Both will need coding
451 adjustment for any calls to smtp_printf() to match the new function
452 signature; a FALSE value for the new argument is always safe.
454 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
455 the file-offset (which the Linux syscall does, and exim expects); this
456 resulted in an indefinite loop.
458 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
459 to do ARC verification. The Authentication-Results: header line added
460 by the configuration then had no ARC item.
466 JH/01 Remove code calling the customisable local_scan function, unless a new
467 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
469 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
470 non-signal-safe functions being used.
472 JH/03 Bug 2269: When presented with a received message having a stupidly large
473 number of DKIM-Signature headers, disable DKIM verification to avoid
474 a resource-consumption attack. The limit is set at twenty.
476 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
477 report of oldest_pass in ${authres } in consequence, and separate out
478 some descriptions of reasons for verification fail.
480 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
481 files in the spool were present and unlocked. A queue-runner could spot
482 them, resulting in a duplicate delivery. Fix that by doing the unlock
483 after the unlink. Investigation by Tim Stewart. Take the opportunity to
484 add more error-checking on spoolfile handling while that code is being
487 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
488 No known attacks, no CVE, this is defensive hardening.
490 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
491 a queue-runner could start a delivery while other operations were ongoing.
492 Cutthrough delivery was a common victim, resulting in duplicate delivery.
493 Found and investigated by Tim Stewart. Fix by using the open message data
494 file handle rather than opening another, and not locally closing it (which
495 releases a lock) for that case, while creating the temporary .eml format
496 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
498 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
499 $sender_verify_failure/$recipient_verify_failure to "random".
501 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
504 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
505 Previously this would segfault.
507 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
510 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
511 like zero, since the resolver should be doing this for us, But we need one
512 as a CNAME but no MX presence gets the CNAME returned; we need to check
513 that doesn't point to an MX to declare it "no MX returned" rather than
514 "error, loop". A new main option is added so the older capability of
515 following some limited number of chain links is maintained.
517 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
519 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
520 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
521 now-deprecated earlier definitions used only specified the range up to TLS
522 1.2 (in the older-version library docs).
524 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
526 JH/15 Rework TLS client-side context management. Stop using a global, and
527 explicitly pass a context around. This enables future use of TLS for
528 connections to service-daemons (eg. malware scanning) while a client smtp
529 connection is using TLS; with cutthrough connections this is quite likely.
531 JH/16 Fix ARC verification to do AS checks in reverse order.
533 JH/17 Support a "tls" option on the ${readsocket } expansion item.
535 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
536 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
537 Previously the "utf8" would be re-prepended for every additional message.
539 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
540 Previously thery were accepted, resulting in issues when attempting to
541 forward messages to a non-supporting MTA.
543 PP/02 Let -n work with printing macros too, not just options.
545 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
546 one parent address was copied, and bogus data was used at delivery-logging
547 time. Either a crash (after delivery) or bogus log data could result.
548 Discovery and analysis by Tim Stewart.
550 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
551 Previously if the string ended mid-character, we did not insert the
552 promised '?' replacement.
554 PP/04 Documentation: current string operators work on bytes, not codepoints.
556 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
557 should pack well giving a smaller memory footprint so better caching and
558 therefore performance. Group the declarations where this can't be done so
559 that the byte-sized flag variables are not interspersed among pointer
560 variables, giving a better chance of good packing by the compiler.
562 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
563 non-null, to avoid issues with sites running BATV. Previously reports were
564 sent with an empty envelope sender so looked like bounces.
566 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
567 The ignore_error flag wasn't being returned from the filter subprocess so
568 was not set for later routers. Investigation and fix by Matthias Kurz.
570 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
571 and a msg:complete for the whole, when a message is manually removed using
572 -Mrm. Developement by Matthias Kurz, hacked on by JH.
574 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
575 a "Gnu special" function, asprintf() in the DB utility binary builds; I
576 hope that is portable enough.
578 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
579 requiring a known-CA anchor certificate; make it now rely entirely on the
580 TLSA as an anchor. Checking the name on the leaf cert against the name
581 on the A-record for the host is still done for TA (but not for EE mode).
583 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
584 included in delivery lines for non-proxied connections, when compiled with
585 SUPPORT_SOCKS and running with proxy logging enabled.
587 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
588 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
589 move the existing event to fire before the normal logging of message
590 failure so that custom logging is bracketed by normal logging.
592 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
593 msg:fail:internal event. Developement by Matthias Kurz.
595 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
596 far too small for todays use of crypto signatures stored there. Go all
597 the way to the max DNS message size of 64kB, even though this might be
598 overmuch for IOT constrained device use.
600 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
601 copy a string over itself. The library routine is documented as not
602 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
604 JH/32 For main options check_spool_space and check_inode_space, where the
605 platform supports 64b integers, support more than the previous 2^31 kB
606 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
607 the previous G, M, k.
609 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
610 $authenticated_fail_id variable on authentication failure. Previously
613 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
614 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
615 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
616 GNUTLS_SEC_PARAM_MEDIUM.
618 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
619 side. Previously we would continue as if no SNI had been received.
621 JH/36 Harden the handling of string-lists. When a list consisted of a sole
622 "<" character, which should be a list-separator specification, we walked
623 off past the nul-terimation.
625 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
626 causes) even when the retry time is not yet met. Previously they were
627 not, meaning that when (say) an account was over-quota and temp-rejecting,
628 and multiple senders' messages were queued, only one sender would get
629 notified on each configured delay_warning cycle.
631 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
633 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
634 cipher-suites, an error can be left on the stack even for a succeeding
635 accept; this results in impossible error messages when a later operation
638 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
639 return error codes indicating retry. Under TLS1.3 this becomes required.
641 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
642 it only wrote the new authenticators, resulting in a lack of tracking of
643 peer changes of ESMTP extensions until the next cache flush.
645 JH/41 Fix the loop reading a message header line to check for integer overflow,
646 and more-often against header_maxsize. Previously a crafted message could
647 induce a crash of the recive process; now the message is cleanly rejected.
649 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
650 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
656 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
657 When redis_servers is set to a list of > 1 element, and the Redis servers
658 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
659 case of MOVED into a DEFER case instead, thus moving the query onto the
660 next server in the list. For a cluster of N elements, all N servers must
661 be defined in redis_servers.
663 GF/02 Catch and remove uninitialized value warning in exiqsumm
664 Check for existence of @ARGV before looking at $ARGV[0]
666 JH/01 Replace the store_release() internal interface with store_newblock(),
667 which internalises the check required to safely use the old one, plus
668 the allocate and data copy operations duplicated in both (!) of the
669 extant use locations.
671 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
672 modifier. This matches the restriction on the commandline.
674 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
675 Previously only the last row was returned.
677 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
678 we assumed that tags in the header were well-formed, and parsed the
679 element content after inspecting only the first char of the tag.
680 Assumptions at that stage could crash the receive process on malformed
683 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
684 While running the DKIM ACL we operate on the Permanent memory pool so that
685 variables created with "set" persist to the DATA ACL. Also (at any time)
686 DNS lookups that fail create cache records using the Permanent pool. But
687 expansions release any allocations made on the current pool - so a dnsdb
688 lookup expansion done in the DKIM ACL releases the memory used for the
689 DNS negative-cache, and bad things result. Solution is to switch to the
690 Main pool for expansions.
691 While we're in that code, add checks on the DNS cache during store_reset,
692 active in the testsuite.
693 Problem spotted, and debugging aided, by Wolfgang Breyha.
695 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
696 When none of the hosts presented to a transport match an already-open
697 connection, close it and proceed with the list. Previously we would
698 queue the message. Spotted by Lena with Yahoo, probably involving
701 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
702 Previously a spurious "250 OK id=" response was appended to the proper
705 JH/08 The "support for" informational output now, which built with Content
706 Scanning support, has a line for the malware scanner interfaces compiled
707 in. Interface can be individually included or not at build time.
709 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
710 by the template makefile "src/EDITME". The "STREAM" support for an older
711 ClamAV interface method is removed.
713 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
714 rows affected is given instead).
716 JH/11 The runtime Berkeley DB library version is now additionally output by
717 "exim -d -bV". Previously only the compile-time version was shown.
719 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
720 SMTP connection. Previously, when one had more receipients than the
721 first, an abortive onward connection was made. Move to full support for
722 multiple onward connections in sequence, handling cutthrough connection
723 for all multi-message initiating connections.
725 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
726 routers. Previously, a multi-recipient message would fail to match the
727 onward-connection opened for the first recipient, and cause its closure.
729 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
730 a timeout on read on a GnuTLS initiating connection, resulting in the
731 initiating connection being dropped. This mattered most when the callout
732 was marked defer_ok. Fix to keep the two timeout-detection methods
735 JH/15 Relax results from ACL control request to enable cutthrough, in
736 unsupported situations, from error to silently (except under debug)
737 ignoring. This covers use with PRDR, frozen messages, queue-only and
740 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
742 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
743 metadata, resulting in a crash in free().
745 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
746 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
747 Broken also in d185889f4, with init system revamp.
749 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
750 Previously we abruptly closed the connection after reading a malware-
751 found indication; now we go on to read the "scan ok" response line,
754 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
755 ACL. Previously, a crash would result.
757 JH/19 Speed up macro lookups during configuration file read, by skipping non-
758 macro text after a replacement (previously it was only once per line) and
759 by skipping builtin macros when searching for an uppercase lead character.
761 JH/20 DANE support moved from Experimental to mainline. The Makefile control
762 for the build is renamed.
764 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
765 was allocated for every new TLS startup, meaning one per message. Fix
766 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
768 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
769 reported the original. Fix to report (as far as possible) the ACL
770 result replacing the original.
772 JH/23 Fix memory leak during multi-message connections using STARTTLS under
773 OpenSSL. Certificate information is loaded for every new TLS startup,
774 and the resources needed to be freed.
776 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
778 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
781 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
782 DATA response info to the (existing) per-recipient response info for
783 the "C=" log element. It can have useful tracking info from the
784 destination system. Patch from Simon Arlott.
786 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
787 length value. Previously this would segfault.
789 HS/02 Support Avast multiline protoocol, this allows passing flags to
790 newer versions of the scanner.
792 JH/28 Ensure that variables possibly set during message acceptance are marked
793 dead before release of memory in the daemon loop. This stops complaints
794 about them when the debug_store option is enabled. Discovered specifically
795 for sender_rate_period, but applies to a whole set of variables.
796 Do the same for the queue-runner and queue-list loops, for variables set
797 from spool message files. Do the same for the SMTP per-message loop, for
798 certain variables indirectly set in ACL operations.
800 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
801 as a multi-recipient message from a mailinglist manager). The coding had
802 an arbitrary cutoff number of characters while checking for more input;
803 enforced by writing a NUL into the buffer. This corrupted long / fast
804 input. The problem was exposed more widely when more pipelineing of SMTP
805 responses was introduced, and one Exim system was feeding another.
806 The symptom is log complaints of SMTP syntax error (NUL chars) on the
807 receiving system, and refused recipients seen by the sending system
808 (propating to people being dropped from mailing lists).
809 Discovered and pinpointed by David Carter.
811 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
812 replaced by the ${authresults } expansion.
814 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
816 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
817 allows proper process termination in container environments.
819 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
820 Previously the "final dot" had a newline after it; ensure it is CR,LF.
822 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
823 and "err_perm", deprecated since 4.83 when the RFC-defined words
824 "temperror" and "permerror" were introduced.
826 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
827 transport-filters or DKIM-signing. The restriction was lost in the
828 consolidation of verify-callout and delivery SMTP handling.
829 Extend the restriction to also cover ARC-signing.
831 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
832 in defer=pass mode supply a 450 to the initiator. Previously the message
835 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
836 tls_require_ciphers is used as before.
838 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
839 "pass_unscanned". Only tmpfails from the scanner are written to
840 the paniclog, as they may require admin intervention (permission
841 denied, license issues). Other scanner errors (like decompression
842 bombs) do not cause a paniclog entry.
844 JH/36 Fix reinitialisation of DKIM logging variable between messages.
845 Previously it was possible to log spurious information in receive log
848 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
849 triggered odd behaviour from Outlook Express clients.
851 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
854 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
855 since the IETF WG has not yet settled on that versus the original
856 "bare" representation.
858 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
859 Previously the millisecond value corrupted the output.
860 Fix also for syslog_pid=no and log_selector +pid, for which the pid
861 corrupted the output.
867 JH/01 Rework error string handling in TLS interface so that the caller in
868 more cases is responsible for logging. This permits library-sourced
869 string to be attached to addresses during delivery, and collapses
870 pairs of long lines into single ones.
872 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
873 during configuration. Wildcards are allowed and expanded.
875 JH/02 Rework error string handling in DKIM to pass more info back to callers.
876 This permits better logging.
878 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
879 do not close it down and have the child transport start it up again on
880 the passed-on TCP connection. Instead, proxy the child (and any
881 subsequent ones) for TLS via a unix-domain socket channel. Logging is
882 affected: the continued delivery log lines do not have any DNSSEC, TLS
883 Certificate or OCSP information. TLS cipher information is still logged.
885 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
886 identical IP addresses on different listening ports. Will also affect
889 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
890 add noisy ifdef guards to special-case this sillyness.
891 Patch from Bernd Kuhls.
893 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
894 than 255 are no longer allowed.
896 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
897 Disable the session-cache too, which might reduce our load. Since we
898 currrectly use a new context for every connection, both as server and
899 client, there is no benefit for these.
900 GnuTLS appears to not support tickets server-side by default (we don't
901 call gnutls_session_ticket_enable_server()) but client side is enabled
902 by default on recent versions (3.1.3 +) unless the PFS priority string
905 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
906 <https://reproducible-builds.org/specs/source-date-epoch/>.
908 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
909 the check for any unsuccessful recipients did not notice the limit, and
910 erroneously found still-pending ones.
912 JH/08 Pipeline CHUNKING command and data together, on kernels that support
913 MSG_MORE. Only in-clear (not on TLS connections).
915 JH/09 Avoid using a temporary file during transport using dkim. Unless a
916 transport-filter is involved we can buffer the headers in memory for
917 creating the signature, and read the spool data file once for the
918 signature and again for transmission.
920 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
921 4.77 as the kernel support then wasn't solid, having issues in 64bit
922 mode. Now, it's been long enough. Add support for FreeBSD also.
924 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
925 case where the routing stage had gathered several addresses to send to
926 a host before calling the transport for the first, we previously failed
927 to close down TLS in the old transport process before passing the TCP
928 connection to the new process. The new one sent a STARTTLS command
929 which naturally failed, giving a failed delivery and bloating the retry
930 database. Investigation and fix prototype from Wolfgang Breyha.
932 JH/12 Fix check on SMTP command input synchronisation. Previously there were
933 false-negatives in the check that the sender had not preempted a response
934 or prompt from Exim (running as a server), due to that code's lack of
935 awareness of the SMTP input buffering.
937 PP/04 Add commandline_checks_require_admin option.
938 Exim drops privileges sanely, various checks such as -be aren't a
939 security problem, as long as you trust local users with access to their
940 own account. When invoked by services which pass untrusted data to
941 Exim, this might be an issue. Set this option in main configuration
942 AND make fixes to the calling application, such as using `--` to stop
945 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
946 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
947 responses to those, into a single TLS record each way (this usually means
948 a single packet). As a side issue, smtp_enforce_sync now works on TLS
951 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
952 affects you only if you're dancing at the edge of the param size limits.
953 If you are, and this message makes sense to you, then: raise the
954 configured limit or use OpenSSL 1.1. Nothing we can do for older
957 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
958 cmdline element to get the documented default one. Previously it was
961 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
962 are unsuable for non-message-specific reasons - eg. connection timeout,
963 banner-time rejection.
965 JH/16 Fix logging of delivery remote port, when specified by router, under
968 PP/06 Repair manualroute's ability to take options in any order, even if one
969 is the name of a transport.
972 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
974 JH/17 Change the list-building routines interface to use the expanding-string
975 triplet model, for better allocation and copying behaviour.
977 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
978 Previously it was constructed the first time a possibly-matching string
979 was met in the configuration file input during startup; now it is done
982 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
983 compatible one, to avoid the (poorly documented) possibility of a config
984 file in the working directory redirecting the DB files, possibly correpting
985 some existing file. CVE-2017-10140 assigned for BDB.
987 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
988 cache-hot. Previously, although the result was properly cached, the
989 initial verify call returned a defer.
991 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
992 the main verify for receipient in uncached-mode.
994 JH/22 Retire historical build files to an "unsupported" subdir. These are
995 defined as "ones for which we have no current evidence of testing".
997 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
998 if present. Previously it was ignored.
1000 JH/24 Start using specified-initialisers in C structure init coding. This is
1001 a C99 feature (it's 2017, so now considered safe).
1003 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1004 if was a fixed-sized field and bitmask ops via macros; it is now more
1007 PP/07 GitHub PR 56: Apply MariaDB build fix.
1008 Patch provided by Jaroslav Škarvada.
1010 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1011 during Coverity cleanups [4.87 JH/47]
1012 Diagnosis and fix provided by Michael Fischer v. Mollard.
1014 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1015 the right size to place the terminating semicolon on its own folded
1016 line, the header hash was calculated to an incorrect value thanks to
1017 the (relaxed) space the fold became.
1019 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1020 and confused the parent.
1022 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1023 which could crash as a result. This could lead to undeliverable messages.
1025 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1028 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1029 matters on fast-turnover and PID-randomising systems, which were getting
1030 out-of-order delivery.
1032 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1033 a possibly-overlapping copy. The symptom was that "Remote host closed
1034 connection in response to HELO" was logged instead of the actual 4xx
1037 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1038 Previously only that bufferd was discarded, resulting in SYMTP command
1041 JH/32 DKIM: when a message has multiple signatures matching an identity given
1042 in dkim_verify_signers, run the dkim acl once for each. Previously only
1043 one run was done. Bug 2189.
1045 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1046 "panic the current process" to "deliberately defer". The panic log is
1047 still written with the problem list name; the mail and reject logs now
1048 get a temp-reject line for the message that was being handled, saying
1049 something like "domains check lookup or other defer". The SMTP 451
1050 message is still "Temporary local problem".
1052 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1053 A crafted sequence of BDAT commands could result in in-use memory beeing
1054 freed. CVE-2017-16943.
1056 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1057 from SMTP input. Previously it was always done; now only done for DATA
1058 and not BDAT commands. CVE-2017-16944.
1060 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1061 to the message (such as an overlong header line). Previously this was
1062 not done and we did not exit BDAT mode. Followon from the previous item
1063 though a different problem.
1069 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1070 than -2003 did; needs libidn2 in addition to libidn.
1072 JH/02 The path option on a pipe transport is now expanded before use.
1074 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1075 Patch provided by "Björn", documentation fix added too.
1077 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1078 missing a wire-to-host endian conversion.
1080 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1081 close after a BDAT command line could be taken as a following command,
1082 giving a synch failure. Fix by only checking for synch immediately
1083 before acknowledging the chunk.
1085 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1086 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1087 macro. Patches provided by Josh Soref.
1089 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1090 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1091 by IANA as a service extension. However, John Klensin suggests that we
1094 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1095 the dkim code may be unix-mode line endings rather than smtp wire-format
1096 CRLF, so prepend a CR to any bare LF.
1098 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1099 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1101 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1102 the first were themselves being wrongly included in the feed into dkim
1103 processing; with most chunk sizes in use this resulted in an incorrect
1104 body hash calculated value.
1106 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1107 DKIM signature block, for verification. Although advised against by
1108 standards it is specifically not ruled illegal.
1110 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1112 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1113 missing a body hash (the bh= tag).
1115 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1116 It seems that HAProxy sends the Proxy Protocol information in clear and
1117 only then does a TLS startup, so do the same.
1119 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1120 TCP connections (such as for Spamd) unless the daemon successfully set
1121 Fast Open mode on its listening sockets. This fixes breakage seen on
1122 too-old kernels or those not configured for Fast Open, at the cost of
1123 requiring both directions being enabled for TFO, and TFO never being used
1124 by non-daemon-related Exim processes.
1126 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1127 endings, at least on the first header line. Try to canonify any that get
1128 past that check, despite the cost.
1130 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1131 now limited to an arbitrary five deep, while parsing addresses with the
1132 strip_excess_angle_brackets option enabled.
1134 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1135 instead leave the unprompted TLS handshake in socket buffer for the
1136 TLS library to consume.
1138 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1140 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1142 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1143 region in one operation, for speed, and this leaves assigned pointers
1144 dangling. Add checks run only under the testsuite which checks all
1145 variables at a store-reset and panics on a dangling pointer; add code
1146 explicitly nulling out all the variables discovered. Fixes one known
1147 bug: a transport crash, where a dangling pointer for $sending_ip_address
1148 originally assigned in a verify callout, is re-used.
1150 PP/06 Drop '.' from @INC in various Perl scripts.
1152 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1154 PP/08 Reduce a number of compilation warnings under clang; building with
1155 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1156 should be warning-free.
1158 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1160 HS/01 Fix portability problems introduced by PP/08 for platforms where
1161 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1163 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1164 chunk. This allows us to accept broken chunked messages. We need a more
1165 general solution here.
1167 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1168 already-broken messages in the queue.
1170 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1172 JH/19 Fix reference counting bug in routing-generated-address tracking.
1178 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1179 supports it and a size is available (ie. the sending peer gave us one).
1181 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1182 years of being deprecated). The replacements are the ACLs
1183 acl_smtp_mime and acl_not_smtp_mime.
1185 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1186 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1187 or even an in-clear connection were permitted. Now, if the host lookup
1188 was dnssec and dane was requested then the host is only used if the
1189 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1190 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1191 if one fails this test.
1192 This means that a poorly-configured remote DNS will make it incommunicado;
1193 but it protects against a DNS-interception attack on it.
1195 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1196 non-noisy when a race steals the message being considered.
1198 JH/05 If main configuration option tls_certificate is unset, generate a
1199 self-signed certificate for inbound TLS connections.
1201 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1202 in rewrites and routers.
1204 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1205 and logged a warning sing 4.83; now they are a configuration file error.
1207 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1208 (lacking @domain). Apply the same qualification processing as RCPT.
1210 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1212 JH/10 Support ${sha256:} applied to a string (as well as the previous
1215 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1216 a cutthrough deliver is pending, as we always want to make a connection.
1217 This also avoids re-routing the message when later placing the cutthrough
1218 connection after a verify cache hit.
1219 Do not update it with the verify result either.
1221 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1222 when routing results in more than one destination address.
1224 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1225 signing (which inhibits the cutthrough capability). Previously only
1226 the presence of an option was tested; now an expansion evaluating as
1227 empty is permissible (obviously it should depend only on data available
1228 when the cutthrough connection is made).
1230 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1231 the relevant preceding SMTP command did not note the pipelining mode.
1233 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1234 Previously they were not counted.
1236 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1237 as one having no matching records. Previously we deferred the message
1238 that needed the lookup.
1240 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1241 distinguished as "(=".
1243 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1244 for missing MX records. Previously it only worked for missing A records.
1246 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1248 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1249 after the data-go-ahead and data-ack. Patch from Jason Betts.
1251 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1252 even for a "none" policy. Patch from Tony Meyer.
1254 JH/22 Fix continued use of a connection for further deliveries. If a port was
1255 specified by a router, it must also match for the delivery to be
1258 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1259 When one of the recipients of a message was unsuitable for the connection
1260 (has no matching addresses), we lost track of needing to mark it
1261 deferred. As a result mail would be lost.
1263 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1265 JH/25 Decoding ACL controls is now done using a binary search; the source code
1266 takes up less space and should be simpler to maintain. Merge the ACL
1267 condition decode tables also, with similar effect.
1269 JH/26 Fix problem with one_time used on a redirect router which returned the
1270 parent address unchanged. A retry would see the parent address marked as
1271 delivered, so not attempt the (identical) child. As a result mail would
1274 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1275 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1276 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1279 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1280 Main config options check_{log,spool}_{inodes,space} are now
1281 100 inodes, 10MB unless set otherwise in the configuration.
1283 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1284 Previously it only applied to the main-section connection policy
1287 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1289 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1290 by me. Added RFC7919 DH primes as an alternative.
1292 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1293 are not in the system include path.
1295 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1296 GnuTLS, when a session startup failed (eg because the client disconnected)
1297 Exim did stdio operations after fclose. This was exposed by a recent
1298 change which nulled out the file handle after the fclose.
1300 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1301 signed directly by the cert-signing cert, rather than an intermediate
1302 OCSP-signing cert. This is the model used by LetsEncrypt.
1304 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1306 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1307 an incoming connection.
1309 HS/02 Bug 1802: Do not half-close the connection after sending a request
1312 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1313 fallback to "prime256v1".
1315 JH/34 SECURITY: Use proper copy of DATA command in error message.
1316 Could leak key material. Remotely exploitable. CVE-2016-9963.
1322 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1323 and 3.4.4 - once the server is enabled to respond to an OCSP request
1324 it does even when not requested, resulting in a stapling non-aware
1325 client dropping the TLS connection.
1327 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1328 support variable-length bit vectors. No functional change.
1330 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1331 The I= interface field on outgoing lines is now after the H= remote
1332 host field, same as incoming lines. There is a separate
1333 outgoing_interface log selector which allows you to disable the
1336 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1337 If not running log_selector +smtp_connection the mainlog would be held
1338 open indefinitely after a "too many connections" event, including to a
1339 deleted file after a log rotate. Leave the per net connection logging
1340 leaving it open for efficiency as that will be quickly detected by the
1341 check on the next write.
1343 HS/01 Bug 1671: Fix post transport crash.
1344 Processing the wait-<transport> messages could crash the delivery
1345 process if the message IDs didn't exist for some reason. When
1346 using 'split_spool_directory=yes' the construction of the spool
1347 file name failed already, exposing the same netto behaviour.
1349 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1350 mime_regex ACL conditions.
1352 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1353 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1354 to HELO, local diagnostic string.
1356 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1357 log line to debug. Even when configured with a tls authenticator many
1358 client connections are expected to not authenticate in this way, so
1359 an authenticate fail is not an error.
1361 HS/02 Add the Exim version string to the process info. This way exiwhat
1362 gives some more detail about the running daemon.
1364 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1365 matter for fast-change records such as DNSBLs.
1367 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1368 retry record, even if constant. There may be multiple transports with
1369 different interface settings and the retry behaviour needs to be kept
1372 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1374 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1376 JH/10 Bug 840: fix log_defer_output option of pipe transport
1378 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1379 hosts_randomize. This matters a lot when combined with mua_wrapper.
1381 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1382 ${quote_pgsql:<string>} operator.
1384 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1386 JH/14 Change header file name for freeradius-client. Relevant if compiling
1387 with Radius support; from the Gentoo tree and checked under Fedora.
1389 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1391 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1392 option transport_filter, meaning no filtering.
1394 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1396 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1397 defaults to "*" (all hosts). The variable is now available when not built
1398 with TLS, default unset, mainly to enable keeping the testsuite sane.
1399 If a server certificate is not supplied (via tls_certificate) an error is
1400 logged, and clients will find TLS connections fail on startup. Presumably
1401 they will retry in-clear.
1402 Packagers of Exim are strongly encouraged to create a server certificate
1403 at installation time.
1405 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1406 with the $config_file variable.
1408 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1409 in transport context, after the attempt, and per-recipient. The latter type
1410 is per host attempted. The event data is the error message, and the errno
1411 information encodes the lookup type (A vs. MX) used for the (first) host,
1412 and the trailing two digits of the smtp 4xx response.
1414 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1415 to write to mainlog (or rejectlog, paniclog) in the window between file
1416 creation and permissions/ownership being changed. Particularly affects
1417 installations where exicyclog is run as root, rather than exim user;
1418 result is that the running daemon panics and dies.
1420 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1422 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1423 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1424 "pri" and "weight". Note that the previous implicit priority given by the
1425 list order is no longer honoured.
1427 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1428 for DKIM processing.
1430 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1431 by defining SUPPORT_SOCKS.
1433 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1434 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1435 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1436 variables are renamed to proxy_{local,external}_{address,port}.
1438 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1439 for a build by defining SUPPORT_I18N
1441 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1442 of the query string, and make ${quote_redis:} do that quoting.
1444 JH/29 Move Events support from Experimental to mainline, enabled by default
1445 and removable for a build by defining DISABLE_EVENT.
1447 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1449 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1450 cached by the daemon.
1452 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1453 by defining LOOKUP_REDIS. The libhiredis library is required.
1455 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1456 keys are given for lookup.
1458 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1459 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1460 only supported when built with TLS support. The PolarSSL SHA routines
1461 are still used when the TLS library is too old for convenient support.
1463 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1464 openssl_options), for security. OpenSSL forces this from version 1.1.0
1465 server-side so match that on older versions.
1467 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1468 allocation for $value could be released as the expansion processing
1469 concluded, but leaving the global pointer active for it.
1471 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1472 and to use the domains and local_parts ACL conditions.
1474 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1475 incorrectly not doubled on cutthrough transmission, hence seen as a
1476 body-termination at the receiving system - resulting in truncated mails.
1477 Commonly the sender saw a TCP-level error, and retransmitted the message
1478 via the normal store-and-forward channel. This could result in duplicates
1479 received - but deduplicating mailstores were liable to retain only the
1480 initial truncated version.
1482 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1484 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1486 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1487 we're in there, support oversigning also; bug 1309.
1489 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1491 HS/04 Add support for keep_environment and add_environment options.
1493 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1494 either intentional arithmetic overflow during PRNG, or testing config-
1497 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1498 delivery resulted in actual delivery. Cancel cutthrough before DATA
1501 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1502 rejects a recipient: pass the reject to the originator.
1504 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1505 Many were false-positives and ignorable, but it's worth fixing the
1508 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1509 for the new environment-manipulation done at startup. Move the routines
1510 from being local to tls.c to being global via the os.c file.
1512 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1513 an extract embedded as result-arg for a map, the first arg for extract
1514 is unavailable so we cannot tell if this is a numbered or keyed
1515 extraction. Accept either.
1521 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1524 JH/02 The smtp transport option "multi_domain" is now expanded.
1526 JH/03 The smtp transport now requests PRDR by default, if the server offers
1529 JH/04 Certificate name checking on server certificates, when exim is a client,
1530 is now done by default. The transport option tls_verify_cert_hostnames
1531 can be used to disable this per-host. The build option
1532 EXPERIMENTAL_CERTNAMES is withdrawn.
1534 JH/05 The value of the tls_verify_certificates smtp transport and main options
1535 default to the word "system" to access the system default CA bundle.
1536 For GnuTLS, only version 3.0.20 or later.
1538 JH/06 Verification of the server certificate for a TLS connection is now tried
1539 (but not required) by default. The verification status is now logged by
1540 default, for both outbound TLS and client-certificate supplying inbound
1543 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1546 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1547 Status Notification (bounce) messages are now MIME format per RFC 3464.
1548 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1549 under the control of the dsn_advertise_hosts option, and routers may
1550 have a dsn_lasthop option.
1552 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1553 default, modifiable by a malware= option. The list separator for
1554 the options can now be changed in the usual way. Bug 68.
1556 JH/10 The smtp_receive_timeout main option is now expanded before use.
1558 JH/11 The incoming_interface log option now also enables logging of the
1559 local interface on delivery outgoing connections.
1561 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1562 if the interface and destination host and port all match.
1564 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1567 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1568 Patch from Andrew Lewis.
1570 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1571 now supports optional time-restrictions, weighting, and priority
1572 modifiers per server. Patch originally by <rommer@active.by>.
1574 JH/16 The spamd_address main option now supports a mixed list of local
1575 and remote servers. Remote servers can be IPv6 addresses, and
1576 specify a port-range.
1578 JH/17 Bug 68: The spamd_address main option now supports an optional
1579 timeout value per server.
1581 JH/18 Bug 1581: Router and transport options headers_add/remove can
1582 now have the list separator specified.
1584 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1587 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1590 JH/21 Support for the A6 type of dns record is withdrawn.
1592 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1593 rather than the verbs used.
1595 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1596 from 255 to 1024 chars.
1598 JH/24 Verification callouts now attempt to use TLS by default.
1600 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1601 are generic router options now. The defaults didn't change.
1603 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1604 Original patch from Alexander Shikoff, worked over by JH.
1606 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1607 files if ZCAT_COMMAND is not executable.
1609 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1611 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1613 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1614 Normally benign, it bites when the pair was led to by a CNAME;
1615 modern usage is to not canonicalize the domain to a CNAME target
1616 (and we were inconsistent anyway for A-only vs AAAA+A).
1618 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1620 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1621 when evaluating $sender_host_dnssec.
1623 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1624 $sender_helo_dnssec variable.
1626 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1628 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1630 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1632 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1633 documented as working, but never had. Support all but $spam_report.
1635 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1636 added for tls authenticator.
1638 HS/03 Add perl_taintmode main config option
1644 TL/01 When running the test suite, the README says that variables such as
1645 no_msglog_check are global and can be placed anywhere in a specific
1646 test's script, however it was observed that placement needed to be near
1647 the beginning for it to behave that way. Changed the runtest perl
1648 script to read through the entire script once to detect and set these
1649 variables, reset to the beginning of the script, and then run through
1650 the script parsing/test process like normal.
1652 TL/02 The BSD's have an arc4random API. One of the functions to induce
1653 adding randomness was arc4random_stir(), but it has been removed in
1654 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1655 function when detected.
1657 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1658 cause callback expansion.
1660 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1661 syntax errors in an expansion can be treated as a string instead of
1662 logging or causing an error, due to the internal use of bool_lax
1663 instead of bool when processing it.
1665 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1666 server certificates when making smtp deliveries.
1668 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1670 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1672 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1674 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1675 Merged patch from Sebastian Wiedenroth.
1677 JH/05 Fix results-pipe from transport process. Several recipients, combined
1678 with certificate use, exposed issues where response data items split
1679 over buffer boundaries were not parsed properly. This eventually
1680 resulted in duplicates being sent. This issue only became common enough
1681 to notice due to the introduction of connection certificate information,
1682 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1684 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1685 size buffer was used, resulting in syntax errors when an expansion
1688 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1689 version 3.3.6 or later.
1691 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1692 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1693 both become "event_action", the variables become $event_name, $event_data
1694 and $event_defer_errno. There is a new variable $verify_mode, usable in
1695 routers, transports and related events. The tls:cert event is now also
1696 raised for inbound connections, if the main configuration event_action
1699 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1700 early OCSP support, but no stapling (appears to be less than 1.0.0).
1702 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1703 server certificate names available under the smtp transport option
1704 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1707 JH/10 Time-related extraction expansions from certificates now use the main
1708 option "timezone" setting for output formatting, and are consistent
1709 between OpenSSL and GnuTLS compilations. Bug 1541.
1711 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1712 encoded parameter in the incoming message. Bug 1558.
1714 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1715 include certificate info, eximon was claiming there were spoolfile
1718 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1720 JH/14 Log delivery-related information more consistently, using the sequence
1721 "H=<name> [<ip>]" wherever possible.
1723 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1724 are problematic for Debian distribution, omit them from the release
1727 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1729 JH/16 Fix string representation of time values on 64bit time_t architectures.
1732 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1733 output list separator was used.
1738 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1739 checkers that were complaining about end of non-void function with no
1742 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1743 This was a regression introduced in 4.83 by another bugfix.
1745 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1747 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1748 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1754 TF/01 Correctly close the server side of TLS when forking for delivery.
1756 When a message was received over SMTP with TLS, Exim failed to clear up
1757 the incoming connection properly after forking off the child process to
1758 deliver the message. In some situations the subsequent outgoing
1759 delivery connection happened to have the same fd number as the incoming
1760 connection previously had. Exim would try to use TLS and fail, logging
1761 a "Bad file descriptor" error.
1763 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1764 utilities have not been installed.
1766 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1767 temporary space as the ACL may create new global variables.
1769 TL/01 LDAP support uses per connection or global context settings, depending
1770 upon the detected version of the libraries at build time.
1772 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1773 to extract and use the src ip:port in logging and expansions as if it
1774 were a direct connection from the outside internet. PPv2 support was
1775 updated based on HAProxy spec change in May 2014.
1777 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1779 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1780 Properly escape header and check for NULL return.
1782 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1785 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1787 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1788 characters in header names, implemented as a verify condition.
1789 Contributed by Michael Fischer v. Mollard.
1791 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1792 results permerror and temperror. Previous values are deprecated but
1793 still accepted. In a future release, err_perm and err_temp will be
1794 completely removed, which will be a backward incompatibility if the
1795 ACL tests for either of these two old results. Patch contributed by
1796 user bes-internal on the mailing list.
1798 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1800 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1801 selectors, in both main and reject logs.
1803 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1806 JH/07 Add malware type "sock" for talking to simple daemon.
1808 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1810 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1811 routers/transports under cutthrough routing.
1813 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1814 numbers. Touch up "bool" conditional to keep the same definition.
1816 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1818 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1820 JH/12 Expand items in router/transport headers_add or headers_remove lists
1821 individually rather than the list as a whole. Bug 1452.
1823 Required for reasonable handling of multiple headers_ options when
1824 they may be empty; requires that headers_remove items with embedded
1825 colons must have them doubled (or the list-separator changed).
1827 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1828 view the policy declared in the DMARC record. Currently, $dmarc_status
1829 is a combined value of both the record presence and the result of the
1832 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1834 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1835 dnslookup router and the smtp transport (applying to the forward
1838 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1839 of ldap servers used for a specific lookup. Patch provided by Heiko
1842 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1843 New variable $lookup_dnssec_authenticated for observability.
1845 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1846 Patch submitted by Lars Timman.
1848 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1850 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1851 Requires trusted mode and valid format message id, aborts otherwise.
1852 Patch contributed by Heiko Schlichting.
1854 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1855 certextract with support for various fields. Bug 1358.
1857 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1858 is requested by default, modifiable by smtp transport option
1861 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1862 operate on certificate variables to give certificate fingerprints
1863 Also new ${sha256:cert_variable}.
1865 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1867 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1868 Christian Aistleitner.
1870 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1872 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1873 file. Patch from Wolfgang Breyha.
1875 JH/25 Expand the coverage of the delivery $host and $host_address to
1876 client authenticators run in verify callout. Bug 1476.
1878 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1879 align with daemon_smtp_ports. Bug 72.
1881 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1882 support and error reporting did not work properly.
1884 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1885 and is readable. Patch from Andrew Colin Kissa.
1887 TL/14 Enhance documentation of ${run expansion and how it parses the
1888 commandline after expansion, particularly in the case when an
1889 unquoted variable expansion results in an empty value.
1891 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1893 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1894 by renaming away from it. Addresses GH issue 15, reported by
1897 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1899 TL/15 SECURITY: prevent double expansion in math comparison functions
1900 (can expand unsanitized data). Not remotely exploitable.
1907 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1909 PP/02 Make -n do something, by making it not do something.
1910 When combined with -bP, the name of an option is not output.
1912 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1915 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1916 $sender_host_name and config options to manage this, and basic check
1919 PP/05 DSCP support for outbound connections and control modifier for inbound.
1921 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1922 (Only plugin which currently uses this is kerberos4, which nobody should
1923 be using, but we should make it available and other future plugins might
1924 conceivably use it, even though it would break NAT; stuff *should* be
1925 using channel bindings instead).
1927 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1928 name; added for Sendmail compatibility; requires admin caller.
1929 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1930 just ignore it); requires trusted caller.
1931 Also parse but ignore: -Ac -Am -X<logfile>
1934 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1936 TL/02 Add +smtp_confirmation as a default logging option.
1938 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1939 Patch by Magnus Holmgren from 2007-02-20.
1941 TL/04 Bugzilla 1281 - Spec typo.
1942 Bugzilla 1283 - Spec typo.
1943 Bugzilla 1290 - Spec grammar fixes.
1945 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1947 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1949 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1950 mailing list by by Dmitry Isaikin.
1952 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1954 JH/02 Support "G" suffix to numbers in ${if comparisons.
1956 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1958 NM/01 Bugzilla 1197 - Spec typo
1959 Bugzilla 1196 - Spec examples corrections
1961 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1963 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1964 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1967 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1968 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1970 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1971 "acl {{name}{arg}...}", and optional args on acl condition
1974 JH/05 Permit multiple router/transport headers_add/remove lines.
1976 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1978 JH/07 Avoid using a waiting database for a single-message-only transport.
1979 Performance patch from Paul Fisher. Bugzilla 1262.
1981 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1984 JH/09 Add $headers_added variable, with content from use of ACL modifier
1985 add_header (but not yet added to the message). Bugzilla 199.
1987 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1988 Pulled from Bugzilla 817 by Wolfgang Breyha.
1990 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1992 (nb: this is the same fix as in Exim 4.80.1)
1994 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1997 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2000 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2002 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2003 advertise SMTP AUTH mechanism to us, instead of a generic
2004 protocol violation error. Also, make Exim more robust to bad
2005 data from the Dovecot auth socket.
2007 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2009 When a queue runner is handling a message, Exim first routes the
2010 recipient addresses, during which it prunes them based on the retry
2011 hints database. After that it attempts to deliver the message to
2012 any remaining recipients. It then updates the hints database using
2015 So if a recipient address works intermittently, it can get repeatedly
2016 deferred at routing time. The retry hints record remains fresh so the
2017 address never reaches the final cutoff time.
2019 This is a fairly common occurrence when a user is bumping up against
2020 their storage quota. Exim had some logic in its local delivery code
2021 to deal with this. However it did not apply to per-recipient defers
2022 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2024 This change adds a proper retry rule check during routing so that the
2025 final cutoff time is checked against the message's age. We only do
2026 this check if there is an address retry record and there is not a
2027 domain retry record; this implies that previous attempts to handle
2028 the address had the retry_use_local_parts option turned on. We use
2029 this as an approximation for the destination being like a local
2030 delivery, as in LMTP.
2032 I suspect this new check makes the old local delivery cutoff check
2033 redundant, but I have not verified this so I left the code in place.
2035 TF/02 Correct gecos expansion when From: is a prefix of the username.
2037 Test 0254 submits a message to Exim with the header
2041 When I ran the test suite under the user fanf2, Exim expanded
2042 the header to contain my full name, whereas it should have added
2043 a Resent-Sender: header. It erroneously treats any prefix of the
2044 username as equal to the username.
2046 This change corrects that bug.
2048 GF/01 DCC debug and logging tidyup
2049 Error conditions log to paniclog rather than rejectlog.
2050 Debug lines prefixed by "DCC: " to remove any ambiguity.
2052 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2054 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2055 Bug spotted by Jeremy Harris; was flawed since initial commit.
2056 Would have resulted in OCSP responses post-SNI triggering an Exim
2057 NULL dereference and crash.
2059 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2061 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2062 Bug detection, analysis and fix by Samuel Thibault.
2063 Bugzilla 1331, Debian bug #698092.
2065 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2067 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2068 Server implementation by Todd Lyons, client by JH.
2069 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2070 config variable "prdr_enable" controls whether the server
2071 advertises the facility. If the client requests PRDR a new
2072 acl_data_smtp_prdr ACL is called once for each recipient, after
2073 the body content is received and before the acl_smtp_data ACL.
2074 The client is controlled by both of: a hosts_try_prdr option
2075 on the smtp transport, and the server advertisement.
2076 Default client logging of deliveries and rejections involving
2077 PRDR are flagged with the string "PRDR".
2079 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2080 fclose(). Diagnosis by Todd Lyons.
2082 PP/17 Update configure.default to handle IPv6 localhost better.
2083 Patch by Alain Williams (plus minor tweaks).
2086 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2087 This is now consistent with GnuTLS, and is now documented: the
2088 previous undocumented portable approach to treating the option as
2089 unset was to force an expansion failure. That still works, and
2090 an empty string is now equivalent.
2092 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2093 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2094 not performing validation itself.
2096 PP/20 Added force_command boolean option to pipe transport.
2097 Patch from Nick Koston, of cPanel Inc.
2099 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2102 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2104 PP/21 Fix eximon continuous updating with timestamped log-files.
2105 Broken in a format-string cleanup in 4.80, missed when I repaired the
2106 other false fix of the same issue.
2107 Report and fix from Heiko Schlichting.
2110 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2111 Report from Prashanth Katuri.
2113 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2114 It's SecureTransport, so affects any MacOS clients which use the
2115 system-integrated TLS libraries, including email clients.
2117 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2118 using a MIME ACL for non-SMTP local injection.
2119 Report and assistance in diagnosis by Warren Baker.
2121 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2123 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2125 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2126 last id that failed so it may be referenced in subsequent ACL's.
2128 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2131 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2132 ldap library initialization, allowing self-signed CA's to be
2133 used. Also properly sets require_cert option later in code by
2134 using NULL (global ldap config) instead of ldap handle (per
2135 session). Bug diagnosis and testing by alxgomz.
2137 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2138 the src/util/ subdirectory.
2140 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2141 renamed to Transport Post Delivery Action by Jeremy Harris, as
2144 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2145 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2146 redis_servers = needs to be configured which will be used by the redis
2147 lookup. Patch from Warren Baker, of The Packet Hub.
2149 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2151 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2152 hostname or reverse DNS when processing a host list. Used suggestions
2153 from multiple comments on this bug.
2155 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2157 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2158 Missed a few lines, added it to make the runtest require no keyboard
2161 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2162 contains upper case chars. Make router use caseful_local_part.
2164 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2165 support when GnuTLS has been built with p11-kit.
2171 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2173 This, or similar/improved, will also be change PP/11 of 4.82.
2179 PP/01 Handle short writes when writing local log-files.
2180 In practice, only affects FreeBSD (8 onwards).
2181 Bugzilla 1053, with thanks to Dmitry Isaikin.
2183 NM/01 Bugzilla 949 - Documentation tweak
2185 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2188 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2190 PP/02 Implemented gsasl authenticator.
2192 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2194 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2195 `pkg-config foo` for cflags/libs.
2197 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2198 with rest of GSASL and with heimdal_gssapi.
2200 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2201 `pkg-config foo` for cflags/libs for the TLS implementation.
2203 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2204 properties get this fed in as external SSF. A number of robustness
2205 and debugging improvements to the cyrus_sasl authenticator.
2207 PP/08 cyrus_sasl server now expands the server_realm option.
2209 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2210 Patch by Jeremy Harris.
2212 PP/10 Added dbmjz lookup type.
2214 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2216 PP/12 MAIL args handles TAB as well as SP, for better interop with
2217 non-compliant senders.
2218 Analysis and variant patch by Todd Lyons.
2220 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2221 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2222 Patch from Dirk Mueller <dmueller@suse.com>
2224 PP/13 tls_peerdn now print-escaped for spool files.
2225 Observed some $tls_peerdn in wild which contained \n, which resulted
2226 in spool file corruption.
2228 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2229 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2230 or write after TLS renegotiation, which otherwise led to messages
2233 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2234 as a tracking header (ie: a signed header comes before the signature).
2235 Patch from Wolfgang Breyha.
2237 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2238 comma-sep list; embedded commas doubled.
2240 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2242 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2244 Report and patch from Dmitry Banschikov.
2246 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2247 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2248 failures. We appear to now support TLS1.1+ with Exim.
2250 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2251 lets Exim select keys and certificates based upon TLS SNI from client.
2252 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2253 before an outbound SMTP session. New log_selector, +tls_sni.
2255 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2256 NULL dereference. Report and patch from Alun Jones.
2258 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2259 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2260 Not seeing resolver debug output on NetBSD, but suspect this is a
2261 resolver implementation change.
2263 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2264 Left warnings. Added "eximon gdb" invocation mode.
2266 PP/21 Defaulting "accept_8bitmime" to true, not false.
2268 PP/22 Added -bw for inetd wait mode support.
2270 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2271 locate the relevant includes and libraries. Made this the default.
2273 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2274 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2276 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2277 This may cause build issues on older platforms.
2279 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2280 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2281 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2282 Added SNI support via GnuTLS too.
2283 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2285 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2287 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2288 Applied second patch from Janne, implementing suggestion to default
2289 multiple-strings-in-record handling to match SPF spec.
2291 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2293 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2294 read-only, out of scope).
2295 Patch from Wolfgang Breyha, report from Stuart Northfield.
2297 PP/29 Fix three issues highlighted by clang analyser static analysis.
2298 Only crash-plausible issue would require the Cambridge-specific
2299 iplookup router and a misconfiguration.
2300 Report from Marcin Mirosław.
2302 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2304 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2305 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2306 As part of this, removing so much warning spew let me fix some minor
2307 real issues in debug logging.
2309 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2310 assignment on my part. Fixed.
2312 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2313 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2314 Janne Snabb (who went above and beyond: thank you).
2316 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2317 string otherwise requires a connection and a bunch more work and it's
2318 relatively easy to get wrong. Should also expose TLS library linkage
2321 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2322 64-bit ${eval} (JH/03).
2324 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2325 GNU libc to support some of the 64-bit stuff, should not lead to
2326 conflicts. Defined before os.h is pulled in, so if a given platform
2327 needs to override this, it can.
2329 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2330 protection layer was required, which is not implemented.
2331 Bugzilla 1254, patch from Wolfgang Breyha.
2333 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2334 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2335 tls_dhparam take prime identifiers. Also unbreak combination of
2336 OpenSSL+DH_params+TLSSNI.
2338 PP/39 Disable SSLv2 by default in OpenSSL support.
2344 PP/01 Solaris build fix for Oracle's LDAP libraries.
2345 Bugzilla 1109, patch from Stephen Usher.
2347 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2349 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2352 TF/02 Fix a couple more cases where we did not log the error message
2353 when unlink() failed. See also change 4.74-TF/03.
2355 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2356 lock up or crash if it happened to be inside a call to libc when it
2357 got a SIGUSR1 from exiwhat.
2359 The SIGUSR1 handler appends the current process status to the process
2360 log which is later printed by exiwhat. It used to use the general
2361 purpose logging code to do this, but several functions it calls are
2362 not safe for signals.
2364 The new output code in the SIGUSR1 handler is specific to the process
2365 log, and simple enough that it's easy to inspect for signal safety.
2366 Removing some special cases also simplifies the general logging code.
2367 Removing the spurious timestamps from the process log simplifies
2370 TF/04 Improved ratelimit ACL condition.
2372 The /noupdate option has been deprecated in favour of /readonly which
2373 has clearer semantics. The /leaky, /strict, and /readonly update modes
2374 are mutually exclusive. The update mode is no longer included in the
2375 database key; it just determines when the database is updated. (This
2376 means that when you upgrade Exim will forget old rate measurements.)
2378 Exim now checks that the per_* options are used with an update mode that
2379 makes sense for the current ACL. For example, when Exim is processing a
2380 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2381 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2382 must specify per_mail/readonly. If you omit the update mode it defaults to
2383 /leaky where that makes sense (as before) or /readonly where required.
2385 The /noupdate option is now undocumented but still supported for
2386 backwards compatibility. It is equivalent to /readonly except that in
2387 ACLs where /readonly is required you may specify /leaky/noupdate or
2388 /strict/noupdate which are treated the same as /readonly.
2390 A useful new feature is the /count= option. This is a generalization
2391 of the per_byte option, so that you can measure the throughput of other
2392 aggregate values. For example, the per_byte option is now equivalent
2393 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2395 The per_rcpt option has been generalized using the /count= mechanism
2396 (though it's more complicated than the per_byte equivalence). When it is
2397 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2398 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2399 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2400 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2401 Note that using per_rcpt with a non-readonly update mode in more than
2402 one ACL will cause the recipients to be double-counted. (The per_mail
2403 and per_byte options don't have this problem.)
2405 The handling of very low rates has changed slightly. If the computed rate
2406 is less than the event's count (usually one) then this event is the first
2407 after a long gap. In this case the rate is set to the same as this event's
2408 count, so that the first message of a spam run is counted properly.
2410 The major new feature is a mechanism for counting the rate of unique
2411 events. The new per_addr option counts the number of different
2412 recipients that someone has sent messages to in the last time period. It
2413 behaves like per_rcpt if all the recipient addresses are different, but
2414 duplicate recipient addresses do not increase the measured rate. Like
2415 the /count= option this is a general mechanism, so the per_addr option
2416 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2417 example, measure the rate that a client uses different sender addresses
2418 with the options per_mail/unique=$sender_address. There are further
2419 details in the main documentation.
2421 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2423 TF/06 Removed a few PCRE remnants.
2425 TF/07 Automatically extract Exim's version number from tags in the git
2426 repository when doing development or release builds.
2428 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2429 Bugzilla 879. Patch from Paul Fisher.
2431 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2432 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2435 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2437 PP/05 Variable $av_failed, true if the AV scanner deferred.
2438 Bugzilla 1078. Patch from John Horne.
2440 PP/06 Stop make process more reliably on build failure.
2441 Bugzilla 1087. Patch from Heiko Schlittermann.
2443 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2444 Bugzilla 1089. Patch from Heiko Schlittermann.
2446 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2447 Bugzilla 1131. Patch from Holger Weiß.
2449 PP/09 Handle IPv6 addresses with SPF.
2450 Bugzilla 860. Patch from Wolfgang Breyha.
2452 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2454 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2457 PP/11 match_* no longer expand right-hand-side by default.
2458 New compile-time build option, EXPAND_LISTMATCH_RHS.
2459 New expansion conditions, "inlist", "inlisti".
2461 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2463 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2465 PP/14 fix log_write() format string regression from TF/03.
2466 Bugzilla 1152. Patch from Dmitry Isaikin.
2472 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2474 PP/02 Harmonised TLS library version reporting; only show if debugging.
2475 Layout now matches that introduced for other libraries in 4.74 PP/03.
2477 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2479 PP/04 New "dns_use_edns0" global option.
2481 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2484 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2485 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2487 TK/01 Updated PolarSSL code to 0.14.2.
2488 Bugzilla 1097. Patch from Andreas Metzler.
2490 PP/07 Catch divide-by-zero in ${eval:...}.
2491 Fixes bugzilla 1102.
2493 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2496 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2497 format-string attack -- SECURITY: remote arbitrary code execution.
2499 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2500 time unintentionally subject to list matching rules, letting the header
2501 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2502 arbitrary string expansion). This allowed for information disclosure.
2504 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2505 INT_MIN/-1 -- value coerced to INT_MAX.
2511 NM/01 Workaround for PCRE version dependency in version reporting
2514 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2515 This fixes portability to compilers other than gcc, notably
2516 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2518 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2519 makefiles for portability to HP-UX and POSIX correctness.
2521 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2522 Also via indented variable definition in the Makefile.
2523 (Debugging by Oliver Heesakkers).
2525 PP/02 Restore caching of spamd results with expanded spamd_address.
2526 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2528 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2529 Improves build reliability. Fix from: Frank Elsner
2531 NM/02 Fix wide character breakage in the rfc2047 coding
2532 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2534 NM/03 Allow underscore in dnslist lookups
2535 Fixes bug 1026. Patch from Graeme Fowler
2537 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2538 Code patches from Adam Ciarcinski of NetBSD.
2540 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2543 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2544 is logged, to avoid truncation. Patch from John Horne.
2546 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2547 Patch from Jakob Hirsch.
2549 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2550 SQL string expansion failure details.
2551 Patch from Andrey Oktyabrski.
2553 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2554 Patch from Simon Arlott.
2556 PP/09 New lookups functionality failed to compile on old gcc which rejects
2557 extern declarations in function scope.
2558 Patch from Oliver Fleischmann
2560 PP/10 Use sig_atomic_t for flags set from signal handlers.
2561 Check getgroups() return and improve debugging.
2562 Fixed developed for diagnosis in bug 927 (which turned out to be
2565 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2566 Patch from Mark Zealey.
2568 PP/12 Bugzilla 1056: Improved spamd server selection.
2569 Patch from Mark Zealey.
2571 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2572 Based on patch from Heiko Schlittermann.
2574 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2575 Patch from Uwe Doering, sign-off by Michael Haardt.
2577 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2578 variable declaration deep within a block. Bug and patch from
2581 PP/15 lookups-Makefile IRIX compatibility coercion.
2583 PP/16 Make DISABLE_DKIM build knob functional.
2585 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2586 Patch by Simon Arlott
2588 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2589 variable argument macros. Our copy now differs from upstream.
2595 TF/01 Failure to get a lock on a hints database can have serious
2596 consequences so log it to the panic log.
2598 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2599 controlled using the smtp_confirmation log selector.
2601 TF/03 Include the error message when we fail to unlink a spool file.
2603 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2604 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2605 for maintaining out-of-tree patches for some time.
2607 PP/01 Bugzilla 139: Documentation and portability issues.
2608 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2609 Handle per-OS dynamic-module compilation flags.
2611 PP/02 Let /dev/null have normal permissions.
2612 The 4.73 fixes were a little too stringent and complained about the
2613 permissions on /dev/null. Exempt it from some checks.
2614 Reported by Andreas M. Kirchwitz.
2616 PP/03 Report version information for many libraries, including
2617 Exim version information for dynamically loaded libraries. Created
2618 version.h, now support a version extension string for distributors
2619 who patch heavily. Dynamic module ABI change.
2621 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2622 privilege escalation vulnerability whereby the Exim run-time user
2623 can cause root to append content of the attacker's choosing to
2626 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2629 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2630 If dropping privileges for untrusted macros, we disabled normal logging
2631 on the basis that it would fail; for the Exim run-time user, this is not
2632 the case, and it resulted in successful deliveries going unlogged.
2633 Fixed. Reported by Andreas Metzler.
2639 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2640 only prepend for the Resent-* case. Fixes regression introduced in
2641 Exim 4.70 by NM/22 for Bugzilla 607.
2643 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2644 increasing numbers of administrators be bitten by this.
2646 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2648 PP/03 Bugzilla 994: added openssl_options main configuration option.
2650 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2652 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2654 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2656 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2657 without a peer certificate, leading to a segfault because of an
2658 assumption that peers always have certificates. Be a little more
2659 paranoid. Problem reported by Martin Tscholak.
2661 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2662 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2663 NB: ClamAV planning to remove STREAM in "middle of 2010".
2664 CL also introduces -bmalware, various -d+acl logging additions and
2665 more caution in buffer sizes.
2667 PP/09 Implemented reverse_ip expansion operator.
2669 PP/10 Bugzilla 937: provide a "debug" ACL control.
2671 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2673 PP/12 Bugzilla 973: Implement --version.
2675 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2677 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2679 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2681 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2682 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2683 ignore trailing whitespace.
2685 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2687 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2690 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2691 Notification from Dr Andrew Aitchison.
2693 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2694 ExtendedDetectionInfo response format.
2695 Notification from John Horne.
2697 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2700 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2701 XSL and documented dependency on system catalogs, with examples of how
2704 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2707 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2708 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2709 configuration file which is writeable by the Exim user or group.
2711 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2712 of configuration files to cover files specified with the -C option if
2713 they are going to be used with root privileges, not just the default
2716 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2717 option (effectively making it always true).
2719 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2720 files to be used while preserving root privileges.
2722 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2723 that rogue child processes cannot use them.
2725 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2726 run-time user, instead of root.
2728 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2729 Exim run-time user without dropping privileges.
2731 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2732 result string, instead of calling string_vformat() twice with the same
2735 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2736 for other users. Others should always drop root privileges if they use
2737 -C on the command line, even for a whitelisted configure file.
2739 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2741 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2747 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2748 $data_path, and $header_path variables; fixed documentation bugs and
2751 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2752 exipick to access non-standard spools, including the "frozen" queue
2755 NM/01 Bugzilla 965: Support mysql stored procedures.
2756 Patch from Alain Williams
2758 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2760 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2761 Patch from Andreas Metzler
2763 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2764 Patch from Kirill Miazine
2766 NM/05 Bugzilla 671: Added umask to procmail example.
2768 JJ/03 installed exipick 20100323.0, fixing doc bug
2770 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2771 directory. Notification and patch from Dan Rosenberg.
2773 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2775 TK/02 Improve log output when DKIM signing operation fails.
2777 MH/01 Treat the transport option dkim_domain as a colon separated
2778 list, not as a single string, and sign the message with each element,
2779 omitting multiple occurences of the same signer.
2781 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2782 Bugzilla 985, 986. Patch by Simon Arlott
2784 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2785 Patch by Simon Arlott
2787 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2788 MBX locking. Notification from Dan Rosenberg.
2794 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2796 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2798 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2800 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2802 NM/04 Bugzilla 915: Flag broken perl installation during build.
2808 TK/01 Added patch by Johannes Berg that expands the main option
2809 "spamd_address" if it starts with a dollar sign.
2811 TK/02 Write list of recipients to X-Envelope-Sender header when building
2812 the mbox-format spool file for content scanning (suggested by Jakob
2815 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2816 (http://www.dcc-servers.net/) support via dccifd. Activated by
2817 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2819 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2820 by Mark Daniel Reidel <mr@df.eu>.
2822 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2823 When building exim an external PCRE library is now needed -
2824 PCRE is a system library on the majority of modern systems.
2825 See entry on PCRE_LIBS in EDITME file.
2827 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2828 conversation. Added nologin parameter to request.
2829 Patch contributed by Kirill Miazine.
2831 TF/01 Do not log submission mode rewrites if they do not change the address.
2833 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2835 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2836 log files in place. Contributed by Roberto Lima.
2838 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2840 TF/03 Bugzilla 615: When checking the local_parts router precondition
2841 after a local_part_suffix or local_part_prefix option, Exim now
2842 does not use the address's named list lookup cache, since this
2843 contains cached lookups for the whole local part.
2845 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2846 Robert Millan. Documentation is in experimental-spec.txt.
2848 TF/04 Bugzilla 668: Fix parallel build (make -j).
2850 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2852 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2853 Patch provided by Jan Srzednicki.
2855 TF/05 Leading white space used to be stripped from $spam_report which
2856 wrecked the formatting. Now it is preserved.
2858 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2859 that they are available at delivery time.
2861 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2863 TF/08 TLS error reporting now respects the incoming_interface and
2864 incoming_port log selectors.
2866 TF/09 Produce a more useful error message if an SMTP transport's hosts
2867 setting expands to an empty string.
2869 NM/06 Bugzilla 744: EXPN did not work under TLS.
2870 Patch provided by Phil Pennock.
2872 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2873 Patch provided by Richard Godbee.
2875 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2876 acl_smtp_notquit, added index entry.
2878 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2879 Patch provided by Eugene Bujak.
2881 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2882 accept(). Patch provided by Maxim Dounin.
2884 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2885 Patch provided by Phil Pennock.
2887 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2889 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2890 Patch provided by Brad "anomie" Jorsch.
2892 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2893 Patch provided by Dean Brooks.
2895 TK/05 Add native DKIM support (does not depend on external libraries).
2897 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2898 Patch provided by Graeme Fowler.
2900 NM/16 Bugzilla 851: Documentation example syntax fix.
2902 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2904 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2907 NM/19 Bugzilla 745: TLS version reporting.
2908 Patch provided by Phil Pennock.
2910 NM/20 Bugzilla 167: bool: condition support.
2911 Patch provided by Phil Pennock.
2913 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2914 clients. Patch provided by Phil Pennock.
2916 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2917 Patch provided by Brad "anomie" Jorsch.
2919 NM/23 Bugzilla 687: Fix misparses in eximstats.
2920 Patch provided by Heiko Schlittermann.
2922 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2923 Patch provided by Heiko Schlittermann.
2925 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2926 plus update to original patch.
2928 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2930 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2931 Patch provided by David Brownlee.
2933 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2935 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2937 NM/30 Bugzilla 888: TLS documentation bugfixes.
2939 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2941 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2942 Unlike the original bugzilla I have changed all shell scripts in src tree.
2944 NM/33 Bugzilla 898: Transport filter timeout fix.
2945 Patch by Todd Rinaldo.
2947 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2948 Patch by Serge Demonchaux.
2950 NM/35 Bugzilla 39: Base64 decode bug fixes.
2951 Patch by Jakob Hirsch.
2953 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2955 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2957 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2959 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2965 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2966 ALT-N's libdkim that I have put here:
2967 http://duncanthrax.net/exim-experimental/
2969 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2970 were called 'true' and it seems that C99 defines that as a reserved
2971 keyword to be used with 'bool' variable types. That means you could
2972 not include C99-style headers which use bools without triggering
2973 build errors in sieve.c.
2975 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2976 as mailq or other aliases. Changed the --help handling significantly
2977 to do whats expected. exim_usage() emits usage/help information.
2979 SC/01 Added the -bylocaldomain option to eximstats.
2981 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2983 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2985 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2991 PH/01 Another patch from the Sieve maintainer.
2993 PH/02 When an IPv6 address is converted to a string for single-key lookup
2994 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2995 dots are used instead of colons so that keys in lsearch files need not
2996 contain colons. This was done some time before quoting was made available
2997 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2998 (notated using the quote facility) so as to distinguish them from IPv4
2999 keys. This meant that lookups for IP addresses in host lists did not work
3000 for iplsearch lookups.
3002 This has been fixed by arranging for IPv6 addresses to be expressed with
3003 colons if the lookup type is iplsearch. This is not incompatible, because
3004 previously such lookups could never work.
3006 The situation is now rather anomalous, since one *can* have colons in
3007 ordinary lsearch keys. However, making the change in all cases is
3008 incompatible and would probably break a number of configurations.
3010 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3013 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3014 conversion specification without a maximum field width, thereby enabling
3015 a rogue spamd server to cause a buffer overflow. While nobody in their
3016 right mind would setup Exim to query an untrusted spamd server, an
3017 attacker that gains access to a server running spamd could potentially
3018 exploit this vulnerability to run arbitrary code as the Exim user.
3020 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3021 $primary_hostname instead of what libspf2 thinks the hosts name is.
3023 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3024 a directory entry by the name of the lookup key. Previously, if a
3025 symlink pointed to a non-existing file or a file in a directory that
3026 Exim lacked permissions to read, a lookup for a key matching that
3027 symlink would fail. Now it is enough that a matching directory entry
3028 exists, symlink or not. (Bugzilla 503.)
3030 PH/03 The body_linecount and body_zerocount variables are now exported in the
3033 PH/04 Added the $dnslist_matched variable.
3035 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3036 This means they are set thereafter only if the connection becomes
3039 PH/06 Added the client_condition to authenticators so that some can be skipped
3040 by clients under certain conditions.
3042 PH/07 The error message for a badly-placed control=no_multiline_responses left
3043 "_responses" off the end of the name.
3045 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3047 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3048 (without spaces) instead of just copying the configuration text.
3050 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3052 PH/11 Added $max_received_linelength.
3054 PH/12 Added +ignore_defer and +include_defer to host lists.
3056 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3057 way in which PCRE > 7.0 is built.
3059 PH/14 Implemented queue_only_load_latch.
3061 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3062 MAIL command. The effect was to mangle the value on 64-bit systems.
3064 PH/16 Another patch from the Sieve maintainer.
3066 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3068 PH/18 If a system quota error occurred while trying to create the file for
3069 a maildir delivery, the message "Mailbox is full" was not appended to the
3070 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3071 only to a quota excession during the actual writing of the file.
3073 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3074 characters?) which causes problems in log lines. The DN values are now
3075 passed through string_printing() before being added to log lines.
3077 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3078 and InterBase are left for another time.)
3080 PH/21 Added message_body_newlines option.
3082 PH/22 Guard against possible overflow in moan_check_errorcopy().
3084 PH/23 POSIX allows open() to be a macro; guard against that.
3086 PH/24 If the recipient of an error message contained an @ in the local part
3087 (suitably quoted, of course), incorrect values were put in $domain and
3088 $local_part during the evaluation of errors_copy.
3094 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3095 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3098 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3099 issue a MAIL command.
3101 PH/02 In an ACL statement such as
3103 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3105 if a client was not listed at all, or was listed with a value other than
3106 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3107 the condition was not true (as it should be), so access was not denied.
3108 The bug was that the ! inversion was incorrectly passed on to the second
3109 item. This has been fixed.
3111 PH/03 Added additional dnslists conditions == and =& which are different from
3112 = and & when the dns lookup returns more than one IP address.
3114 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3115 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3117 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3118 FSYNC, which compiles an option called disable_fsync that allows for
3119 bypassing fsync(). The documentation is heavily laced with warnings.
3121 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3123 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3124 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3125 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3126 including adding "make clean"; (3) Added -fPIC when compiling the test
3127 dynamically loaded module, to get rid of a warning.
3129 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3130 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3131 = 0s. The bug is otherwise harmless.
3133 PH/07 There was a bug in the dovecot authenticator such that the value of
3134 $auth1 could be overwritten, and so not correctly preserved, after a
3135 successful authentication. This usually meant that the value preserved by
3136 the server_setid option was incorrect.
3138 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3140 PH/09 Installed PCRE release 7.0.
3142 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3143 run for batched SMTP input. It is now run at the start of every message
3144 in the batch. While fixing this I discovered that the process information
3145 (output by running exiwhat) was not always getting set for -bs and -bS
3146 input. This is fixed, and it now also says "batched" for BSMTP.
3148 PH/11 Added control=no_pipelining.
3150 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3151 patch, slightly modified), and move the expansion of helo_data till after
3152 the connection is made in the smtp transport (so it can use these
3155 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3157 PH/14 Added log_selector = +pid.
3159 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3161 PH/16 Add ${if forany and ${if forall.
3163 PH/17 Added dsn_from option to vary the From: line in DSNs.
3165 PH/18 Flush SMTP output before performing a callout, unless control =
3166 no_callout_flush is set.
3168 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3169 was true (the default) a successful delivery failed to delete the retry
3170 item, thus causing premature timeout of the address. The bug is now
3173 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3175 PH/21 Long custom messages for fakedefer and fakereject are now split up
3176 into multiline responses in the same way that messages for "deny" and
3177 other ACL rejections are.
3179 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3180 with slight modification.
3182 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3183 draft, changing the syntax and factoring some duplicate code".
3185 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3186 for deliveries of the second and subsequent messages over the same SMTP
3189 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3190 ${reduce, with only minor "tidies".
3192 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3194 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3195 expansion side effects.
3197 PH/27 When a message times out after an over-quota error from an Exim-imposed
3198 quota, the bounce message says "mailbox is full". This message was not
3199 being given when it was a system quota that was exceeded. It now should
3202 MH/03 Made $recipients available in local_scan(). local_scan() already has
3203 better access to the recipient list through recipients_list[], but
3204 $recipients can be useful in postmaster-provided expansion strings.
3206 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3207 in the case of a MAIL command with additional options following the
3208 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3209 were accidentally chopped off.
3211 PH/29 SMTP synchronization checks are implemented when a command is read -
3212 there is a check that no more input is waiting when there shouldn't be
3213 any. However, for some commands, a delay in an ACL can mean that it is
3214 some time before the response is written. In this time, more input might
3215 arrive, invalidly. So now there are extra checks after an ACL has run for
3216 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3217 pipelining has not been advertised.
3219 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3221 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3222 returned in the SMTP response when smtp_return_error_details was false.
3223 This has been fixed.
3225 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3226 instead of the C I/O that was originally supplied, because problems were
3227 reported on Solaris.
3229 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3230 Exim which did not show up earlier: it was assuming that a call to
3231 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3232 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3233 was a macro that became an assignment, so it seemed to work. This has
3234 changed to a proper function call with a void return, hence the compile
3235 error. Exim's code has been fixed.
3237 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3240 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3242 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3244 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3245 because it turns out to be too big or there is a timeout) there is no
3246 "Completed" line in the log. When some messages of this type were
3247 selected by exigrep, they were listed as "not completed". Others were
3248 picked up by some special patterns. I have improved the selection
3249 criteria to be more general.
3251 PH/38 The host_find_failed option in the manualroute router can now be set
3252 to "ignore", to completely ignore a host whose IP address cannot be
3253 found. If all hosts are ignored, the behaviour is controlled by the new
3254 host_all_ignored option.
3256 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3257 homing or because of multiple MX records with /mx) generated more than
3258 one IP address, and the following item turned out to be the local host,
3259 all the secondary addresses of the first item were incorrectly removed
3260 from the list, along with the local host and any following hosts (which
3261 is what is supposed to happen).
3263 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3264 whoever called Exim into the -H file. In the case of the daemon it was
3265 behaving confusingly. When first started, it used values for whoever
3266 started the daemon, but after a SIGHUP it used the Exim user (because it
3267 calls itself on a restart). I have changed the code so that it now always
3270 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3271 message are rejected with the same error (e.g. no authentication or bad
3272 sender address), and a DATA command is nevertheless sent (as can happen
3273 with PIPELINING or a stupid MUA), the error message that was given to the
3274 RCPT commands is included in the rejection of the DATA command. This is
3275 intended to be helpful for MUAs that show only the final error to their
3278 PH/42 Another patch from the Sieve maintainer.
3280 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3281 Eximstats - Fixed some broken HTML links and added missing column headers
3283 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3284 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3286 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3288 PH/43 Yet another patch from the Sieve maintainer.
3290 PH/44 I found a way to check for a TCP/IP connection going away before sending
3291 the response to the final '.' that terminates a message, but only in the
3292 case where the client has not sent further data following the '.'
3293 (unfortunately, this is allowed). However, in many cases there won't be
3294 any further data because there won't be any more messages to send. A call
3295 to select() can be used: if it shows that the input is "ready", there is
3296 either input waiting, or the socket has been closed. An attempt to read
3297 the next input character can distinguish the two cases. Previously, Exim
3298 would have sent an OK response which the client would never have see.
3299 This could lead to message repetition. This fix should cure that, at
3300 least in a lot of common cases.
3302 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3303 advertised in response to EHLO.
3309 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3310 fixed by 4.65/MH/01 (is this a record?) are fixed:
3312 (i) An empty string was always treated as zero by the numeric comparison
3313 operators. This behaviour has been restored.
3315 (ii) It is documented that the numeric comparison operators always treat
3316 their arguments as decimal numbers. This was broken in that numbers
3317 starting with 0 were being interpreted as octal.
3319 While fixing these problems I realized that there was another issue that
3320 hadn't been noticed. Values of message_size_limit (both the global option
3321 and the transport option) were treated as octal if they started with 0.
3322 The documentation was vague. These values are now always treated as
3323 decimal, and I will make that clear in the documentation.
3329 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3330 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3333 MH/01 Don't check that the operands of numeric comparison operators are
3334 integers when their expansion is in "skipping" mode (fixes bug
3335 introduced by 4.64-PH/07).
3337 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3338 child addresses, Exim now panics and dies. Previously, because the count
3339 is held in a short int, deliveries were likely to be lost. As such a
3340 large number of recipients for a single message is ridiculous
3341 (performance will be very, very poor), I have chosen to impose a limit
3342 rather than extend the field.
3348 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3349 leftover -K file (the existence of which was triggered by #402).
3350 While we were at it, introduced process PID as part of the -K
3351 filename. This should rule out race conditions when creating
3354 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3355 processing considerably. Previous code took too long for large mails,
3356 triggering a timeout which in turn triggers #401.
3358 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3359 in the DK code in transports.c. sendfile() is not really portable,
3360 hence the _LINUX specificness.
3362 TF/01 In the add_headers option to the mail command in an Exim filter,
3363 there was a bug that Exim would claim a syntax error in any
3364 header after the first one which had an odd number of characters
3367 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3368 callout verification, Exim cached a "reject" for the entire domain. This
3369 is correct for most verifications, but it is not correct for a recipient
3370 verification with use_sender or use_postmaster set, because in that case
3371 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3372 case of MAIL FROM:<> rejection from other early rejections (e.g.
3373 rejection of HELO). When verifying a recipient using a non-null MAIL
3374 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3375 Whatever the result of the callout, the value of the domain cache is
3376 left unchanged (for any other kind of callout, getting as far as trying
3377 RCPT means that the domain itself is ok).
3379 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3382 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3383 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3384 session, the daemon ignores ECONNECT errors and logs others; it now
3385 ignores EPIPE as well.
3387 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3388 (quoted-printable decoding).
3390 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3391 later the small subsequent patch to fix an introduced bug.
3393 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3395 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3397 PH/08 An error is now given if message_size_limit is specified negative.
3399 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3400 to be given (somewhat) arbitrary names.
3402 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3405 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3406 miscellaneous code fixes
3408 PH/10 Added the log_reject_target ACL modifier to specify where to log
3411 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3412 hostname. This is wrong, because it relates to the incoming message (and
3413 probably the interface on which it is arriving) and not to the outgoing
3414 callout (which could be using a different interface). This has been
3415 changed to use the value of the helo_data option from the smtp transport
3416 instead - this is what is used when a message is actually being sent. If
3417 there is no remote transport (possible with a router that sets up host
3418 addresses), $smtp_active_hostname is used.
3420 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3421 tweaks were necessary in order to get it to work (see also 21 below):
3422 (a) The code assumed that strncpy() returns a negative number on buffer
3423 overflow, which isn't the case. Replaced with Exim's string_format()
3425 (b) There were several signed/unsigned issues. I just did the minimum
3426 hacking in of casts. There is scope for a larger refactoring.
3427 (c) The code used strcasecmp() which is not a standard C function.
3428 Replaced with Exim's strcmpic() function.
3429 (d) The code set only $1; it now sets $auth1 as well.
3430 (e) A simple test gave the error "authentication client didn't specify
3431 service in request". It would seem that Dovecot has changed its
3432 interface. Fortunately there's a specification; I followed it and
3433 changed what the client sends and it appears to be working now.
3435 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3438 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3439 address A is aliased to B and C, where B exists and C does not. Without
3440 -v the output is "A verified" because verification stops after a
3441 successful redirection if more than one address is generated. However,
3442 with -v the child addresses are also verified. Exim was outputting "A
3443 failed to verify" and then showing the successful verification for C,
3444 with its parentage. It now outputs "B failed to verify", showing B's
3445 parentage before showing the successful verification of C.
3447 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3448 look up a TXT record in a specific list after matching in a combined
3451 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3452 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3453 they consult the DNS. I had assumed they would set it the way they
3454 wanted; and indeed my experiments on Linux seem to show that in some
3455 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3456 To be on the safe side, however, I have now made the interface to
3457 host_find_byname() similar to host_find_bydns(), with an argument
3458 containing the DNS resolver options. The host_find_byname() function now
3459 sets these options at its start, just as host_find_bydns() does. The smtp
3460 transport options dns_qualify_single and dns_search_parents are passed to
3461 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3462 of host_find_byname() use the default settings of RES_DEFNAMES
3463 (qualify_single) but not RES_DNSRCH (search_parents).
3465 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3466 spool_read_header() do less string testing, by means of a preliminary
3467 switch on the second character of optional "-foo" lines. (This is
3468 overdue, caused by the large number of possibilities that now exist.
3469 Originally there were few.) While I was there, I also converted the
3470 str(n)cmp tests so they don't re-test the leading "-" and the first
3471 character, in the hope this might squeeze out yet more improvement.
3473 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3474 flag allowing group syntax was set by the header_syntax check but not
3475 turned off, possible causing trouble later; (2) The flag was not being
3476 set at all for the header_verify test, causing "group"-style headers to
3477 be rejected. I have now set it in this case, and also caused header_
3478 verify to ignore an empty address taken from a group. While doing this, I
3479 came across some other cases where the code for allowing group syntax
3480 while scanning a header line wasn't quite right (mostly, not resetting
3481 the flag correctly in the right place). These bugs could have caused
3482 trouble for malformed header lines. I hope it is now all correct.
3484 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3485 with the "reply" argument non-NULL. The code, however (which originally
3486 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3487 but it didn't always do it. This confused somebody who was copying the
3488 code for some other use. I have removed all the tests.
3490 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3491 feature that was used to support insecure browsers during the U.S. crypto
3492 embargo. It requires special client support, and Exim is probably the
3493 only MTA that supported it -- and would never use it because real RSA is
3494 always available. This code has been removed, because it had the bad
3495 effect of slowing Exim down by computing (never used) parameters for the
3496 RSA_EXPORT functionality.
3498 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3499 authenticator to fail if there's a tab character in the incoming data
3500 (there should never be unless someone is messing about, as it's supposed
3501 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3502 if the connection is using TLS or if the remote IP is the same as the
3503 local IP, and the "valid-client-cert option" if a client certificate has
3506 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3507 authenticators. This can be used for authorization after authentication
3508 succeeds. (In the case of plaintext, it servers for both authentication
3511 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3512 if any retry times were supplied.
3514 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3515 connection, where there is no client IP address to check. In this
3516 situation, the verify now always succeeds.
3518 PH/25 Applied John Jetmore's -Mset patch.
3520 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3522 PH/27 In a string expansion for a processed (not raw) header when multiple
3523 headers of the same name were present, leading whitespace was being
3524 removed from all of them, but trailing whitespace was being removed only
3525 from the last one. Now trailing whitespace is removed from each header
3526 before concatenation. Completely empty headers in a concatenation (as
3527 before) are ignored.
3529 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3530 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3532 PH/29 [Removed. This was a change that I later backed out, and forgot to
3533 correct the ChangeLog entry (that I had efficiently created) before
3534 committing the later change.]
3536 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3537 address errors (4xx response to RCPT) over the same connection as other
3538 messages routed to the same hosts. Such deliveries are always "forced",
3539 so retry times are not inspected. This resulted in far too many retries
3540 for the affected addresses. The effect occurred only when there were more
3541 hosts than the hosts_max_try setting in the smtp transport when it had
3542 the 4xx errors. Those hosts that it had tried were not added to the list
3543 of hosts for which the message was waiting, so if all were tried, there
3544 was no problem. Two fixes have been applied:
3546 (i) If there are any address or message errors in an SMTP delivery, none
3547 of the hosts (tried or untried) are now added to the list of hosts
3548 for which the message is waiting, so the message should not be a
3549 candidate for sending over the same connection that was used for a
3550 successful delivery of some other message. This seems entirely
3551 reasonable: after all the message is NOT "waiting for some host".
3552 This is so "obvious" that I'm not sure why it wasn't done
3553 previously. Hope I haven't missed anything, but it can't do any
3554 harm, as the worst effect is to miss an optimization.
3556 (ii) If, despite (i), such a delivery is accidentally attempted, the
3557 routing retry time is respected, so at least it doesn't keep
3558 hammering the server.
3560 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3561 in ${readsocket because some servers need this prod.
3563 PH/32 Added some extra debug output when updating a wait-xxx database.
3565 PH/33 The hint "could be header name not terminated by colon", which has been
3566 given for certain expansion errors for a long time, was not being given
3567 for the ${if def:h_colon_omitted{... case.
3569 PH/34 The spec says: "With one important exception, whenever a domain list is
3570 being scanned, $domain contains the subject domain." There was at least
3571 one case where this was not true.
3573 PH/35 The error "getsockname() failed: connection reset by peer" was being
3574 written to the panic log as well as the main log, but it isn't really
3575 panic-worthy as it just means the connection died rather early on. I have
3576 removed the panic log writing for the ECONNRESET error when getsockname()
3579 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3580 runs only) independently of the message's sender address. This meant
3581 that, if the 4xx error was in fact related to the sender, a different
3582 message to the same recipient with a different sender could confuse
3583 things. In particular, this can happen when sending to a greylisting
3584 server, but other circumstances could also provoke similar problems.
3585 I have changed the default so that the retry time for these errors is now
3586 based a combination of the sender and recipient addresses. This change
3587 can be overridden by setting address_retry_include_sender=false in the
3590 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3591 remote server are returned as part of bounce messages. This was not
3592 happening for LMTP over a pipe (the lmtp transport), but now it is the
3593 same for both kinds of LMTP.
3595 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3596 in header lines that were in fact CNAMEs. This is no longer the case.
3598 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3599 and queue runs started by the daemon processed all messages. This has
3600 been fixed so that -R and -S can now usefully be given with -q<time>.
3602 PH/40 Import PCRE release 6.7 (fixes some bugs).
3604 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3606 PH/42 Give an error if -q is specified more than once.
3608 PH/43 Renamed the variables $interface_address and $interface_port as
3609 $received_ip_address and $received_port, to make it clear that these
3610 values apply to message reception, and not to the outgoing interface when
3611 a message is delivered. (The old names remain recognized, of course.)
3613 PH/44 There was no timeout on the connect() call when using a Unix domain
3614 socket in the ${readsocket expansion. There now is.
3616 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3617 be meaningful with "accept".
3619 SC/01 Eximstats V1.43
3620 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3622 SC/02 Eximstats V1.44
3623 Use a glob alias rather than an array ref in the generated
3624 parser. This improves both readability and performance.
3626 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3627 Collect SpamAssassin and rejection statistics.
3628 Don't display local sender or destination tables unless
3629 there is data to show.
3630 Added average volumes into the top table text output.
3632 SC/04 Eximstats V1.46
3633 Collect data on the number of addresses (recipients)
3634 as well as the number of messages.
3636 SC/05 Eximstats V1.47
3637 Added 'Message too big' to the list of mail rejection
3638 reasons (thanks to Marco Gaiarin).
3640 SC/06 Eximstats V1.48
3641 Mainlog lines which have GMT offsets and are too short to
3642 have a flag are now skipped.
3644 SC/07 Eximstats V1.49 (Alain Williams)
3645 Added the -emptyok flag.
3647 SC/08 Eximstats V1.50
3648 Fixes for obtaining the IP address from reject messages.
3650 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3651 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3652 whitespace changes from 4.64-PH/27
3654 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3657 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3658 are found, allow negative numbers in numeric criteria)
3660 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3662 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3663 to match changes made in 4.64-PH/43
3665 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3667 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3668 "domain" is actually an IP address, and force a failure. This locks out
3669 those revolvers/nameservers that support "A-for-A" lookups, in
3670 contravention of the specifications.
3672 PH/48 When a host name was looked up from an IP address, and the subsequent
3673 forward lookup of the name timed out, the host name was left in
3674 $sender_host_name, contrary to the specification.
3676 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3677 restricted to single-key lookups, Exim was not diagnosing an error if
3678 * or *@ was used with a query-style lookup.
3680 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3682 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3683 long ago, but noone interested enough thought of it. Let's just say that
3684 the "1.1" means that there are some new functions that weren't there at
3685 some point in the past.
3687 PH/51 Error processing for expansion failure of helo_data from an smtp
3688 transport during callout processing was broken.
3690 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3691 tested/used via the -bh/-bhc/-bs options.
3693 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3694 bug, fixed in subsequent PCRE releases).
3696 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3697 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3699 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3705 SC/01 Use a glob alias rather than an array ref in eximstats generated
3706 parser. This improves both readability and performance.
3708 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3709 Don't display local sender or destination tables in eximstats unless
3710 there is data to show.
3711 Added average volumes into the eximstats top table text output.
3713 SC/03 Collect data on the number of addresses (recipients) as well
3714 as the number of messages in eximstats.
3716 TF/01 Correct an error in the documentation for the redirect router. Exim
3717 does (usually) call initgroups() when daemonizing.
3719 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3720 with consistent privilege compared to when running as a daemon.
3722 TF/03 Note in the spec that $authenticated_id is not set for local
3723 submissions from trusted users.
3725 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3726 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3728 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3729 by adding some example configuration directives to the default
3730 configuration file. A little bit of work is required to uncomment the
3731 directives and define how usernames and passwords are checked, but
3732 there is now a framework to start from.
3734 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3735 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3736 without this. I don't know how relevant this is to other LDAP libraries.
3738 PH/02 Add the verb name to the "unknown ACL verb" error.
3740 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3742 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3744 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3745 directory not expanded when it should be if an expanded home directory
3746 was set for the address (which is overridden by the transport).
3748 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3751 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3752 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3753 because it is too late at that time, and has no effect.
3755 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3756 security issue with \' (bugzilla #107). I could not use the
3757 PQescapeStringConn() function, because it needs a PGconn value as one of
3760 PH/08 When testing addresses using -bt, indicate those final addresses that
3761 are duplicates that would not cause an additional delivery. At least one
3762 person was confused, thinking that -bt output corresponded to deliveries.
3763 (Suppressing duplicates isn't a good idea as you lose the information
3764 about possibly different redirections that led to the duplicates.)
3766 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3767 systems where poll() doesn't work, in particular OS X.
3769 PH/10 Added more information to debugging output for retry time not reached.
3771 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3772 operations in malware.c.
3774 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3777 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3778 syslog_duplication set false, log lines that would normally be written
3779 both the the main log and to the reject log were not written to syslog at
3782 PH/14 In the default configuration, change the use of "message" in ACL warn
3783 statements to "add_header".
3785 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3786 not followed by a command (e.g. "seen endif").
3788 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3789 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3792 PH/17 Added extra conditions to the default value of delay_warning_condition
3796 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3797 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3798 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3801 The Auto-Submitted: and various List- headers are standardised, whereas I
3802 don't think Precedence: ever was.
3804 PH/18 Refactored debugging code in route_finduser() to show more information,
3805 in particular, the error code if getpwnam() issues one.
3807 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3808 This is apparently needed in addition to the PH/07 change above to avoid
3809 any possible encoding problems.
3811 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3812 but not after initializing Perl.
3814 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3815 output them only if debugging. By default they are written stderr,
3816 apparently, which is not desirable.
3818 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3821 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3824 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3826 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3827 authenticated or an ident call has been made. Suppress the default
3828 values for $authenticated_id and $authenticated_sender (but permit -oMai
3829 and -oMas) when testing with -bh.
3831 PH/24 Re-jigged the order of the tests in the default configuration so that the
3832 tests for valid domains and recipients precede the DNS black list and CSA
3833 tests, on the grounds that those ones are more expensive.
3835 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3836 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3837 This bug exists in every version of Exim that I still have, right back to
3840 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3841 However, an attempt to turn on case-sensitivity in a regex key by
3842 including (?-i) didn't work because the subject string was already
3843 lowercased, and the effects were non-intuitive. It turns out that a
3844 one-line patch can be used to allow (?-i) to work as expected.
3850 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3851 other effects) broke the use of negated acl sub-conditions.
3853 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3856 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3857 "Deny" causes Exim to reject the incoming connection with a 554 error.
3858 Unfortunately, if there is a major crisis, such as a disk failure,
3859 tcp-wrappers gives "deny", whereas what one would like would be some
3860 kind of temporary error. A kludge has been added to help with this.
3861 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3862 554 error is used if errno is still zero or contains ENOENT (which occurs
3863 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3866 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3868 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3869 errors. Otherwise a message that provokes a temporary error (when other
3870 messages do not) can cause a whole host to time out.
3872 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3873 addresses were routed directly to files or pipes from a redirect router.
3874 File deliveries just didn't batch; pipe deliveries might have suffered
3877 PH/06 A failure to get a lock for a hints database would erroneously always say
3878 "Failed to get write lock", even when it was really a read lock.
3880 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3881 of 0600. This has been changed to use the value of the lockfile_mode
3882 option (which defaults to 0600).
3884 PH/08 Applied small patch from the Sieve maintainer.
3886 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3887 folder from quota calculations, a direct delivery into this folder messed
3888 up the contents of the maildirsize file. This was because the regex was
3889 used only to exclude .Trash (or whatever) when the size of the mailbox
3890 was calculated. There was no check that a delivery was happening into an
3891 excluded directory. This bug has been fixed by ignoring all quota
3892 processing for deliveries into excluded directories.
3894 PH/10 Added the maildirfolder_create_regex option to appendfile.
3900 PH/01 The code for finding all the local interface addresses on a FreeBSD
3901 system running IPv6 was broken. This may well have applied to all BSD
3902 systems, as well as to others that have similar system calls. The broken
3903 code found IPv4 interfaces correctly, but gave incorrect values for the
3904 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3905 that it would not match correctly against @[] and not recognize the IPv6
3908 PH/02 The ipliteral router was not recognizing addresses of the form user@
3909 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3911 PH/03 Added disable_ipv6.
3913 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3914 decoded form, because it is most often used to construct To: headers
3915 lines in autoreplies, and the decoded form may well be syntactically
3916 invalid. However, $reply_address has leading white space removed, and all
3917 newlines turned into spaces so that the autoreply transport does not
3920 PH/05 If group was specified without a user on a router, and no group or user
3921 was specified on a transport, the group from the router was ignored.
3923 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3924 for visible compile-time settings that can be used to change these
3925 numbers, for those that want even more. Backwards compatibility with old
3926 spool files has been maintained. However, going back to a previous Exim
3927 release will lost any variables that are in spool files.
3929 PH/07 Two small changes when running in the test harness: increase delay when
3930 passing a TCP/IP connection to a new process, in case the original
3931 process has to generate a bounce, and remove special handling of
3932 127.0.0.2 (sic), which is no longer necessary.
3934 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3935 be the same on different OS.
3937 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3940 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3941 whether --show-vars was specified or not
3943 JJ/02 exipick: Added support for new ACL variable spool format introduced
3946 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3947 syntactically invalid From: or Reply-to: line, and a filter used this to
3948 generate an autoreply, and therefore failed to obtain an address for the
3949 autoreply, Exim could try to deliver to a non-existent relative file
3950 name, causing unrelated and misleading errors. What now happens is that
3951 it logs this as a hard delivery error, but does not attempt to create a
3954 PH/11 The exinext utility has a -C option for testing purposes, but although
3955 the given file was scanned by exinext itself; it wasn't being passed on
3956 when Exim was called.
3958 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3959 an end-of-file indication when reading a command response.
3961 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3962 compiled. In many other places in Exim, IPv6 addresses are always
3963 recognized, so I have changed this. It also means that IPv4 domain
3964 literals of the form [IPV4:n.n.n.n] are now always recognized.
3966 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3967 used if the router is not running as root, for example, when verifying at
3968 ACL time, or when using -bh. The debugging output from this situation was
3969 non-existent - all you got was a failure to exec. I have made two
3972 (a) Failures to set uid/gid, the current directory, or a process leader
3973 in a subprocess such as that created by queryprogram now generate
3974 suitable debugging output when -d is set.
3976 (b) The queryprogram router detects when it is not running as root,
3977 outputs suitable debugging information if -d is set, and then runs
3978 the subprocess without attempting to change uid/gid.
3980 PH/15 Minor change to Makefile for building test_host (undocumented testing
3983 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3984 additional section of a DNS packet that returns MX or SRV records.
3985 Instead, it always explicitly searches for A/AAAA records. This avoids
3986 major problems that occur when a DNS server includes only records of one
3987 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3988 fixed another bug: if SRV records were looked up and the corresponding
3989 address records were *not* found in the additional section, the port
3990 values from the SRV records were lost.
3992 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3993 using the correct key (the original address) when searching the retry
3994 rules in order to find which one to use for generating the retry hint.
3996 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3997 adding the default one. Similarly, if it contains a Reply-To: header, the
3998 errors_reply_to option, if set, is not used.
4000 PH/19 When calculating a retry time, Exim used to measure the "time since
4001 failure" by looking at the "first failed" field in the retry record. Now
4002 it does not use this if it is later than than the arrival time of the
4003 message. Instead it uses the arrival time. This makes for better
4004 behaviour in cases where some deliveries succeed, thus re-setting the
4005 "first failed" field. An example is a quota failure for a huge message
4006 when small messages continue to be delivered. Without this change, the
4007 "time since failure" will always be short, possible causing more frequent
4008 delivery attempts for the huge message than are intended.
4009 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4011 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4012 $1, $2, $3) because the numerical variables can be reset during some
4013 expansion items (e.g. "match"), thereby losing the authentication data.
4015 PH/21 Make -bV show the size of off_t variables so that the test suite can
4016 decide whether to run tests for quotas > 2G.
4018 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4019 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4020 filecount value is greater than 2G or if a quota value is greater than 2G
4021 on a system where the size of off_t is not greater than 4, a panic error
4024 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4025 never match. The debug and -bh output now contains an explicit error
4026 message indicating a malformed IPv4 address or mask.
4028 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4029 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4030 PH/23 above applies.
4032 PH/25 Do not write to syslog when running in the test harness. The only
4033 occasion when this arises is a failure to open the main or panic logs
4034 (for which there is an explicit test).
4036 PH/26 Added the /no_tell option to "control=freeze".
4038 PH/27 If a host name lookup failed very early in a connection, for example, if
4039 the IP address matched host_lookup and the reverse lookup yielded a name
4040 that did not have a forward lookup, an error message of the form "no IP
4041 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4042 could be logged. Now it outputs the IP address instead of "NULL".
4044 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4045 allows the sender and the authenticated sender to be set when
4046 submitting a message from within Exim. Since child_open_exim() is
4047 documented for local_scan(), the new function should be too.
4049 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4050 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4051 results in an empty string is now treated as unset.
4053 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4055 PH/31 Added sender_verify_fail logging option.
4057 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4058 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4059 tidied the source and removed it altogether.
4061 PH/33 When a queue run was abandoned because the load average was too high, a
4062 log line was always written; now it is written only if the queue_run log
4063 selector is set. In addition, the log line for abandonment now contains
4064 information about the queue run such as the pid. This is always present
4065 in "start" and "stop" lines but was omitted from the "abandon" line.
4067 PH/34 Omit spaces between a header name and the colon in the error message that
4068 is given when verify = headers_syntax fails (if there are lots of them,
4069 the message gets confusing).
4071 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4072 names, as there are now some PTR records that contain slashes. This check
4073 is only to protect against broken name servers that fall over on strange
4074 characters, so the fact that it applies to all lookups doesn't matter.
4076 PH/36 Now that the new test suite is complete, we can remove some of the
4077 special code in Exim that was needed for the old test suite. For example,
4078 sorting DNS records because real resolvers return them in an arbitrary
4079 order. The new test suite's fake resolver always returns records in the
4082 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4083 bounces) except when queue_only is set, to avoid logging races between
4084 the different processes.
4086 PH/38 Panic-die if .include specifies a non-absolute path.
4088 PH/39 A tweak to the "H" retry rule from its user.
4090 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4091 a label. They prevented compilation on older perls.
4093 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4094 a warning to be raised on newish perls.
4096 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4097 on queue. Changes to match documented behaviour of showing count of
4098 messages matching specified criteria.
4100 PH/40 Changed the default ident timeout from 30s to 5s.
4102 PH/41 Added support for the use of login_cap features, on those BSD systems
4103 that have them, for controlling the resources used by pipe deliveries.
4105 PH/42 The content-scanning code uses fopen() to create files in which to put
4106 message data. Previously it was not paying any attention to the mode of
4107 the files. Exim runs with umask(0) because the rest of the code creates
4108 files with open(), and sets the required mode explicitly. Thus, these
4109 files were ending up world-writeable. This was not a big issue, because,
4110 being within the spool directory, they were not world-accessible. I have
4111 created a function called modefopen, which takes an additional mode
4112 argument. It sets umask(777), creates the file, chmods it to the required
4113 mode, then resets the umask. All the relevant calls to fopen() in the
4114 content scanning code have been changed to use this function.
4116 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4117 to 24 hours. This avoids potential overflow problems when processing G
4118 and H retry rules. I suspect nobody ever tinkers with this value.
4120 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4122 PH/45 When the plaintext authenticator is running as a client, the server's
4123 challenges are checked to ensure they are valid base64 strings. By
4124 default, the authentication attempt is cancelled if an invalid string is
4125 received. Setting client_ignore_invalid_base64 true ignores these errors.
4126 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4127 they are received. Thus, the responses can be made to depend on the
4128 challenges. If an invalid string is ignored, an empty string is placed in
4131 PH/46 Messages that are created by the autoreply transport now contains a
4132 References: header, in accordance with RFCs 2822 and 3834.
4134 PH/47 Added authenticated_sender_force to the smtp transport.
4136 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4138 PH/49 Installed latest patch from the Sieve maintainer.
4140 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4141 was also set, the appendfile transport was unnecessarily scanning a
4142 directory of message files (e.g. for maildir delivery) to find the count
4143 of files (along with the size), even though it did not need this
4144 information. It now does the scan only if it needs to find either the
4145 size of the count of files.
4147 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4149 PH/52 Two bugs concerned with error handling when the smtp transport is
4152 (i) Exim was not creating retry information for temporary errors given
4153 for individual recipients after the DATA command when the smtp transport
4154 was used in LMTP mode. This meant that they could be retried too
4155 frequently, and not timed out correctly.
4157 (ii) Exim was setting the flag that allows error details to be returned
4158 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4159 recipients that were returned after the DATA command.
4161 PH/53 This is related to PH/52, but is more general: for any failing address,
4162 when detailed error information was permitted to be returned to the
4163 sender, but the error was temporary, then after the final timeout, only
4164 "retry timeout exceeded" was returned. Now it returns the full error as
4165 well as "retry timeout exceeded".
4167 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4168 do this, and (what is worse) MTAs that accept it.
4170 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4171 will now be deprecated.
4173 PH/56 New os.c-cygwin from the Cygwin maintainer.
4175 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4176 formats (previously only available in exim formats via -bpr, -bpru,
4177 and -bpra. Now also available in native and exiqgrep formats)
4179 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4180 with very large, slow to parse queues
4182 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4184 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4186 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4187 responses to authentication challenges, though it was showing the
4188 challenges; (ii) I've removed the CR characters from the debug output for
4191 PH/58 Allow for the insertion of a newline as well as a space when a string
4192 is turned into more than one encoded-word during RFC 2047 encoding. The
4193 Sieve code now uses this.
4195 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4196 data_4xx, lost_connection, tls_required.
4198 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4199 message was being sent as an SMTP response.
4201 PH/61 Add -l and -k options to exicyclog.
4203 PH/62 When verifying, if an address was redirected to one new address, so that
4204 verification continued, and the new address failed or deferred after
4205 having set something in $address_data, the value of $address_data was not
4206 passed back to the ACL. This was different to the case when no
4207 redirection occurred. The value is now passed back in both cases.
4209 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4210 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4211 use login_cap.h, so on its own it isn't the distinguishing feature. The
4212 new name refers directly to the setclassresources() function.
4214 PH/65 Added configuration files for NetBSD3.
4216 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4218 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4219 is preferred over IPv4.
4221 PH/68 The bounce_return_message and bounce_return_body options were not being
4222 honoured for bounces generated during the reception of non-SMTP messages.
4223 In particular, this applied to messages rejected by the ACL. This bug has
4224 been fixed. However, if bounce_return_message is true and bounce_return_
4225 body is false, the headers that are returned for a non-SMTP message
4226 include only those that have been read before the error was detected.
4227 (In the case of an ACL rejection, they have all been read.)
4229 PH/69 The HTML version of the specification is now built in a directory called
4230 spec_html instead of spec.html, because the latter looks like a path with
4231 a MIME-type, and this confuses some software.
4233 PH/70 Catch two compiler warnings in sieve.c.
4235 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4236 function verify_get_ident() calls ip_connect() to connect a socket, but
4237 if the "connect()" function timed out, ip_connect() used to close the
4238 socket. However, verify_get_ident() also closes the socket later, and in
4239 between Exim writes to the log, which may get opened at this point. When
4240 the socket was closed in ip_connect(), the log could get the same file
4241 descriptor number as the socket. This naturally causes chaos. The fix is
4242 not to close the socket in ip_connect(); the socket should be closed by
4243 the function that creates it. There was only one place in the code where
4244 this was missing, in the iplookup router, which I don't think anybody now
4245 uses, but I've fixed it anyway.
4247 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4248 well as to direct DNS lookups. Otherwise the handling of names in host
4249 lists is inconsistent and therefore confusing.
4255 PH/01 Two changes to the default runtime configuration:
4257 (1) Move the checks for relay_from_hosts and authenticated clients from
4258 after to before the (commented out) DNS black list checks.
4260 (2) Add control=submission to the relay_from_hosts and authenticated
4261 clients checks, on the grounds that messages accepted by these
4262 statements are most likely to be submissions.
4264 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4266 (1) Generate an error if the third argument for the ${prvs expansion is
4269 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4272 (3) Reset the variables that are obtained from the first argument of
4273 ${prvscheck and used in the second argument before leaving the code,
4274 because their memory is reclaimed, so using them afterwards may do
4277 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4278 one (it's much easier than Tom thought :-).
4280 (5) Because of (4), we can now allow for the use of $prvscheck_result
4281 inside the third argument.
4283 PH/03 For some reason, the default setting of PATH when running a command from
4284 a pipe transport was just "/usr/bin". I have changed it to
4287 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4288 anything to be listed in the output from -bV.
4290 PH/05 When a filter generated an autoreply, the entire To: header line was
4291 quoted in the delivery log line, like this:
4293 => >A.N.Other <ano@some.domain> <original@ddress> ...
4295 This has been changed so that it extracts the operative address. There
4296 may be more than one such address. If so, they are comma-separated, like
4299 => >ano@some.domain,ona@other.domain <original@ddress> ...
4301 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4302 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4303 not being looked up in the rDNS to get a host name, Exim was showing the
4304 IP address twice in Received: lines, even though the IP addresses were
4305 identical. For example:
4307 Received: from [1.2.3.4] (helo=[1.2.3.4])
4309 However, if the real host name was known, it was omitting the HELO data
4310 if it matched the actual IP address. This has been tidied up so that it
4311 doesn't show the same IP address twice.
4313 PH/07 When both +timestamp and +memory debugging was on, the value given by
4314 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4315 called by the debug printing, thereby overwriting the timestamp buffer.
4316 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4318 PH/08 When the original message was included in an autoreply transport, it
4319 always said "this is a copy of the message, including all the headers",
4320 even if body_only or headers_only was set. It now gives an appropriate
4323 PH/09 Applied a patch from the Sieve maintainer which:
4325 o fixes some comments
4326 o adds the (disabled) notify extension core
4327 o adds some debug output for the result of if/elsif tests
4328 o points to the current vacation draft in the documentation
4329 and documents the missing references header update
4333 o fixes a bug in processing the envelope test (when testing
4334 multiple envelope elements, the last element determined the
4337 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4338 Electronic Mail") by including:
4340 Auto-submitted: auto-generated
4342 in the messages that it generates (bounce messages and others, such as
4343 warnings). In the case of bounce messages for non-SMTP messages, there was
4344 also a typo: it was using "Auto_submitted" (underscore instead of
4345 hyphen). Since every message generated by Exim is necessarily in response
4346 to another message, thes have all been changed to:
4348 Auto-Submitted: auto-replied
4350 in accordance with these statements in the RFC:
4352 The auto-replied keyword:
4354 - SHOULD be used on messages sent in direct response to another
4355 message by an automatic process,
4357 - MUST NOT be used on manually-generated messages,
4359 - MAY be used on Delivery Status Notifications (DSNs) and Message
4360 Disposition Notifications (MDNs),
4362 - MUST NOT be used on messages generated by automatic or periodic
4363 processes, except for messages which are automatic responses to
4366 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4367 to the default Received: header definition.
4369 PH/12 Added log selector acl_warn_skipped (default on).
4371 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4372 variables because (a) they are in the wrong storage pool and (b) even if
4373 they were copied, it wouldn't work properly because of the caching.
4375 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4376 checking when decoding. Apparently there are clients that generate
4377 overlong encoded strings. Why am I not surprised?
4379 PH/15 If the first argument of "${if match_address" was not empty, but did not
4380 contain an "@" character, Exim crashed. Now it writes a panic log message
4381 and treats the condition as false.
4383 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4385 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4386 extension "envelope-auth". The code is finished and in agreement with
4387 other implementations, but there is no documentation so far and in fact,
4388 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4389 not changing the active code.
4391 Print executed "if" and "elsif" statements when debugging is used. This
4392 helps a great deal to understand what a filter does.
4394 Document more things not specified clearly in RFC3028. I had all this
4395 sorted out, when out of a sudden new issues came to my mind. Oops."
4397 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4400 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4401 canonical form (as documented). However, after a host name lookup from
4402 the IP address, check_host() was doing a simple string comparison with
4403 addresses acquired from the DNS when checking that the found name did
4404 have the original IP as one of its addresses. Since any found IPv6
4405 addresses are likely to be in abbreviated form, the comparison could
4406 fail. Luckily, there already exists a function for doing the comparison
4407 by converting both addresses to binary, so now that is used instead of
4408 the text comparison.
4410 PH/20 There was another similar case to PH/19, when a complete host name was
4411 given in a host list; looking up its IP address could give an abbreviated
4412 form, whereas the current host's name might or might not be abbreviated.
4413 The same fix has been applied.
4419 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4420 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4423 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4424 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4426 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4428 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4429 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4430 Also, for "Auto-Submitted", treat anything other than "no" as
4431 non-personal, in accordance with RFC 3834. (Previously it treated
4432 anything starting "auto-" as non-personal.)
4434 TF/01 The control=submission/name=... option had a problem with syntax
4435 errors if the name included a slash character. The /name= option
4436 now slurps the rest of the string, so it can include any characters
4437 but it must come last in the list of options (after /sender_retain
4440 PH/05 Some modifications to the interface to the fake nameserver for the new
4448 TK/01 Added the "success_on_redirect" address verification option. See
4449 NewStuff for rationale and an example.
4451 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4453 PH/02 Patch to exigrep to allow it to work on syslog lines.
4455 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4456 fread() to skip over the body file's header line, because in Cygwin the
4457 header line is locked and is inaccessible.
4459 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4460 co-exist for some time) to make it clear that it is the Exim ID that is
4461 referenced, not the Message-ID: header line.
4463 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4464 string_format() function, because snprintf() does not exist on all
4467 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4468 ${stat: expansion item.
4470 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4471 protocol synchronization error", to keep the pedants happy.
4473 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4474 well as for IRIX systems, when gcc is being used. See the host.c source
4477 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4479 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4482 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4483 transport and to the smtp transport in LMTP mode.
4485 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4487 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4488 run a filter in a subprocess. This could lead to confusion in subsequent
4489 lookups in the parent process. There should also be a search_tidyup() at
4490 the end of the subprocess.
4492 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4493 only if the host matched helo_try_verify_hosts, which caused the
4494 verification to occur when the EHLO/HELO command was issued. The ACL just
4495 tested the remembered result. Now, if a previous verification attempt has
4496 not happened, "verify = helo" does it there and then.
4498 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4500 TK/03 Fix log output including CR from clamd.
4502 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4503 exist provoked a memory error which could cause a segfault.
4505 PH/15 Installed PCRE 6.2
4507 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4509 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4510 of the problem. Specifically, suggested +O2 rather than +O1 for the
4513 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4515 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4516 record, the port was not correctly logged when the outgoing_port log
4517 selector was set (it logged the transort's default port).
4519 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4520 fallback_hosts, and "hosts" in the smtp transport.
4522 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4523 host errors such as "Connection refused".
4525 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4526 authentication with radiusclient 0.4.9:
4528 - Error returned from rc_read_config was caught wrongly
4529 - Username/password not passed on to radius server due to wrong length.
4531 The presumption is that some radiusclient API changes for 4.51/PH/17
4532 were not taken care of correctly. The code is still untested by me (my
4533 Linux distribution still has 0.3.2 of radiusclient), but it was
4534 contributed by a Radius user.
4536 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4537 expanding the "port" option of the smtp transport.
4539 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4540 while reading a MIME header. Thanks to Tom Hughes for a patch.
4542 PH/24 Include config.h inside local_scan.h so that configuration settings are
4545 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4546 that in an ACL for RCPT (for example), you can examine exactly what was
4549 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4550 commands, but it was not correctly comparing the address with the actual
4551 client host address. Thus, it would show the EHLO address in Received:
4552 header lines when this was not necessary.
4554 PH/27 Added the % operator to ${eval:}.
4556 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4557 it should be ignoring failures (because with -C, for example, it has lost
4558 privilege). It wasn't ignoring creation failures other than "already
4561 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4564 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4565 because an input line was too long, either on its own, or by virtue of
4566 too many continuations, the temporary file was not being removed, and the
4567 return code was incorrect.
4569 PH/31 Missing "BOOL" in function definition in filtertest.c.
4571 PH/32 Applied Sieve patches from the maintainer.
4573 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4575 PH/33 Added "verify = not_blind".
4577 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4578 Local/Makefile (with some defaults set). These are used in built scripts
4579 such as exicyclog, but they have never been used in the exim_install
4580 script (though there are many overriding facilities there). I have
4581 arranged that the exim_install script now takes note of these two
4584 PH/35 Installed configuration files for Dragonfly.
4586 PH/36 When a locally submitted message by a trusted user did not contain a
4587 From: header, and the sender address was obtained from -f or from an SMTP
4588 MAIL command, and the trusted user did not use -F to supply a sender
4589 name, $originator_name was incorrectly used when constructing a From:
4590 header. Furthermore, $originator_name was used for submission mode
4591 messages from external hosts without From: headers in a similar way,
4592 which is clearly wrong.
4594 PH/37 Added control=suppress_local_fixups.
4596 PH/38 When log_selector = +received_sender was set, and the addition of the
4597 sender made the log line's construction buffer exactly full, or one byte
4598 less than full, an overflow happened when the terminating "\n" was
4601 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4602 when the result of a list match is failure because a DNS lookup failed.
4604 PH/40 RM_COMMAND is now used in the building process.
4606 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4607 the "build-* directories that it finds.
4609 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4610 address in a domain literal was a prefix of an interface address.
4612 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4613 when verifying a sender address, unless rewrite_headers is false.
4615 PH/44 Wrote a long comment about why errors_to addresses are verified as
4616 recipients, not senders.
4618 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4619 the ratelimit ACL was added.
4621 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4623 PH/46 Added extra information about PostgreSQL errors to the error string.
4625 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4626 avoiding the need to install special zones in a real server. This is
4627 backwards compatible; if it can't find the fake resolver, it drops back.
4628 Thus, both old and new test suites can be run.
4630 TF/02 Added util/ratelimit.pl
4632 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4633 clock is set back in time.
4635 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4636 Candler <B.Candler@pobox.com>.
4638 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4639 OK for addresses that are the result of redirections.
4641 PH/48 A number of further additions for the benefit of the new test suite,
4642 including a fake gethostbyname() that interfaces to the fake DNS resolver
4645 TF/06 The fix for widen_domains has also been applied to qualify_single and
4646 search_parents which are the other dnslookup options that can cause
4649 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4652 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4654 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4655 service" if it encountered an MX record with an empty target hostname.
4656 The message is now "an MX or SRV record indicated no SMTP service".
4658 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4659 if the DNS of the sending site is misconfigured. This is quite a
4660 common situation. This change restores the behaviour of treating a
4661 helo verification defer as a failure.
4663 PH/51 If self=fail was set on a router, the bounce message did not include the
4664 actual error message.
4670 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4672 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4673 command itself ended in error, the underlying message about the transport
4674 filter timeout was being overwritten with the pipe command error. Now the
4675 underlying error message should be appended to the second error message.
4677 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4679 PH/02 Reduce the amount of output that "make" produces by default. Full output
4680 can still be requested.
4682 PH/03 The warning log line about a condition test deferring for a "warn" verb
4683 was being output only once per connection, rather than after each
4684 occurrence (because it was using the same function as for successful
4685 "warn" verbs). This seems wrong, so I have changed it.
4687 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4688 it should not have, which might have caused a crash in the right
4689 circumstances, but probably never did.
4691 PH/04 Installed a modified version of Tony Finch's patch to make submission
4692 mode fix the return path as well as the Sender: header line, and to
4693 add a /name= option so that you can make the user's friendly name appear
4696 TF/03 Added the control = fakedefer ACL modifier.
4698 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4699 Mark Lowes for thorough testing.
4701 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4703 TK/03 Merged latest SRS patch from Miles Wilton.
4705 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4706 with the definition in sysexits.h (which is #included earlier).
4707 Fortunately, Exim does not actually use EX_OK. The code used to try to
4708 preserve the sysexits.h value, by assuming that macro definitions were
4709 scanned for macro replacements. I have been disabused of this notion,
4710 so now the code just undefines EX_OK before #including unistd.h.
4712 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4713 in the smtp transport. When a block could not be written in a single
4714 write() function, the timeout was being re-applied to each part-write.
4715 This seems wrong - if the receiver was accepting one byte at a time it
4716 would take for ever. The timeout is now adjusted when this happens. It
4717 doesn't have to be particularly precise.
4719 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4720 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4722 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4723 without a domain if the check to <postmaster@domain> fails.
4725 SC/01 Eximstats: added -xls and the ability to specify output files
4726 (patch written by Frank Heydlauf).
4728 SC/02 Eximstats: use FileHandles for outputting results.
4730 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4732 SC/04 Eximstats: fixed display of large numbers with -nvr option
4734 SC/05 Eximstats: fixed merging of reports with empty tables.
4736 SC/06 Eximstats: added the -include_original_destination flag
4738 SC/07 Eximstats: removed tabs and trailing whitespace.
4740 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4742 TK/06 MBOX spool code: Add real "From " MBOX separator line
4743 so the .eml file is really in mbox format (even though
4744 most programs do not really care). Patch from Alex Miller.
4746 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4747 The latter is generated from $received_to and is only set if the
4748 message has one envelope recipient. SA can use these headers,
4749 obviously out-of-the-box. Patch from Alex Miller.
4751 PH/08 The ${def test on a variable was returning false if the variable's
4752 value was "0", contrary to what the specification has always said!
4753 The result should be true unless the variable is empty.
4755 PH/09 The syntax error of a character other than { following "${if
4756 def:variable_name" (after optional whitespace) was not being diagnosed.
4757 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4758 accidental colon was present, for example, could give incorrect results.
4760 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4761 result is used (not including appendfile, where other changes are about
4764 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4765 This involved changing a lot of size variables from int to off_t. It
4766 should work with maildirs and everything.
4768 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4769 spamd dying while we are connected to it.
4771 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4774 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4775 function for BDB 4.3.
4777 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4779 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4780 the sources for PCRE have been re-arranged and more files are now
4783 PH/15 The code I had for printing potentially long long variables in PH/11
4784 above was not the best (it lost precision). The length of off_t variables
4785 is now inspected at build time, and an appropriate printing format (%ld
4786 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4787 to be "long long int" or "long int". This is needed for the internal
4788 formatting function string_vformat().
4790 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4791 the configuration file to be ":syslog", then the script "guesses" where
4792 the logs files are, rather than using the compiled in default. In our
4793 case the guess is not the same as the compiled default, so the script
4794 suddenly stopped working when I started to use syslog. The patch checks
4795 to see if log_file_path is "". If so, it attempts to read it from exim
4796 with no configuration file to get the compiled in version, before it
4797 falls back to the previous guessing code."
4799 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4800 implementing BATV in an Exim configuration. See NewStuff for the gory
4803 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4804 Makefile that are specific to HP-UX.
4806 PH/18 If the "use_postmaster" option was set for a recipient callout together
4807 with the "random" option, the postmaster address was used as the MAIL
4808 FROM address for the random test, but not for the subsequent recipient
4809 test. It is now used for both.
4811 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4812 patch removes a few documentation additions to RFC 3028, because the
4813 latest draft now contains them. It adds the new en;ascii-case comparator
4814 and a new error check for 8bit text in MIME parts. Comparator and
4815 require names are now matched exactly. I enabled the subaddress
4816 extension, but it is not well tested yet (read: it works for me)."
4818 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4819 rework some of the code of TK/09 above to avoid the hardwired use of
4820 "%lld" and "long long". Replaced the call to snprintf() with a call to
4823 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4824 records point to non-existent hosts", "retry timeout exceeded", and
4825 "retry time not reached for any host after a long failure period".
4827 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4828 experimental DomainKeys support:
4830 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4831 (2) On an error such as an illegally used "control", the wrong name for
4832 the control was given.
4834 These problems did NOT occur unless DomainKeys support was compiled.
4836 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4838 PH/24 Added ${if match_ip condition.
4840 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4841 databases so that it will be absolutely obvious if a crash occurs in the
4842 DB library. This is a regular occurrence (often caused by mis-matched
4845 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4846 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4847 picked up on a user's system that detects such things. There doesn't seem
4848 to be a gcc warning option for this - only an attribute that has to be
4849 put on the function's prototype. It seems that in Fedora Core 4 they have
4850 set this on a number of new functions. No doubt there will be more in due
4853 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4854 specify a transport. However, if an address that was verified by such a
4855 router was the subject of a callout, Exim crashed because it tried to
4856 read the rcpt_include_affixes from the non-existent transport. Now it
4857 just assumes that the setting of that option is false. This bug was
4858 introduced by 4.51/PH/31.
4860 PH/28 Changed -d+all to exclude +memory, because that information is very
4861 rarely of interest, but it makes the output a lot bigger. People tend to
4862 do -d+all out of habit.
4864 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4865 code in os-type was giving problems when libc.so lives in lib64, like on
4868 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4869 aren't the modern standard, and it seems that some systems' include files
4870 don't always have them. Exim was already checking for some of the newer
4871 ones like T_AAAA, and defining it itself. I've added checks for all the
4872 record types that Exim uses.
4874 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4875 not automatically generating a new one, as it is supposed to. This
4876 prevented TLS from working. If the file did exist, but contained invalid
4877 data, a new version was generated, as expected. It was only the case of a
4878 non-existent file that was broken.
4880 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4881 with a change in libdomainkeys > 0.64.
4883 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4884 from DNS. If the selector record carries the flag, it now has
4885 precedence over the domain-wide flag.
4887 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4889 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4890 the use of a port name that isn't defined in /etc/services) occurred, the
4891 message was deferred as in a normal delivery, and thus remained on the
4892 spool, instead of being failed because of the mua_wrapper setting. This
4893 is now fixed, and I tidied up some of the mua_wrapper messages at the
4896 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4897 the messages in a hash of arrays rather than using individual hashes.
4898 This is a bit cleaner and results in dramatic memory savings, albeit
4899 at a slight CPU cost.
4901 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4902 as requested by Marc Sherman.
4904 SC/10 Eximstats: added histograms for user specified patterns as requested
4907 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4909 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4910 fopen() in the content-scanning modules that did not already have it.
4916 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4917 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4919 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4921 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4923 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4924 to test Sieve filters that use "vacation".
4926 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4927 that changes the way the GnuTLS parameters are stored in the cache file.
4928 The new format can be generated externally. For backward compatibility,
4929 if the data in the cache doesn't make sense, Exim assumes it has read an
4930 old-format file, and it generates new data and writes a new file. This
4931 means that you can't go back to an older release without removing the
4934 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4935 work if there are any delivery delays because "one_time" forces the
4936 parent to be marked "delivered", so its unseen clone is never tried
4937 again. For this reason, Exim now forbids the simultaneous setting of
4940 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4941 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4942 entry if you want to know the details. The fix, however, neglected to
4943 consider the case when local delivery batching is involved. The test for
4944 "previously delivered" was not happening when checking to see if an
4945 address could be batched with a previous (undelivered) one; under
4946 certain circumstances this could lead to multiple deliveries to the same
4949 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4950 in its include files, and this causes problems building Exim.
4952 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4953 header_syntax) but Exim was just ignoring anything given after a slash.
4954 In particular, this caused confusion with an attempt to use "verify =
4955 reverse_host_lookup/defer_ok". An error is now given when options are
4956 supplied for verify items that do not have them. (Maybe reverse_host_
4957 lookup should have a defer_ok option, but that's a different point.)
4959 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4960 defined by RFC 821) to 2048, because there were problems with some AUTH
4961 commands, and RFC 1869 says the size should be increased for extended
4962 SMTP commands that take arguments.
4964 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4967 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4968 "unknown" error; now it says that the functionality isn't in the binary.
4970 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4971 an address' error message when a string expansion fails (syntax or
4972 whatever). Otherwise the password may appear in the log. Following change
4973 PH/42 below, there is no longer a chance of it appearing in a bounce
4976 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4978 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4979 of its addresses was ever tried. (Bugzilla bug #2.)
4981 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4982 the result incorrectly in the debug output. (It correctly added a newline
4983 to what was transported.)
4985 TF/01 Added $received_time.
4987 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4988 commented out examples of how to interface to a virus scanner and to
4989 SpamAssassin. Also added commented examples of av_scanner and
4990 spamd_address settings.
4992 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4993 and controls are allowed in which ACLs. There were a couple of minor
4994 errors. Some of the entries in the conditions table (which is a table of
4995 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4996 negation of where the condition IS allowed.
4998 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5000 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5001 header file does not have a version number, so I've had to invent a new
5002 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5003 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5004 radiusclient), but it was contributed by a Radius user.
5006 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5007 files or directories, for OpenSSL.
5009 PH/19 When an Exim process that is running as root has to create an Exim log
5010 file, it does so in a subprocess that runs as exim:exim so as to get the
5011 ownership right at creation (otherwise, other Exim processes might see
5012 the file with the wrong ownership). There was no test for failure of this
5013 fork() call, which would lead to the process getting stuck as it waited
5014 for a non-existent subprocess. Forks do occasionally fail when resources
5015 run out. I reviewed all the other calls to fork(); they all seem to check
5018 PH/20 When checking for unexpected SMTP input at connect time (before writing
5019 the banner), Exim was not dealing correctly with a non-positive return
5020 from the read() function. If the client had disconnected by this time,
5021 the result was a log entry for a synchronization error with an empty
5022 string after "input=" when read() returned zero. If read() returned -1
5023 (an event I could not check), uninitialized data bytes were printed.
5024 There were reports of junk text (parts of files, etc) appearing after
5027 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5029 PH/22 Added support for macro redefinition, and (re)definition in between
5030 driver and ACL definitions.
5032 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5033 forgetting to use the resulting value; it was using the unexpanded value.
5035 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5036 hadn't been configured. The fix is from Juergen Kreileder, who
5037 understands it better than I do:
5039 "Here's what I see happening with three configured cyrus_sasl
5040 authenticators configured (plain, login, cram-md5):
5042 On startup auth_cyrus_sasl_init() gets called for each of these.
5043 This means three calls to sasl_listmech() without a specified mech_list.
5044 => SASL tests which mechs of all available mechs actually work
5045 => three warnings about OTP not working
5046 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5048 With the patch, sasl_listmech() also gets called three times. But now
5049 SASL's mech_list option is set to the server_mech specified in the the
5050 authenticator. Or in other words, the answer from sasl_listmech()
5051 gets limited to just the mech you're testing for (which is different
5053 => the return list contains just 'plain' or 'login', 'cram-md5' or
5054 nothing depending on the value of ob->server_mech.
5056 I've just tested the patch: Authentication still works fine,
5057 unavailable mechs specified in the exim configuration are still
5058 caught, and the auth.log warnings about OTP are gone."
5060 PH/25 When debugging is enabled, the contents of the command line are added
5061 to the debugging output, even when log_selector=+arguments is not
5064 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5065 answer is "GNU", and only if the return is "GNU/something" is the answer
5068 PH/27 $acl_verify_message is now set immediately after the failure of a
5069 verification in an ACL, and so is available in subsequent modifiers. In
5070 particular, the message can be preserved by coding like this:
5072 warn !verify = sender
5073 set acl_m0 = $acl_verify_message
5075 Previously, $acl_verify_message was set only while expanding "message"
5076 and "log_message" when a very denied access.
5078 PH/28 Modified OS/os.c-Linux with
5080 -#ifndef OS_LOAD_AVERAGE
5081 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5083 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5084 nomenclature these days.)
5086 PH/29 Installed patch from the Sieve maintainer that adds the options
5087 sieve_useraddress and sieve_subaddress to the redirect router.
5089 PH/30 In these circumstances:
5090 . Two addresses routed to the same list of hosts;
5091 . First host does not offer TLS;
5092 . First host accepts first address;
5093 . First host gives temporary error to second address;
5094 . Second host offers TLS and a TLS session is established;
5095 . Second host accepts second address.
5096 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5097 and peerdn, if requested) that were in fact used only for the second
5100 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5101 attention to any local part prefix or suffix that was matched by the
5102 router that accepted the address. It now behaves in the same way as it
5103 does for delivery: the affixes are removed from the local part unless
5104 rcpt_include_affixes is set on the transport.
5106 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5107 timeout during the DATA phase of an incoming message.
5109 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5110 applied a patch sanctioned by the Sieve maintainer.
5112 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5113 the uid or gid is negative. A case of a negative gid caused this to be
5114 noticed. The fix allows for either to be negative.
5116 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5117 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5120 PH/36 The change PH/12 above was broken. Fixed it.
5122 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5123 the grounds that routing the same address twice would always produce the
5124 same answer. This might have been true once, but it is certainly no
5125 longer true now. Routing a child address may depend on the previous
5126 routing that produced that child. Some complicated redirection strategies
5127 went wrong when messages had multiple recipients, and made Exim's
5128 behaviour dependent on the order in which the addresses were given.
5130 I have moved the duplicate checking until after the routing is complete.
5131 Exim scans the addresses that are assigned to local and remote
5132 transports, and removes any duplicates. This means that more work will be
5133 done, as duplicates will always all be routed, but duplicates are
5134 presumably rare, so I don't expect this is of any significance.
5136 For deliveries to pipes, files, and autoreplies, the duplicate checking
5137 still happens during the routing process, since they are not going to be
5140 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5141 It corrects a timeout issue with spamd. This is Ian's comment: "The
5142 background is that sometimes spamd either never reads data from a
5143 connection it has accepted, or it never writes response data. The exiscan
5144 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5145 blindly assumes that writes won't block so it may never time out."
5147 PH/39 Allow G after quota size as well as K and M.
5149 PH/40 The value set for $authenticated_id in an authenticator may not contain
5150 binary zeroes or newlines because the value is written to log lines and
5151 to spool files. There was no check on this. Now the value is run through
5152 the string_printing() function so that such characters are converted to
5153 printable escape sequences.
5155 PH/41 $message_linecount is a new variable that contains the total number of
5156 lines in the message. Compare $body_linecount, which is the count for the
5159 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5160 bounce and delay warning messages, except in certain special cases, which
5163 (a) An SMTP error message from a remote host;
5164 (b) A message specified in a :fail: redirection;
5165 (c) A message specified in a "fail" command in a system filter;
5166 (d) A message specified in a FAIL return from the queryprogram router;
5167 (e) A message specified by the cannot_route_message router option.
5169 In these cases only, Exim does include the error details in bounce and
5170 warning messages. There are also a few cases where bland messages such
5171 as "unrouteable address" or "local delivery error" are given.
5173 PH/43 $value is now also set for the "else" part of a ${run expansion.
5175 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5176 being worked on, but at least Exim now implements the latest version to
5179 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5180 process to complete was treated as a delivery failure, a timeout while
5181 writing the message to the pipe was logged, but erroneously treated as a
5182 successful delivery. Such timeouts include transport filter timeouts. For
5183 consistency with the overall process timeout, these timeouts are now
5184 treated as errors, giving rise to delivery failures by default. However,
5185 there is now a new Boolean option for the pipe transport called
5186 timeout_defer, which, if set TRUE, converts the failures into defers for
5187 both kinds of timeout. A transport filter timeout is now identified in
5190 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5191 systems where "make" and "gmake" are different, calling "gmake" at top
5192 level broke things. I've arranged for the value of $(MAKE) to be passed
5193 from the Makefile to this script so that it can call the same version of
5197 A note about Exim versions 4.44 and 4.50
5198 ----------------------------------------
5200 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5201 changes of various kinds. As a consequence, a big documentation update was
5202 needed. This delayed the release for rather longer than seemed good, especially
5203 in the light of a couple of (minor) security issues. Therefore, the changes
5204 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5205 release. So 4.44 and 4.50 are in effect two different branches that both start
5208 I have left the 4.50 change log unchanged; it contains all the changes since
5209 4.43. The change log for 4.44 is below; many of its items are identical to
5210 those for 4.50. This seems to be the most sensible way to preserve the
5211 historical information.
5217 1. Minor wording change to the doc/README.SIEVE file.
5219 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5220 computation of the current number of files was incorrect.
5222 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5223 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5226 4. Give more explanation in the error message when the command for a transport
5227 filter fails to execute.
5229 5. There are several places where Exim runs a non-Exim command in a
5230 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5231 was being done only for the command run by the queryprogram router. It is
5232 now done for all such subprocesses. The other cases are: ${run, transport
5233 filters, and the commands run by the lmtp and pipe transports.
5235 6. Added CONFIGURE_GROUP build-time option.
5237 7. Some older OS have a limit of 256 on the maximum number of file
5238 descriptors. Exim was using setrlimit() to set 1000 as a large value
5239 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5240 systems. I've change it so that if it can't get 1000, it tries for 256.
5242 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5243 was an oversight, and furthermore, ever since the addition of extra
5244 controls (e.g. 4.43/32), the checks on when to allow different forms of
5245 "control" were broken. There should now be diagnostics for all cases when a
5246 control that does not make sense is encountered.
5248 9. Added the /retain_sender option to "control=submission".
5250 10. $recipients is now available in the predata ACL (oversight).
5252 11. Tidy the search cache before the fork to do a delivery from a message
5253 received from the command line. Otherwise the child will trigger a lookup
5254 failure and thereby defer the delivery if it tries to use (for example) a
5255 cached ldap connection that the parent has called unbind on.
5257 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5258 of $address_data from the recipient verification was clobbered by the
5259 sender verification.
5261 13. The value of address_data from a sender verification is now available in
5262 $sender_address_data in subsequent conditions in the ACL statement.
5264 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5266 15. Added a new option "connect=<time>" to callout options, to set a different
5269 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5270 was its contents. (It was OK if the option was not defined at all.)
5272 17. A "Completed" log line is now written for messages that are removed from
5273 the spool by the -Mrm option.
5275 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5276 information about exactly what failed.
5278 19. Added -dd to debug only the daemon process.
5280 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5281 handles timeouts, both on the server side and network timeouts. Renamed the
5282 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5284 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5285 It is now set to "smtps".
5287 22. $host_address is now set to the target address during the checking of
5288 ignore_target_hosts.
5290 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5291 being passed; this would have caused $sender_host_name to have been used if
5292 matching the list had actually called for a host name (not very likely,
5293 since this list is usually IP addresses). A host name is now passed as
5296 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5297 code that specifies a non-restarting handler (typically sigaction() in
5298 modern systems) in an attempt to fix a rare and obscure crash bug.
5300 25. Narrowed the window for a race in the daemon that could cause it to ignore
5301 SIGCHLD signals. This is not a major problem, because they are used only to
5302 wake it up if nothing else does.
5304 26. A malformed maildirsize file could cause Exim to calculate negative values
5305 for the mailbox size or file count. Odd effects could occur as a result.
5306 The maildirsize information is now recalculated if the size or filecount
5309 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5310 support for a long time. Removed HAVE_SYS_VFS_H.
5312 28. Installed the latest version of exipick from John Jetmore.
5314 29. In an address list, if the pattern was not a regular expression, an empty
5315 subject address (from a bounce message) matched only if the pattern was an
5316 empty string. Non-empty patterns were not even tested. This was the wrong
5317 because it is perfectly reasonable to use an empty address as part of a
5318 database query. An empty address is now tested by patterns that are
5319 lookups. However, all the other forms of pattern expect the subject to
5320 contain a local part and a domain, and therefore, for them, an empty
5321 address still always fails if the pattern is not itself empty.
5323 30. Exim went into a mad DNS loop when attempting to do a callout where the
5324 host was specified on an smtp transport, and looking it up yielded more
5325 than one IP address.
5327 31. Re-factored the code for checking spool and log partition space into a
5328 function that finds that data and another that does the check. The former
5329 is then used to implement four new variables: $spool_space, $log_space,
5330 $spool_inodes, and $log_inodes.
5332 32. The RFC2047 encoding function was originally intended for short strings
5333 such as real names; it was not keeping to the 75-character limit for
5334 encoded words that the RFC imposes. It now respects the limit, and
5335 generates multiple encoded words if necessary. To be on the safe side, I
5336 have increased the buffer size for the ${rfc2047: expansion operator from
5339 33. It is now permitted to omit both strings after an "if" condition; if the
5340 condition is true, the result is "true". As before, when the second string
5341 is omitted, a false condition yields an empty string. This makes it less
5342 cumbersome to write custom ACL and router conditions.
5344 34. Failure to deliver a bounce message always caused it to be frozen, even if
5345 there was an errors_to setting on the router. The errors_to setting is now
5348 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5349 canonical form (fully expanded) before being placed in
5350 $sender_host_address.
5352 36. The table in the code that translates DNS record types into text (T_A to
5353 "A" for instance) was missing entries for NS and CNAME. It is just possible
5354 that this could have caused confusion if both these types were looked up
5355 for the same domain, because the text type is used as part of Exim's
5356 per-process caching. But the chance of anyone hitting this buglet seems
5359 37. The dnsdb lookup has been extended in a number of ways.
5361 (1) There is a new type, "zns", which walks up the domain tree until it
5362 finds some nameserver records. It should be used with care.
5364 (2) There is a new type, "mxh", which is like "mx" except that it returns
5365 just the host names, not the priorities.
5367 (3) It is now possible to give a list of domains (or IP addresses) to be
5368 looked up. The behaviour when one of the lookups defers can be
5369 controlled by a keyword.
5371 (4) It is now possible to specify the separator character for use when
5372 multiple records are returned.
5374 38. The dnslists ACL condition has been extended: it is now possible to supply
5375 a list of IP addresses and/or domains to be looked up in a particular DNS
5378 39. Added log_selector=+queue_time_overall.
5380 40. When running the queue in the test harness, wait just a tad after forking a
5381 delivery process, to get repeatability of debugging output.
5383 41. Include certificate and key file names in error message when GnuTLS fails
5384 to set them up, because the GnuTLS error message doesn't include the name
5385 of the failing file when there is a problem reading it.
5387 42. Allow both -bf and -bF in the same test run.
5389 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5391 44. The "Exiscan patch" is now merged into the mainline Exim source.
5393 45. Sometimes the final signoff response after QUIT could fail to get
5394 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5395 < 0 before doing a fflush(). This bug looks as though it goes back to the
5396 introduction of TLS in release 3.20, but "sometimes" must have been rare
5397 because the tests only now provoked it.
5399 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5400 (this can affect the format of dates).
5402 47. exim_tidydb, when checking for the continued existence of a message for
5403 which it has found a message-specific retry record, was not finding
5404 messages that were in split spool directories. Consequently, it was
5405 deleting retry records that should have stayed in existence.
5407 48. Steve fixed some bugs in eximstats.
5409 49. The SPA authentication driver was not abandoning authentication and moving
5410 on to the next authenticator when an expansion was forced to fail,
5411 contradicting the general specification for all authenticators. Instead it
5412 was generating a temporary error. It now behaves as specified.
5414 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5415 (the order specifies the preference for clients). The order is now AES256,
5416 AES128, 3DES, ARCFOUR128.
5418 51. Small patch to Sieve code - explicitly set From: when generating an
5421 52. Exim crashed if a remote delivery caused a very long error message to be
5422 recorded - for instance if somebody sent an entire SpamAssassin report back
5423 as a large number of 550 error lines. This bug was coincidentally fixed by
5424 increasing the size of one of Exim's internal buffers (big_buffer) that
5425 happened as part of the Exiscan merge. However, to be on the safe side, I
5426 have made the code more robust (and fixed the comments that describe what
5429 53. Now that there can be additional text after "Completed" in log lines (if
5430 the queue_time_overall log selector is set), a one-byte patch to exigrep
5431 was needed to allow it to recognize "Completed" as not the last thing in
5434 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5435 patch that reportedly fixes this has been added. I am not expert enough to
5436 create a test for it. This is what the patch creator wrote:
5438 "I found a little strange behaviour of ldap code when working with
5439 Windows 2003 AD Domain, where users was placed in more than one
5440 Organization Units. When I tried to give exim partial DN, the exit code
5441 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5442 But simultaneously result of request was absolutely normal ldap result,
5443 so I produce this patch..."
5445 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5446 REFERENCE, so I have modified the code to exclude the patch when that macro
5449 55. Some experimental protocols are using DNS PTR records for new purposes. The
5450 keys for these records are domain names, not reversed IP addresses. The
5451 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5452 leaves it alone. Component reversal etc. now happens only for IP addresses.
5455 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5457 57. Double the size of the debug message buffer (to 2048) so that more of very
5458 long debug lines gets shown.
5460 58. The exicyclog utility now does better if the number of log files to keep
5461 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5463 59. Two changes related to the smtp_active_hostname option:
5465 (1) $smtp_active_hostname is now available as a variable.
5466 (2) The default for smtp_banner uses $smtp_active_hostname instead
5467 of $primary_hostname.
5469 60. The host_aton() function is supposed to be passed a string that is known
5470 to be a valid IP address. However, in the case of IPv6 addresses, it was
5471 not checking this. This is a hostage to fortune. Exim now panics and dies
5472 if the condition is not met. A case was found where this could be provoked
5473 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5474 components; fortuitously, this particular loophole had already been fixed
5475 by change 4.50/55 above.
5477 If there are any other similar loopholes, the new check in host_aton()
5478 itself should stop them being exploited. The report I received stated that
5479 data on the command line could provoke the exploit when Exim was running as
5480 exim, but did not say which command line option was involved. All I could
5481 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5482 running as the user.
5485 61. There was a buffer overflow vulnerability in the SPA authentication code
5486 (which came originally from the Samba project). I have added a test to the
5487 spa_base64_to_bits() function which I hope fixes it.
5490 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5491 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5493 63. The daemon start-up calls getloadavg() while still root for those OS that
5494 need the first call to be done as root, but it missed one case: when
5495 deliver_queue_load_max is set with deliver_drop_privilege. This is
5496 necessary for the benefit of the queue runner, because there is no re-exec
5497 when deliver_drop_privilege is set.
5499 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5500 This has been fixed.
5502 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5503 was in use, was not putting the data itself into the right store pool;
5504 consequently, it could be overwritten for a subsequent message in the same
5505 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5508 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5510 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5511 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5512 the function were treating the return as a boolean value, which happened to
5513 work because 0=false and not-0=true, but is not correct code.
5515 68. The host_aton() function was not handling scoped IPv6 addresses (those
5516 with, for example, "%eth0" on the end) correctly.
5518 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5519 negated items (that is, ~something) in unsigned ints. Some compilers
5520 apparently mutter when there is no cast.
5522 70. If an address verification called from an ACL failed, and did not produce a
5523 user-specific message (i.e. there was only a "system" message), nothing was
5524 put in $acl_verify_message. In this situation, it now puts the system
5527 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5528 session; change 4.31/43 added the unwanted input to the log line - except
5529 that it did not do this in the start of session case. It now does.
5531 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5532 This is wrong and can cause the other end to generate a synchronization
5533 error if it is another Exim or anything else that does the synchronization
5534 check. A QUIT command is no longer sent after a timeout.
5536 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5537 during host lookups.
5539 74. The defer_ok option of callout verification was not working if it was used
5540 when verifying addresses in header lines, that is, for this case:
5542 verify = header_sender/callout=defer_ok
5544 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5545 those file descriptors could be used for SMTP connections. If anything
5546 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5547 could be sent to the SMTP client, causing chaos. The daemon now opens
5548 stdin, stdout, and stderr to /dev/null when it puts itself into the
5551 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5552 log by default. The user can override this with suitable Perl magic.
5554 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5555 the log message when discard triggers, was not working for the DATA ACL or
5556 for the non-SMTP ACL.
5558 78. Error message wording change in sieve.c.
5560 79. If smtp_accept_max_per_host was set, the number of connections could be
5561 restricted to fewer than expected, because the daemon was trying to set up
5562 a new connection before checking whether the processes handling previous
5563 connections had finished. The check for completed processes is now done
5564 earlier. On busy systems, this bug wouldn't be noticed because something
5565 else would have woken the daemon, and it would have reaped the completed
5568 80. If a message was submitted locally by a user whose login name contained one
5569 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5570 It caused a spool format error. I have fixed the spool reading code. A
5571 related problem was that the "from" clause in the Received: line became
5572 illegal because of the space(s). It is now covered by ${quote_local_part.
5574 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5577 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5579 83. Patch from Sieve maintainer to fix unterminated string problem in
5580 "vacation" handling.
5582 84. Some minor changes to the Linux configuration files to help with other
5583 OS variants using glibc.
5585 85. One more patch for Sieve to update vacation handling to latest spec.
5588 ----------------------------------------------------
5589 See the note above about the 4.44 and 4.50 releases.
5590 ----------------------------------------------------
5596 1. Change 4.43/35 introduced a bug that caused file counts to be
5597 incorrectly computed when quota_filecount was set in an appendfile
5600 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5601 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5604 3. Give more explanation in the error message when the command for a transport
5605 filter fails to execute.
5607 4. There are several places where Exim runs a non-Exim command in a
5608 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5609 was being done only for the command run by the queryprogram router. It is
5610 now done for all such subprocesses. The other cases are: ${run, transport
5611 filters, and the commands run by the lmtp and pipe transports.
5613 5. Some older OS have a limit of 256 on the maximum number of file
5614 descriptors. Exim was using setrlimit() to set 1000 as a large value
5615 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5616 systems. I've change it so that if it can't get 1000, it tries for 256.
5618 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5619 was an oversight, and furthermore, ever since the addition of extra
5620 controls (e.g. 4.43/32), the checks on when to allow different forms of
5621 "control" were broken. There should now be diagnostics for all cases when a
5622 control that does not make sense is encountered.
5624 7. $recipients is now available in the predata ACL (oversight).
5626 8. Tidy the search cache before the fork to do a delivery from a message
5627 received from the command line. Otherwise the child will trigger a lookup
5628 failure and thereby defer the delivery if it tries to use (for example) a
5629 cached ldap connection that the parent has called unbind on.
5631 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5632 of $address_data from the recipient verification was clobbered by the
5633 sender verification.
5635 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5636 was its contents. (It was OK if the option was not defined at all.)
5638 11. A "Completed" log line is now written for messages that are removed from
5639 the spool by the -Mrm option.
5641 12. $host_address is now set to the target address during the checking of
5642 ignore_target_hosts.
5644 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5645 being passed; this would have caused $sender_host_name to have been used if
5646 matching the list had actually called for a host name (not very likely,
5647 since this list is usually IP addresses). A host name is now passed as
5650 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5651 code that specifies a non-restarting handler (typically sigaction() in
5652 modern systems) in an attempt to fix a rare and obscure crash bug.
5654 15. Narrowed the window for a race in the daemon that could cause it to ignore
5655 SIGCHLD signals. This is not a major problem, because they are used only to
5656 wake it up if nothing else does.
5658 16. A malformed maildirsize file could cause Exim to calculate negative values
5659 for the mailbox size or file count. Odd effects could occur as a result.
5660 The maildirsize information is now recalculated if the size or filecount
5663 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5664 support for a long time. Removed HAVE_SYS_VFS_H.
5666 18. Updated exipick to current release from John Jetmore.
5668 19. Allow an empty sender to be matched against a lookup in an address list.
5669 Previously the only cases considered were a regular expression, or an
5672 20. Exim went into a mad DNS lookup loop when doing a callout where the
5673 host was specified on the transport, if the DNS lookup yielded more than
5676 21. The RFC2047 encoding function was originally intended for short strings
5677 such as real names; it was not keeping to the 75-character limit for
5678 encoded words that the RFC imposes. It now respects the limit, and
5679 generates multiple encoded words if necessary. To be on the safe side, I
5680 have increased the buffer size for the ${rfc2047: expansion operator from
5683 22. Failure to deliver a bounce message always caused it to be frozen, even if
5684 there was an errors_to setting on the router. The errors_to setting is now
5687 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5688 canonical form (fully expanded) before being placed in
5689 $sender_host_address.
5691 24. Updated eximstats to version 1.33
5693 25. Include certificate and key file names in error message when GnuTLS fails
5694 to set them up, because the GnuTLS error message doesn't include the name
5695 of the failing file when there is a problem reading it.
5697 26. Expand error message when OpenSSL has problems setting up cert/key files.
5700 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5701 (this can affect the format of dates).
5703 28. exim_tidydb, when checking for the continued existence of a message for
5704 which it has found a message-specific retry record, was not finding
5705 messages that were in split spool directories. Consequently, it was
5706 deleting retry records that should have stayed in existence.
5708 29. eximstats updated to version 1.35
5709 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5710 1.35 - bugfix such that pie charts by volume are generated correctly
5712 30. The SPA authentication driver was not abandoning authentication and moving
5713 on to the next authenticator when an expansion was forced to fail,
5714 contradicting the general specification for all authenticators. Instead it
5715 was generating a temporary error. It now behaves as specified.
5717 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5718 (the order specifies the preference for clients). The order is now AES256,
5719 AES128, 3DES, ARCFOUR128.
5721 31. Small patch to Sieve code - explicitly set From: when generating an
5724 32. Exim crashed if a remote delivery caused a very long error message to be
5725 recorded - for instance if somebody sent an entire SpamAssassin report back
5726 as a large number of 550 error lines. This bug was coincidentally fixed by
5727 increasing the size of one of Exim's internal buffers (big_buffer) that
5728 happened as part of the Exiscan merge. However, to be on the safe side, I
5729 have made the code more robust (and fixed the comments that describe what
5732 33. Some experimental protocols are using DNS PTR records for new purposes. The
5733 keys for these records are domain names, not reversed IP addresses. The
5734 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5735 leaves it alone. Component reversal etc. now happens only for IP addresses.
5738 34. The host_aton() function is supposed to be passed a string that is known
5739 to be a valid IP address. However, in the case of IPv6 addresses, it was
5740 not checking this. This is a hostage to fortune. Exim now panics and dies
5741 if the condition is not met. A case was found where this could be provoked
5742 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5743 components; fortuitously, this particular loophole had already been fixed
5744 by change 4.50/55 or 4.44/33 above.
5746 If there are any other similar loopholes, the new check in host_aton()
5747 itself should stop them being exploited. The report I received stated that
5748 data on the command line could provoke the exploit when Exim was running as
5749 exim, but did not say which command line option was involved. All I could
5750 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5751 running as the user.
5754 35. There was a buffer overflow vulnerability in the SPA authentication code
5755 (which came originally from the Samba project). I have added a test to the
5756 spa_base64_to_bits() function which I hope fixes it.
5759 36. The daemon start-up calls getloadavg() while still root for those OS that
5760 need the first call to be done as root, but it missed one case: when
5761 deliver_queue_load_max is set with deliver_drop_privilege. This is
5762 necessary for the benefit of the queue runner, because there is no re-exec
5763 when deliver_drop_privilege is set.
5765 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5766 was in use, was not putting the data itself into the right store pool;
5767 consequently, it could be overwritten for a subsequent message in the same
5768 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5771 38. Sometimes the final signoff response after QUIT could fail to get
5772 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5773 < 0 before doing a fflush(). This bug looks as though it goes back to the
5774 introduction of TLS in release 3.20, but "sometimes" must have been rare
5775 because the tests only now provoked it.
5781 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5782 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5783 Now it returns an integer. A number of calls were still expecting a T/F
5784 return. Fortuitously, in all cases, the tests worked in OK situations,
5785 which is the norm. However, things would have gone wrong on any write
5786 failures on the smtp file descriptor. This function is used when sending
5787 messages over SMTP and also when doing verify callouts.
5789 2. When Exim is called to do synchronous delivery of a locally submitted
5790 message (the -odf or -odi options), it no longer closes stderr before doing
5793 3. Implemented the mua_wrapper option.
5795 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5797 5. Implemented the functions header_remove(), header_testname(),
5798 header_add_at_position(), and receive_remove_recipient(), and exported them
5801 6. If an ACL "warn" statement specified the addition of headers, Exim already
5802 inserted X-ACL-Warn: at the start if there was no header name. However, it
5803 was not making this test for the second and subsequent header lines if
5804 there were newlines in the string. This meant that an invalid header could
5805 be inserted if Exim was badly configured.
5807 7. Allow an ACL "warn" statement to add header lines at the start or after all
5808 the Received: headers, as well as at the end.
5810 8. Added the rcpt_4xx retry error code.
5812 9. Added postmaster_mailfrom=xxx to callout verification option.
5814 10. Added mailfrom=xxxx to the callout verification option, for verify=
5817 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5818 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5820 12. Inserted some casts to stop certain compilers warning when using pointer
5821 differences as field lengths or precisions in printf-type calls (mostly
5822 affecting debugging statements).
5824 13. Added optional readline() support for -be (dynamically loaded).
5826 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5827 same clock tick as a message's arrival, so that its received time was the
5828 same as the "first fail" time on the retry record, and that message
5829 remained on the queue past the ultimate address timeout, every queue runner
5830 would try a delivery (because it was past the ultimate address timeout) but
5831 after another failure, the ultimate address timeout, which should have then
5832 bounced the address, did not kick in. This was a "< instead of <=" error;
5833 in most cases the first failure would have been in the next clock tick
5834 after the received time, and all would be well.
5836 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5837 being recognized when the domain list was tested by the match_domain
5838 condition in an expansion string.
5840 16. Added the ${str2b64: operator.
5842 17. Exim was always calling setrlimit() to set a large limit for the number of
5843 processes, without checking whether the existing limit was already
5844 adequate. (It did check for the limit on file descriptors.) Furthermore,
5845 errors from getrlimit() and setrlimit() were being ignored. Now they are
5846 logged to the main and panic logs, but Exim does carry on, to try to do its
5847 job under whatever limits there are.
5849 18. Imported PCRE 5.0.
5851 19. Trivial typo in log message " temporarily refused connection" (the leading
5854 20. If the log selector return_path_on_delivery was set and an address was
5855 redirected to /dev/null, the delivery process crashed because it assumed
5856 that a return path would always be set for a "successful" delivery. In this
5857 case, the whole delivery is bypassed as an optimization, and therefore no
5860 21. Internal re-arrangement: the function for sending a challenge and reading
5861 a response while authentication was assuming a zero-terminated challenge
5862 string. It's now changed to take a pointer and a length, to allow for
5863 binary data in such strings.
5865 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5867 23. Exim was not respecting finduser_retries when seeking the login of the
5868 uid under which it was called; it was always trying 10 times. (The default
5869 setting of finduser_retries is zero.) Also, it was sleeping after the final
5870 failure, which is pointless.
5872 24. Implemented tls_on_connect_ports.
5874 25. Implemented acl_smtp_predata.
5876 26. If the domain in control=submission is set empty, Exim assumes that the
5877 authenticated id is a complete email address when it generates From: or
5878 Sender: header lines.
5880 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5881 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5882 chown and chgrp in /bin and hostname in /usr/bin.
5884 28. Exim was keeping the "process log" file open after each use, just as it
5885 does for the main log. This opens the possibility of it remaining open for
5886 long periods when the USR1 signal hits a daemon. Occasional processlog
5887 errors were reported, that could have been caused by this. Anyway, it seems
5888 much more sensible not to leave this file open at all, so that is what now
5891 29. The long-running daemon process does not normally write to the log once it
5892 has entered its main loop, and it closes the log before doing so. This is
5893 so that log files can straightforwardly be renamed and moved. However,
5894 there are a couple of unusual error situations where the daemon does write
5895 log entries, and I had neglected to close the log afterwards.
5897 30. The text of an SMTP error response that was received during a remote
5898 delivery was being truncated at 512 bytes. This is too short for some of
5899 the long messages that one sometimes sees. I've increased the limit to
5902 31. It is now possible to make retry rules that apply only when a message has a
5903 specific sender, in particular, an empty sender.
5905 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5906 it possible to be selective about when SMTP synchronization is enforced.
5908 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5910 32. Implemented hosts_connection_nolog.
5912 33. Added an ACL for QUIT.
5914 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5917 35. Added mailbox_size and mailbox_filecount to appendfile.
5919 36. Added control = no_multiline_responses to ACLs.
5921 37. There was a bug in the logic of the code that waits for the clock to tick
5922 in the case where the clock went backwards by a substantial amount such
5923 that the microsecond fraction of "now" was more than the microsecond
5924 fraction of "then" (but the whole seconds number was less).
5926 38. Added support for the libradius Radius client library this is found on
5927 FreeBSD (previously only the radiusclient library was supported).
5933 1. When certain lookups returned multiple values in the form name=value, the
5934 quoting of the values was not always being done properly. Specifically:
5935 (a) If the value started with a double quote, but contained no whitespace,
5937 (b) If the value contained whitespace other than a space character (i.e.
5938 tabs or newlines or carriage returns) it was not quoted.
5939 This fix has been applied to the mysql and pgsql lookups by writing a
5940 separate quoting function and calling it from the lookup code. The fix
5941 should probably also be applied to nisplus, ibase and oracle lookups, but
5942 since I cannot test any of those, I have not disturbed their existing code.
5944 2. A hit in the callout cache for a specific address caused a log line with no
5945 reason for rejecting RCPT. Now it says "Previous (cached) callout
5946 verification failure".
5948 3. There was an off-by-one bug in the queryprogram router. An over-long
5949 return line was truncated at 256 instead of 255 characters, thereby
5950 overflowing its buffer with the terminating zero. As well as fixing this, I
5951 have increased the buffer size to 1024 (and made a note to document this).
5953 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5954 when Exim is waiting for an SMTP response from a remote server, Exim
5955 restarts its select() call on the socket, thereby resetting its timeout.
5956 This is not a problem when such interrupts are rare. Somebody set up a cron
5957 job to run exiwhat every 2 minutes, which is less than the normal select()
5958 timeout (5 or 10 minutes). This meant that the select() timeout never
5959 kicked in because it was always reset. I have fixed this by comparing the
5960 time when an interrupt arrives with the time at the start of the first call
5961 to select(). If more time than the timeout has elapsed, the interrupt is
5962 treated as a timeout.
5964 5. Some internal re-factoring in preparation for the addition of Sieve
5965 extensions (by MH). In particular, the "personal" test is moved to a
5966 separate function, and given an option for scanning Cc: and Bcc: (which is
5967 not set for Exim filters).
5969 6. When Exim created an email address using the login of the caller as the
5970 local part (e.g. when creating a From: or Sender: header line), it was not
5971 quoting the local part when it contained special characters such as @.
5973 7. Installed new OpenBSD configuration files.
5975 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5976 try to make them clearer.
5978 9. Callout options, other than the timeout value, were being ignored when
5979 verifying sender addresses in header lines. For example, when using
5981 verify = header_sender/callout=no_cache
5983 the cache was (incorrectly) being used.
5985 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5986 only the Cygwin environment.
5988 11. When return_path_on_delivery was set as a log selector, if different remote
5989 addresses in the same message used different return paths and parallel
5990 remote delivery occurred, the wrong values would sometimes be logged.
5991 (Whenever a remote delivery process finished, the return path value from
5992 the most recently started remote delivery process was logged.)
5994 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5995 lines when AUTH and/or TLS are in use. This is the "received protocol"
5996 field. Exim used to use "asmtp" for authenticated SMTP, without any
5997 indication (in the protocol name) for TLS use. Now it follows the RFC and
5998 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5999 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6000 names appear in log lines as well as in Received: header lines.
6002 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6003 extensions, and comparison tests, and to fix some bugs.
6005 14. Changes to the "personal" filter test:
6007 (1) The test was buggy in that it was just doing the equivalent of
6008 "contains" tests on header lines. For example, if a user's address was
6009 anne@some.where, the "personal" test would incorrectly be true for
6011 To: susanne@some.where
6013 This test is now done by extracting each address from the header in turn,
6014 and checking the entire address. Other tests that are part of "personal"
6015 are now done using regular expressions (for example, to check local parts
6016 of addresses in From: header lines).
6018 (2) The list of non-personal local parts in From: addresses has been
6019 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6020 taken from the Sieve specification recommendations.
6022 (3) If the message contains any header line starting with "List-" it is
6023 treated as non-personal.
6025 (4) The test for "circular" in the Subject: header line has been removed
6026 because it now seems ill-conceived.
6028 15. Minor typos in src/EDITME comments corrected.
6030 16. Installed latest exipick from John Jetmore.
6032 17. If headers_add on a router specified a text string that was too long for
6033 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6034 of string_sprintf() is now avoided.
6036 18. $message_body_size was not set (it was always zero) when running the DATA
6037 ACL and the local_scan() function.
6039 19. For the "mail" command in an Exim filter, no default was being set for
6040 the once_repeat time, causing a random time value to be used if "once" was
6041 specified. (If the value happened to be <= 0, no repeat happened.) The
6042 default is now 0s, meaning "never repeat". The "vacation" command was OK
6043 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6044 (I found it when inspecting the code).
6046 20. There is now an overall timeout for performing a callout verification. It
6047 defaults to 4 times the callout timeout, which applies to individual SMTP
6048 commands during the callout. The overall timeout applies when there is more
6049 than one host that can be tried. The timeout is checked before trying the
6050 next host. This prevents very long delays if there are a large number of
6051 hosts and all are timing out (e.g. when the network connections are timing
6052 out). The value of the overall timeout can be changed by specifying an
6053 additional sub-option for "callout", called "maxwait". For example:
6055 verify = sender/callout=5s,maxwait=20s
6057 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6058 to the end before writing, but this should make it even safer).
6060 22. Exim was forgetting that it had advertised PIPELINING for the second and
6061 subsequent messages on an SMTP connection. It was also not resetting its
6062 memory on STARTTLS and an internal HELO.
6064 23. When Exim logs an SMTP synchronization error within a session, it now
6065 records whether PIPELINING has been advertised or not.
6067 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6068 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6069 rather than long int.
6071 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6073 26. Added the never_mail option to autoreply.
6079 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6080 crash if the getsockname() call failed; this can happen if a connection is
6081 closed very soon after it is established. The problem was simply in the
6082 order in which certain operations were done, causing Exim to try to write
6083 to the SMTP stream before it had set up the file descriptor. The bug has
6084 been fixed by making things happen in the correct order.
6090 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6091 before the connection was closed, thus losing the rejection response.
6093 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6094 some early Solaris releases, but causes trouble in current releases where
6095 socklen_t is defined.
6097 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6100 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6103 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6104 that releasing the top of it at the end releases what was used for sub-
6105 expansions (unless the block got too big). However, discard this block if
6106 the first thing is a variable or header, so that we can use its block when
6107 it is dynamic (useful for very large $message_headers, for example).
6109 6. Lookups now cache *every* query, not just the most recent. A new, separate
6110 store pool is used for this. It can be recovered when all lookup caches are
6111 flushed. Lookups now release memory at the end of their result strings.
6112 This has involved some general refactoring of the lookup sources.
6114 7. Some code has been added to the store_xxx() functions to reduce the amount
6115 of flapping under certain conditions.
6117 8. log_incoming_interface used to affect only the <= reception log lines. Now
6118 it causes the local interface and port to be added to several more SMTP log
6119 lines, for example "SMTP connection from", and rejection lines.
6121 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6123 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6125 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6126 could be overwritten at the end of the current message (or the start of a
6127 new message if it was set in a HELO ACL). The value is now preserved for
6128 the duration of the SMTP connection.
6130 12. If a transport had a headers_rewrite setting, and a matching header line
6131 contained an unqualified address, that address was qualified, even if it
6132 did not match any rewriting rules. The underlying bug was that the values
6133 of the flags that permit the existence of unqualified sender and recipient
6134 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6135 non-local messages, and by -bnq for local messages) were not being
6136 preserved with the message after it was received.
6138 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6139 "next input=" as part of the text comprising the host identity instead of
6140 the correct text. The code was using the same buffer for two different
6141 strings. However, depending on which order the printing function evaluated
6142 its arguments, the bug did not always show up. Under Linux, for example, my
6143 test suite worked just fine.
6145 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6146 doesn't work with some older versions of Perl. It has been changed to "my",
6147 which in any case is probably the better facility to use.
6149 15. A really picky compiler found some instances of statements for creating
6150 error messages that either had too many or two few arguments for the format
6153 16. The size of the buffer for calls to the DNS resolver has been increased
6154 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6155 for addresses that have a lot of PTR records. This alleviates a problem; it
6156 does not fully solve it.
6158 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6159 buffer now truncates the list and logs the incident, which is the same
6160 action as happens when Exim is looking up a host name and its aliases.
6161 Previously in this situation something unpredictable would happen;
6162 sometimes it was "internal error: store_reset failed".
6164 18. If a server dropped the connection unexpectedly when an Exim client was
6165 using GnuTLS and trying to read a response, the client delivery process
6166 crashed while trying to generate an error log message.
6168 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6169 string, for example:
6171 warn message = H1: something\nH2: something
6173 the text was added as a single header line from Exim's point of view
6174 though it ended up OK in the delivered message. However, searching for the
6175 second and subsequent header lines using $h_h2: did not work. This has been
6176 fixed. Similarly, if a system filter added multiple headers in this way,
6177 the routers could not see them.
6179 20. Expanded the error message when iplsearch is called with an invalid key to
6180 suggest using net-iplsearch in a host list.
6182 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6183 ACLs are no longer actually imposed (and a message to that effect is
6186 22. If a "gecos" field in a passwd entry contained escaped characters, in
6187 particular, if it contained a \" sequence, Exim got it wrong when building
6188 a From: or a Sender: header from that name. A second bug also caused
6189 incorrect handling when an unquoted " was present following a character
6190 that needed quoting.
6192 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6193 was not being matched caselessly.
6195 24. Arranged for all hyphens in the exim.8 source to be escaped with
6198 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6199 properly. Recipient callout cache records were still being keyed to include
6200 the sender, even when use_sender was set false. This led to far more
6201 callouts that were necessary. The sender is no longer included in the key
6202 when use_sender is false.
6204 26. Added "control = submission" modifier to ACLs.
6206 27. Added the ${base62d: operator to decode base 62 numbers.
6208 28. dnsdb lookups can now access SRV records.
6210 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6211 the configuration file.
6213 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6214 (-v) mode. This makes the output for a verbose queue run more intelligible.
6216 31. Added a use_postmaster feature to recipient callouts.
6218 32. Added the $body_zerocount variable, containing the number of binary zero
6219 bytes in the message body.
6221 33. The time of last modification of the "new" subdirectory is now used as the
6222 "mailbox time last read" when there is a quota error for a maildir
6225 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6227 35. Added +ignore_unknown as a special item in host lists.
6229 36. Code for decoding IPv6 addresses in host lists is now included, even if
6230 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6231 address was recognized as an IP address, but was then not correctly decoded
6232 into binary, causing unexpected and incorrect effects when compared with
6239 1. Very minor rewording of debugging text in manualroute to say "list of
6240 hosts" instead of "hostlist".
6242 2. If verify=header_syntax was set, and a header line with an unqualified
6243 address (no domain) and a large number of spaces between the end of the
6244 name and the colon was received, the reception process suffered a buffer
6245 overflow, and (when I tested it) crashed. This was caused by some obsolete
6246 code that should have been removed. The fix is to remove it!
6248 3. When running in the test harness, delay a bit after writing a bounce
6249 message to get a bit more predictability in the log output.
6251 4. Added a call to search_tidyup() just before forking a reception process. In
6252 theory, someone could use a lookup in the expansion of smtp_accept_max_
6253 per_host which, without the tidyup, could leave open a database connection.
6255 5. Added the variables $recipient_data and $sender_data which get set from a
6256 lookup success in an ACL "recipients" or "senders" condition, or a router
6257 "senders" option, similar to $domain_data and $local_part_data.
6259 6. Moved the writing of debug_print from before to after the "senders" test
6262 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6263 problems for message scanning, either using a data ACL, or using
6264 local_scan() because the Received: header was not generated till after they
6265 were called (in order to set the time as the time of reception completion).
6266 I have revised the way this works. The header is now generated after the
6267 body is received, but before the ACL or local_scan() are called. After they
6268 are run, the timestamp in the header is updated.
6274 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6275 before starting a queue runner without re-exec. This happened only when
6276 deliver_drop_privilege was set or when the Exim user was set to root. The
6277 effect of the bug was that timeouts during subsequent deliveries caused
6278 crashes instead of being properly handled. The handler is now left at its
6279 default (and expected) setting.
6281 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6282 message, again when deliver_drop_privilege is set or Exim is run as root.
6283 The bug described in (1) was not present in this case, but the tidying up
6284 of the other signals was missing. I have made the two cases consistent.
6286 3. The ignore_target_hosts setting on a manualroute router was being ignored
6287 for hosts that were looked up using the /MX notation.
6289 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6292 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6293 operated on the sender address. After changing the $sender_address to <>
6294 for the sender address verify, Exim was re-instated it as the original
6295 (before rewriting) address, but remembering that it had rewritten it, so it
6296 wasn't rewriting it again. This bug also had the effect of breaking the
6297 sender address verification caching when the sender address was rewritten.
6299 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6300 This has been changed so that if the ip literal address matches
6301 ignore_target_hosts, the router declines.
6303 7. Added expansion conditions match_domain, match_address, and match_local_
6304 part (NOT match_host).
6306 8. The placeholder for the Received: header didn't have a length field set.
6308 9. Added code to Exim itself and to exim_lock to test for a specific race
6309 condition that could lead to file corruption when using MBX delivery. The
6310 issue is with the lockfile that is created in /tmp. If this file is removed
6311 after a process has opened it but before that process has acquired a lock,
6312 there is the potential for a second process to recreate the file and also
6313 acquire a lock. This could lead to two Exim processes writing to the file
6314 at the same time. The added code performs the same test as UW imapd; it
6315 checks after acquiring the lock that its file descriptor still refers to
6316 the same named file.
6318 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6319 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6322 11. Added the smtp_active_hostname option. If used, this will typically be made
6323 to depend on the incoming interface address. Because $interface_address is
6324 not set up until the daemon has forked a reception process, error responses
6325 that can happen earlier (such as "too many connections") no longer contain
6328 12. If an expansion in a condition on a "warn" statement fails because a lookup
6329 defers, the "warn" statement is abandoned, and the next ACL statement is
6330 processed. Previously this caused the whole ACL to be aborted.
6332 13. Added the iplsearch lookup type.
6334 14. Added ident_timeout as a log selector.
6336 15. Added tls_certificate_verified as a log selector.
6338 16. Added a global option tls_require_ciphers (compare the smtp transport
6339 option of the same name). This controls incoming TLS connections.
6341 17. I finally figured out how to make tls_require_ciphers do a similar thing
6342 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6343 before starting the TLS session.
6345 18. Tabs are now shown as \t in -bP output.
6347 19. If the log selector return_path_on_delivery was set, Exim crashed when
6348 bouncing a message because it had too many Received: header lines.
6350 20. If two routers both had headers_remove settings, and the first one included
6351 a superfluous trailing colon, the final name in the first list and the
6352 first name in the second list were incorrectly joined into one item (with a
6353 colon in the middle).
6359 1. Added -C and -D options to the exinext utility, mainly to make it easier
6360 to include in the automated testing, but these could be helpful when
6361 multiple configurations are in use.
6363 2. The exinext utility was not formatting the output nicely when there was
6364 an alternate port involved in the retry record key, nor when there was a
6365 message id as well (for retries that were specific to a specific message
6366 and a specific host). It was also confused by IPv6 addresses, because of
6367 the additional colons they contain. I have fixed the IPv4 problem, and
6368 patched it up to do a reasonable job for IPv6.
6370 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6371 delivery, the log line now contains "pipelined" if PIPELINING was used.
6373 4. An SMTP transport process used to panic and die if the bind() call to set
6374 an explicit outgoing interface failed. This has been changed; it is now
6375 treated in the same way as a connect() failure.
6377 5. A reference to $sender_host_name in the part of a conditional expansion
6378 that was being skipped was still causing a DNS lookup. This no longer
6381 6. The def: expansion condition was not recognizing references to header lines
6382 that used bh_ and bheader_.
6384 7. Added the _cache feature to named lists.
6386 8. The code for checking quota_filecount in the appendfile transport was
6387 allowing one more file than it should have been.
6389 9. For compatibility with Sendmail, the command line option
6397 and sets the incoming protocol and host name (for trusted callers). The
6398 host name and its colon can be omitted when only the protocol is to be set.
6399 Note the Exim already has two private options, -pd and -ps, that refer to
6400 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6401 "s", but I don't think that's a major issue.
6403 10. A number of refactoring changes to the code, none of which should affect
6406 (a) The number of logging options was getting close to filling up the
6407 32-bit word that was used as a bit map. I have split them into two classes:
6408 those that are passed in the argument to log_write(), and those that are
6409 only ever tested independently outside of that function. These are now in
6410 separate 32-bit words, so there is plenty of room for expansion again.
6411 There is no change in the user interface or the logging behaviour.
6413 (b) When building, for example, log lines, the code previously used a
6414 macro that called string_cat() twice, in order to add two strings. This is
6415 not really sufficiently general. Furthermore, there was one instance where
6416 it was actually wrong because one of the argument was used twice, and in
6417 one call a function was used. (As it happened, calling the function twice
6418 did not affect the overall behaviour.) The macro has been replaced by a
6419 function that can join an arbitrary number of extra strings onto a growing
6422 (c) The code for expansion conditions now uses a table and a binary chop
6423 instead of a serial search (which was left over from when there were very
6424 few conditions). Also, it now recognizes conditions like "pam" even when
6425 the relevant support is not compiled in: a suitably worded error message is
6426 given if an attempt is made to use such a condition.
6428 11. Added ${time_interval:xxxxx}.
6430 12. A bug was causing one of the ddress fields not to be passed back correctly
6431 from remote delivery subprocesses. The field in question was not being
6432 subsequently used, so this caused to problems in practice.
6434 13. Added new log selectors queue_time and deliver_time.
6436 14. Might have fixed a bug in maildirsizefile handling that threw up
6437 "unexpected character" debug warnings, and recalculated the data
6438 unnecessarily. In any case, I expanded the warning message to give more
6441 15. Added the message "Restricted characters in address" to the statements in
6442 the default ACL that block characters like @ and % in local parts.
6444 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6445 Three changes have been made:
6447 (a) There was a serious bug; a negative response to MAIL caused the whole
6448 recipient domain to be cached as invalid, thereby blocking all messages
6449 to all local parts at the same domain, from all senders. This bug has
6450 been fixed. The domain is no longer cached after a negative response to
6451 MAIL if the sender used is not empty.
6453 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6456 (c) A new callout option, "use_sender" has been added for people who want
6457 the modified behaviour.
6463 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6466 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6467 indeed breaks things for older releases.
6469 3. Added additional logging to the case where there is a problem reading data
6470 from a filter that is running in a subprocess using a pipe, in order to
6471 try to track down a specific problem.
6473 4. Testing facility fudge: when running in the test harness and attempting
6474 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6475 getting "No route to host". Convert this to a timeout.
6477 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6480 6. Some OS don't have socklen_t but use size_t instead. This affects the
6481 fifth argument of getsockopt() amongst other things. This is now
6482 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6483 can be set for individual OS. I have set it for SunOS5, OSF1, and
6484 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6485 some earlier ones do not.
6487 7. Change 4.30/15 was not doing the test caselessly.
6489 8. The standard form for an IPv6 address literal was being rejected by address
6490 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6491 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6492 this, as well as the form without the "IPv6" on the front (but only when
6493 address literals are enabled, of course).
6495 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6497 10. Exim crashed if a message with an empty sender address specified by -f
6498 encountered a router with an errors_to setting. This could be provoked only
6499 by a command such as
6503 where an empty string was supplied; "<>" did not hit this bug.
6505 11. Installed PCRE release 4.5.
6507 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6508 remained set. It is now erased.
6510 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6511 times from message ids (which are base 36 rather than the normal 62).
6513 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6514 were being counted as actual protocol errors, and logged if the log
6515 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6516 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6517 and DATA following a set of rejected RCPTs do not count as protocol errors.
6518 In other words, Exim assumes they were pipelined, though this may not
6519 actually be the case. Of course, in all cases the client gets an
6520 appropriate error code.
6522 15. If a lookup fails in an ACL condition, a message about the failure may
6523 be available; it is used if testing the ACL cannot continue, because most
6524 such messages specify what the cause of the deferral is. However, some
6525 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6526 that caused an old message to be retained and used if a later statement
6527 caused a defer, replacing the real cause of the deferral.
6529 16. If an IP address had so many PTR records that the DNS lookup buffer
6530 was not large enough to hold them, Exim could crash while trying to process
6531 the truncated data. It now detects and logs this case.
6533 17. Further to 4.21/58, another change has been made: if (and only if) the
6534 first line of a message (the first header line) ends with CRLF, a bare LF
6535 in a subsequent header line has a space inserted after it, so as not to
6536 terminate the header.
6538 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6539 unnecessarily, used atoi() instead of strtol(), and didn't check the
6540 termination when getting file sizes from file names by regex.
6542 19. Completely re-implemented the support for maildirsize files, in the light
6543 of a number of problems with the previous contributed implementation
6544 (4.30/29). In particular:
6546 . If the quota is zero, the maildirsize file is maintained, but no quota is
6549 . If the maildir directory does not exist, it is created before any attempt
6550 to write a maildirsize file.
6552 . The quota value in the file is just a cache; if the quota is changed in
6553 the transport, the new value overrides.
6555 . A regular expression is available for excluding directories from the
6558 20. The autoreply transport checks the characters in options that define the
6559 message's headers; it allows continued headers, but it was checking with
6560 isspace() after an embedded newline instead of explicitly looking for a
6563 21. If all the "regular" hosts to which an address was routed had passed their
6564 expiry times, and had not reached their retry times, the address was
6565 bounced, even if fallback hosts were defined. Now Exim should go on to try
6568 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6569 equivalent code in the SMTP transport. Some hosts send humungous responses
6570 to HELO/EHLO, more than 1024 it seems.
6572 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6573 gives compiler warnings in some environments. I've now done it "properly",
6576 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6577 (because of problems with the built-in one) was declared to return uschar *
6578 instead of char *, causing compiler failure.
6580 25. Fixed a file descriptor leak when processing alias/forward files.
6582 26. Fixed a minor format string issue in dbfn.c.
6584 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6586 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6587 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6590 29. When a sender address is verified, it is cached, to save repeating the test
6591 when there is more than one recipient in a message. However, when the
6592 verification involves a callout, it is possible for different callout
6593 options to be set for different recipients. It is too complicated to keep
6594 track of this in the cache, so now Exim always runs a verification when a
6595 callout is required, relying on the callout cache for the optimization.
6596 The overhead is duplication of the address routing, but this should not be
6599 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6600 to be verified with callout=postmaster, and the main callout worked but the
6601 postmaster check failed, the verification correctly failed. However, if a
6602 subsequent RCPT command asked for sender verification *without* the
6603 postmaster check, incorrect caching caused this verification also to fail,
6606 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6607 it was not caching the DNS options (qualify_single, search_parents) that
6608 were used when the lookup failed. A subsequent lookup with different
6609 options therefore always gave the same answer, though there were cases
6610 where it should not have. (Example: a "domains = !$mx_any" option on a
6611 dnslookup router: the "domains" option is always processed without any
6612 widening, but the router might have qualify_single set.) Now Exim uses the
6613 cached value only when the same options are set.
6615 32. Added John Jetmore's "exipick" utility to the distribution.
6617 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6618 than a timeout (e.g. a certificate is required, and is not provided), an
6619 Exim server now closes the connection immediately. Previously it waited for
6620 the client to close - but if the client is SSL, it seems that they each
6621 wait for each other, leading to a delay before one of them times out.
6623 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6624 maintained 0.8.x compatibility because I don't think many are using it, and
6625 it is clearly obsolete.
6627 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6630 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6631 client certificate was expired. A simple patch fixes this, though I don't
6632 understand the full logic of why the verify callback is called multiple
6635 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6636 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6637 which causes problems with some clients (such as the Certicom SSL Plus
6638 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6639 disables the coutermeasure allowing Eudora to connect."
6641 38. Exim was not checking that a write() to a log file succeeded. This could
6642 lead to Bad Things if a log got too big, in particular if it hit a file
6643 size limit. Exim now panics and dies if it cannot write to a log file, just
6644 as it does if it cannot open a log file.
6646 39. Modified OS/Makefile-Linux so that it now contains
6648 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6650 The two -D definitions ensure that Exim is compiled with large file
6651 support, which makes it possible to handle log files that are bigger than
6654 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6655 instance) a domain was checked against a named list that involved a lookup,
6656 causing $domain_data to be set, then another domain was checked against the
6657 same list, then the first domain was re-checked, the value of $domain_data
6658 after the final check could be wrong. In particular, if the second check
6659 failed, it could be set empty. This bug probably also applied to
6662 41. The strip_trailing_dot option was not being applied to the address given
6663 with the -f command-line option.
6665 42. The code for reading a message's header from the spool was incrementing
6666 $received_count, but never initializing it. This meant that the value was
6667 incorrect (doubled) while delivering a message in the same process in which
6668 it was received. In the most common configuration of Exim, this never
6669 happens - a fresh exec is done - but it can happen when
6670 deliver_drop_privilege is set.
6672 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6673 it now includes up to 150 characters of the unexpected data in the log
6676 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6677 and building data strings. The size of both of these buffers was 10 000
6678 bytes - far larger than anybody would *ever* want, thought I. Needless to
6679 say, somebody hit the limit. I have increased the maximum line length to
6680 20 000 and the maximum data length of concatenated lines to 100 000. I have
6681 also fixed two bugs, because there was no checking on these buffers. Tsk,
6682 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6683 buffer is too small.
6685 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6686 lsearch lookups. Now it does.
6688 46. When parsing a route_list item in a manualroute router, a fixed-length
6689 buffer was used for the list of hosts. I made this 1024 bytes long,
6690 thinking that nobody would ever have a list of hosts that long. Wrong.
6691 Somebody had a whole pile of complicated expansion conditions, and the
6692 string was silently truncated, leading to an expansion error. It turns out
6693 that it is easier to change to an unlimited length (owing to other changes
6694 that have happened since this code was originally written) than to build
6695 structure for giving a limitation error. The length of the item that
6696 expands into the list of hosts is now unlimited.
6698 47. The lsearch lookup could not handle data where the length of text line was
6699 more than 4095 characters. Such lines were truncated, leading to shortened
6700 data being returned. It should now handle lines of any length.
6702 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6703 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6706 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6707 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6708 no longer happens in comment lines. A list of replacements is now placed
6709 at the head of all of the source files, except those whose only change is
6710 to replace PERL_COMMAND in the very first #! line.
6712 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6713 messages on the queue, with a bottom-up merge sort, using code contributed
6714 by Michael Haardt. This should make operations like -bp somewhat faster on
6715 large queues. It won't affect queue runners, except when queue_run_in_order
6718 51. Installed eximstats 1.31 in the distribution.
6720 52. Added support for SRV lookups to the dnslookup router.
6722 53. If an ACL referred to $message_body or $message_body_end, the value was not
6723 reset for any messages that followed in the same SMTP session.
6725 54. The store-handling optimization for building very long strings was not
6726 differentiating between the different store pools. I don't think this
6727 actually made any difference in practice, but I've tidied it.
6729 55. While running the routers to verify a sender address, $sender_address
6730 was still set to the sender address. This is wrong, because when routing to
6731 send a bounce to the sender, it would be empty. Therefore, I have changed
6732 it so that, while verifying a sender address, $sender_address is set to <>.
6733 (There is no change to what happens when verifying a recipient address.)
6735 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6736 target A or AAAA records (if not already returned) without resetting the
6737 qualify_single or search_parents options of the DNS resolver. These are
6738 inappropriate in this case because the targets of MX and SRV records must
6739 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6740 target that, when qualified, matched something in the local domain. These
6741 two options are now turned off when doing these lookups.
6743 57. It seems that at least some releases of Reiserfs (which does not have the
6744 concept of a fixed number of inodes) returns zero and not -1 for the
6745 number of available inodes. This interacted badly with check_spool_inodes,
6746 which assumed that -1 was the "no such thing" setting. What I have done is
6747 to check that the total number of inodes is greater than zero before doing
6748 the test of how many are available.
6750 58. When a "warn" ACL statement has a log_message modifier, the message is
6751 remembered, and not repeated. This is to avoid a lot of repetition when a
6752 message has many recipients that cause the same warning to be written.
6753 However, Exim was preserving the list of already written lines for an
6754 entire SMTP session, which doesn't seem right. The memory is now reset if a
6755 new message is started.
6757 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6758 debugging output unless log_rewrite was also set.
6760 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6761 of (int)(handle) when we know that handle contains (void *)(-1).
6763 61. The Exim daemon panic-logs an error return when it closes the incoming
6764 connection. However "connection reset by peer" seems to be common, and
6765 isn't really an error worthy of noting specially, so that particular error
6768 62. When Exim is trying to find all the local interfaces, it used to panic and
6769 die if the ioctl to get the interface flags failed. However, it seems that
6770 on at least one OS (Solaris 9) it is possible to have an interface that is
6771 included in the list of interfaces, but for which you get a failure error
6772 for this call. This happens when the interface is not "plumbed" into a
6773 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6774 failure of the "get flags" call assumes that the interface is down.
6776 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6777 makes life easier for people who are doing arithmetic on fields extracted
6778 from dates, where you often get leading zeros that should not be
6779 interpreted as octal.
6781 64. Added qualify_domain to the redirect router, to override the global
6784 65. If a pathologically long header line contained very many addresses (the
6785 report of this problem mentioned 10 000) and each of them was rewritten,
6786 Exim could use up a very large amount of memory. (It kept on making new
6787 copies of the header line as it rewrote, and never released the old ones.)
6788 At the expense of a bit more processing, the header rewriting function has
6789 been changed so that it no longer eats memory in this way.
6791 66. The generation of the Received: header has been moved from the time that a
6792 message starts to be received, to the time that it finishes. The timestamp
6793 in the Received: header should now be very close to that of the <= log
6794 line. There are two side-effects of this change:
6796 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6797 logged header lines no longer include the local Received: line, because
6798 it has not yet been created. The same applies to a copy of the message
6799 that is returned to a non-SMTP sender when a message is rejected.
6801 (b) When a filter file is tested using -bf, no additional Received: header
6802 is added to the test message. After some thought, I decided that this
6805 This change does not affect the value of $received_for. It is still set
6806 after address rewriting, but before local_scan() is called.
6808 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6810 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6811 gave an unhelpful panic error message, and a defer error. I have managed to
6812 change this behaviour so that it now rejects any supplied certificate,
6813 which seems right, as the list of acceptable certificates is empty.
6815 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6816 gave an unhelpful defer error. I have not managed to make this reject any
6817 supplied certificates, but the error message it gives is "no certificate
6818 supplied", which is not helpful.
6820 70. exigrep's output now also includes lines that are not associated with any
6821 message, but which match the given pattern. Implemented by a patch from
6822 Martin Sluka, which also tidied up the Perl a bit.
6824 71. Recipient callout verification, like sender verification, was using <> in
6825 the MAIL FROM command. This isn't really the right thing, since the actual
6826 sender may affect whether the remote host accepts the recipient or not. I
6827 have changed it to use the actual sender in the callout; this means that
6828 the cache record is now keyed on a recipient/sender pair, not just the
6829 recipient address. There doesn't seem to be a real danger of callout loops,
6830 since a callout by the remote host to check the sender would use <>.
6831 [SEE ABOVE: changed after hitting problems.]
6833 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6834 temporary errors. However, in the case of such a code being given after
6835 the end of a data transmission (i.e. after ".") Exim was failing to write
6836 a retry record for the message. (Yes, there was some broken host that was
6837 actually sending 8xx at this point.)
6839 73. An unknown lookup type in a host list could cause Exim to panic-die when
6840 the list was checked. (An example that provoked this was putting <; in the
6841 middle of a list instead of at the start.) If this happened during a DATA
6842 ACL check, a -D file could be left lying around. This kind of configuration
6843 error no longer causes Exim to die; instead it causes a defer error. The
6844 incident is still logged to the main and panic logs.
6846 74. Buglet left over from Exim 3 conversion. The message "too many messages
6847 in one connection" was written to the rejectlog but not the mainlog, except
6848 when address rewriting (yes!) was being logged.
6850 75. Added write_rejectlog option.
6852 76. When a system filter was run not as root (that is, when system_filter_user
6853 was set), the values of the $n variables were not being returned to the
6854 main process; thus, they were not subsequently available in the $sn
6857 77. Added +return_path_on_delivery log selector.
6859 78. A connection timeout was being treated differently from recipients deferred
6860 when testing hosts_max_try with a message that was older than the host's
6861 retry timeout. (The host should not be counted, thus allowing all hosts to
6862 be tried at least once before bouncing.) This may have been the cause of an
6863 occasionally reported bug whereby a message would remain on the queue
6864 longer than the retry timeout, but would be bounced if a delivery was
6865 forced. I say "may" because I never totally pinned down the problem;
6866 setting up timeout/retry tests is difficult. See also the next item.
6868 79. The ultimate address timeout was not being applied to errors that involved
6869 a combination of host plus message (for example, a timeout on a MAIL
6870 command). When an address resolved to a number of possible hosts, and they
6871 were not all tried for each delivery (e.g. because of hosts_max_try), a
6872 message could remain on the queue longer than the retry timeout.
6874 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6877 81. Fixed an obscure SMTP outgoing bug which required at least the following
6878 conditions: (a) there was another message waiting for the same server;
6879 (b) the server returned 5xx to all RCPT commands in the first message so
6880 that the message was not completed; (c) the server dropped the connection
6881 or gave a negative response to the RSET that Exim sends to abort the
6882 transaction. The observed case was a dropped connection after DATA that had
6883 been sent in pipelining mode. That is, the server had advertised PIPELINING
6884 but was not implementing it correctly. The effect of the bug was incorrect
6885 behaviour, such as trying another host, and this could lead to a crash.
6891 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6892 and daemon.c were passed as pointers to ints; they should have been
6893 pointers to socklen_t variables (which are typically unsigned ints).
6895 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6898 3. Fixed a really odd bug that affected only the testing scheme; patching a
6899 certain fixed string in the binary changed the value of another string that
6900 happened to be identical to the end of the original first string.
6902 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6903 name", it returns that address as the IP address. On some operating
6904 systems (e.g. Solaris), it also passes back the IP address string as the
6905 "host name". However, on others (e.g. Linux), it passes back an empty
6906 string. Exim wasn't checking for this, and was changing the host name to an
6907 empty string, assuming it had been canonicalized.
6909 5. Although rare, it is permitted to have more than one PTR record for a given
6910 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6911 all the names associated with an address, because they do in Solaris.
6912 However, it seems that they do not in Linux for data that comes from the
6913 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6914 I found this out when I moved to a new Linux workstation and tried to run
6915 the Exim test suite.
6917 To get round this problem I have changed the code so that it now does its
6918 own call to the DNS to look up PTR records when searching for a host name.
6919 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6920 addresses that are only in /etc/hosts are still found.
6922 This behaviour is, however, controlled by an option called host_lookup_
6923 order, which defaults to "bydns:byaddr". If people want to use the other
6924 order, or indeed, just use one or the other means of lookup, they can
6925 specify it in this variable.
6927 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6928 some operating systems, this comes back from gethostbyaddr() as an empty
6929 string, and this is what Exim used to test for. However, it seems that in
6930 other systems, "." is yielded. Exim now tests for this case too.
6932 7. The values of check_spool_space and check_log_space are now held internally
6933 as a number of kilobytes instead of an absolute number of bytes. If a
6934 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6935 kilobyte. This means that much larger values can be stored.
6937 8. Exim monitor: an attempt to get the action menu when not actually pointing
6938 at a message produces an empty menu entitled "No message selected". This
6939 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6940 no entries in it ("Shell widget menu has zero width and/or height"). So I
6941 have added a single, blank menu entry in this case.
6943 9. Added ${quote_local_part.
6945 10. MIME decoding is now applied to the contents of Subject: header lines when
6948 11. Now that a reference to $sender_host_address automatically causes a reverse
6949 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6950 host lookup before query-style lookups in lists that might use this
6951 variable. This has therefore been abolished, and the "net-" prefix is no
6952 longer necessary for query-style lookups.
6954 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6955 have been a typo for LFLAGS, so it has been changed.
6957 13. The install script calls Exim with "-C /dev/null" in order to find the
6958 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6959 to be output. However, since Exim outputs its version number before the
6960 error, it didn't break the script. It just looked ugly. I fixed this by
6961 always allowing "-C /dev/null" if the caller is root.
6963 14. Ignore overlarge ACL variable number when reading spool file - insurance
6964 against a later release with more variables having written the file.
6966 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6967 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6968 this, as well as the form without the "IPv6" on the front.
6970 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6971 OS/Makefile-Darwin file.
6973 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6974 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6976 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6977 variable. However, some OS use ints for inodes. I've added cast to long int
6978 to get rid of the compiler warning.
6980 19. I had forgotten to lock out "/../" in configuration file names when
6981 ALT_CONFIG_PREFIX was set.
6983 20. Routers used for verification do not need to specify transports. However,
6984 if such a router generated a host list, and callout was configured, Exim
6985 crashed, because it could not find a port number from the (non-existent)
6986 transport. It now assumes port 25 in this circumstance.
6988 21. Added the -t option to exigrep.
6990 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6991 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6992 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6993 them is set, LOOKUP_LSEARCH is forced.
6995 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6997 24. Added sender and host information to the "rejected by local_scan()" log
6998 line; previously there was no indication of these.
7000 25. Added .include_if_exists.
7002 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7003 Linux. It turns out that not all file systems support this. Apparently some
7004 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7005 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7006 not supported on the file descriptor, is now ignored when Exim is trying to
7007 sync a directory. This applies only to Linux.
7009 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7011 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7012 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7015 29. Added support for maildirsize files from supplied patch (modified a bit).
7017 30. The use of :fail: followed by an empty string could lead Exim to respond to
7018 sender verification failures with (e.g.):
7020 550 Verification failed for <xxx>
7021 550 Sender verify failed
7023 where the first response line was missing the '-' that indicates it is not
7024 the final line of the response.
7026 31. The loop for finding the name of the user that called Exim had a hardwired
7027 limit of 10; it now uses the value of finduser_retries, which is used for
7028 all other user lookups.
7030 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7033 33. Exim was neglecting to zero errno before one call of strtol() when
7034 expanding a string and expecting an integer value. On some systems this
7035 resulted in spurious "integer overflow" errors. Also, it was casting the
7036 result into an int without checking.
7038 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7039 did not work. The code looks as if it has *never* worked, though it appears
7040 to have been documented since at least release 1.62. I have made it work.
7042 35. The "timeout_DNS" error in retry rules, also documented since at least
7043 1.62, also never worked. As it isn't clear exactly what this means, and
7044 clearly it isn't a major issue, I have abolished the feature by treating it
7045 as "timeout", and writing a warning to the main and panic logs.
7047 36. The display of retry rules for -brt wasn't always showing the error code
7050 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7051 timeout_connect_A, timeout_connect_MX.
7053 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7054 to the empty sender.
7056 39. The daemon was not analysing the content of -oX till after it had closed
7057 stderr and disconnected from the controlling terminal. This meant that any
7058 syntax errors were only noted on the panic log, and the return code from
7059 the command was 0. By re-arranging the code a little, I've made the
7060 decoding happen first, so such errors now appear on stderr, and the return
7061 code is 1. However, the actual setting up of the sockets still happens in
7062 the disconnected process, so errors there are still only recorded on the
7065 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7066 connections (as happens on some IP stacks) was logged at start up time as
7067 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7068 it from "IPv6 and IPv4", which means that two separate sockets are being
7071 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7072 says whether AF_INET or AF_INET6 was passed as an argument.
7074 42. Exiwhat output was messed up when time zones were included in log
7077 43. Exiwhat now gives more information about the daemon's listening ports,
7078 and whether -tls-on-connect was used.
7080 44. The "port" option of the smtp transport is now expanded.
7082 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7083 silently ignored. Now an error message is written to the main and panic
7086 46. There's a new ACL modifier called "logwrite" which writes to a log file
7087 as soon as it is encountered.
7089 47. Added $local_user_uid and $local_user_gid at routing time.
7091 48. Exim crashed when trying to verify a sender address that was being
7094 49. Exim was recognizing only a space character after ".include". It now also
7095 recognizes a tab character.
7097 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7098 extracting the relevant information from the specification. The man page no
7099 longer contains scrambled data for the -d option, and I've added a section
7100 at the front about calling Exim under different names.
7102 51. Added "extra_headers" argument to the "mail" command in filter files.
7104 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7107 53. Installed eximstats 1.29.
7109 54. Added transport_filter_timeout as a generic transport option.
7111 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7112 Cc: header lines. This was required by RFC 822, but it not required by RFC
7115 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7116 incoming messages that did not have them. Now it does so only if the
7117 message originates locally, that is, if there is no associated remote host
7118 address. When Resent- header lines are present, this applies to the Resent-
7119 lines rather than the non-Resent- lines.
7121 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7122 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7124 58. Messages for configuration errors now include the name of the main
7125 configuration file - useful now that there may be more than one file in a
7126 list (.included file names were always shown).
7128 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7129 for those rare installations that do not start the daemon as root or run it
7130 setuid root. I've cut out the call to initgroups() if the daemon is not
7133 60. The Exim user and group can now be bound into the binary as text strings
7134 that are looked up at the start of Exim's processing.
7136 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7138 62. Added $mailstore_basename variable.
7140 63. Installed patch to sieve.c from Michael Haardt.
7142 64. When Exim failed to open the panic log after failing to open the main log,
7143 the original message it was trying to log was written to stderr and debug
7144 output, but if they were not available (the usual case in production), it
7145 was lost. Now it is written to syslog before the two lines that record the
7146 failures to open the logs.
7148 65. Users' Exim filters run in subprocesses under the user's uid. It is
7149 possible for a "deliver" command or an alias in a "personal" command to
7150 provoke an address rewrite. If logging of address rewriting is configured,
7151 this fails because the process is not running as root or exim. There may be
7152 a better way of dealing with this, but for the moment (because 4.30 needs
7153 to be released), I have disabled address rewrite logging when running a
7154 filter in a non-root, non-exim process.
7160 1. The buildconfig auxiliary program wasn't quoting the value set for
7161 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7162 not defined. This bug was masked in 4.22 by the effect that was fixed in
7165 2. Some messages that were rejected after a message id was allocated were
7166 shown as "incomplete" by exigrep. It no longer does this for messages that
7167 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7169 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7170 have allow_domain_literals set, the ID did not get logged in the <= line.
7171 Domain literals are now always recognized in Message-ID: header lines.
7173 4. The first argument for a ${extract expansion item is the key name or field
7174 number. Leading and trailing spaces in this item were not being ignored,
7175 causing some misleading effects.
7177 5. When deliver_drop_privilege was set, single queue runner processes started
7178 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7179 same command in the process it spins off) were not dropping privilege.
7181 6. When the daemon running as "exim" started a queue runner, it always
7182 re-executed Exim in the spun-off process. This is a waste of effort when
7183 deliver_drop_privilege is set. The new process now just calls the
7184 queue-runner function directly.
7190 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7193 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7194 ignored. Though the use of -oP was forcing the writing of a pid file, it
7195 was always written to the default place.
7197 3. If the message "no IP address found for host xxxx" is generated during
7198 incoming verification, it is now followed by identification of the incoming
7199 connection (so you can more easily find what provoked it).
7201 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7203 5. Added some features to "harden" Exim a bit more against certain attacks:
7205 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7206 be put in Local/Makefile. This is like the never_users runtime option,
7207 but it cannot be overridden. The default setting is "root".
7209 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7210 prefix string with which any file named in a -C command line option
7213 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7214 is retained for -C and -D only if the caller of Exim is root. Without
7215 it, the exim user may also use -C and -D and retain privilege.
7217 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7218 command line option is disabled.
7220 6. Macro names set by the -D option must start with an upper case letter, just
7221 like macro names defined in the configuration file.
7223 7. Added "dereference=" facility to LDAP.
7225 8. Two instances of the typo "uknown" in the source files are fixed.
7227 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7228 the Configure-Makefile script screwed up while processing it.
7230 10. Incorporated PCRE 4.4.
7232 11. The SMTP synchronization check was not operating right at the start of an
7233 SMTP session. For example, it could not catch a HELO sent before the client
7234 waited for the greeting. There is now a check for outstanding input at the
7235 point when the greeting is written. Because of the duplex, asynchronous
7236 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7237 way, but not yet received, when the check is performed.
7239 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7240 on TCP/IP sockets, because this apparently causes some broken clients to
7243 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7244 unchanged) from the Cygwin maintainer.
7246 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7247 was supported instead of testing for mbx. Effectively a typo.
7249 15. The spa authenticator server code was not checking that the input it
7250 received was valid base64.
7252 16. The debug output line for the "set" modifier in ACLs was not showing the
7253 name of the variable that was being set.
7255 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7257 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7258 lookup on its own. Something else was needed to trigger the lookup. For
7259 example, a match in host_lookup or the need for a host name in a host list.
7260 Now, if $sender_host_name is referenced and the host name has not yet been
7261 looked up, a lookup is performed. If the lookup fails, the variable remains
7262 empty, and $host_lookup_failed is set to "1".
7264 19. Added "eqi" as a case-independent comparison operator.
7266 20. The saslauthd authentication condition could segfault if neither service
7267 nor realm was specified.
7269 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7270 error message that was logged was misleading, and incoming SMTP
7271 connections were dropped. The message is now more accurate, and temporary
7272 errors are given to SMTP connections.
7274 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7275 (except RSET) with a 421 error, until QUIT is received. However, it was
7276 failing to send a response to QUIT.
7278 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7279 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7280 ACL is now run after the helo_try_verify_hosts code.
7282 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7283 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7284 case-independent, so other case variants are also recognized). Apparently
7285 some systems use these upper case variants.
7287 25. If more than two messages were waiting for the same host, and a transport
7288 filter was specified for the transport, Exim sent two messages over the
7289 same TCP/IP connection, and then failed with "socket operation on non-
7290 socket" when it tried to send the third.
7292 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7294 27. The extern definition of crypt16() in expand.c was not being excluded when
7295 the OS had its own crypt16() function.
7297 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7298 as a preferred synonym for return_size_limit, both as an option and as an
7301 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7303 30. Changed the default configuration ACL to relax the local part checking rule
7304 for addresses that are not in any local domains. For these addresses,
7305 slashes and pipe symbols are allowed within local parts, but the sequence
7306 /../ is explicitly forbidden.
7308 31. SPA server authentication was not clearing the challenge buffer before
7311 32. log_message in a "warn" ACL statement was writing to the reject log as
7312 well as to the main log, which contradicts the documentation and doesn't
7313 seem right (because no rejection is happening). So I have stopped it.
7315 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7316 However, I am unable to do any testing of this.
7318 34. Fixed an infelicity in the appendfile transport. When checking directories
7319 for a mailbox, to see if any needed to be created, it was accidentally
7320 using path names with one or more superfluous leading slashes; tracing
7321 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7323 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7324 contents are added to the log line that is written for every discarded
7325 recipient. (Previously a log_message setting was ignored.)
7327 36. The ${quote: operator now quotes the string if it is empty.
7329 37. The install script runs exim in order to find its version number. If for
7330 some reason other than non-existence or emptiness, which it checks, it
7331 could not run './exim', it was installing it with an empty version number,
7332 i.e. as "exim-". This error state is now caught, and the installation is
7335 38. An argument was missing from the function that creates an error message
7336 when Exim fails to connect to the socket for saslauthd authentication.
7337 This could cause Exim to crash, or give a corrupted message.
7339 39. Added isip, isip4, and isip6 to ${if conditions.
7341 40. The ACL variables $acl_xx are now saved with the message, and can be
7342 accessed later in routers, transports, and filters.
7344 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7345 strings in the file are not string-expanded.
7347 42. If a MAIL command specified a SIZE value that was too large to fit into an
7348 int variable, the check against message_size_limit failed. Such values are
7349 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7350 day this will have to be increased, but I don't think I want to be around
7351 when emails are that large.
7358 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7359 iconv() is not standard in FreeBSD.
7361 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7362 IPv6 enabled. The observed symptom was a segmentation fault on return from
7363 the function os_common_find_running_interfaces() in src/os.c.
7365 3. In the check_special_case() function in daemon.c I had used "errno" as an
7366 argument name, which causes warnings on some systems. This was basically a
7367 typo, since it was named "eno" in the comments!
7369 4. The code that waits for the clock to tick (at a resolution of some fraction
7370 of a second) so as to ensure message-id uniqueness was always waiting for
7371 at least one whole tick, when it could have waited for less. [This is
7372 almost certainly not relevant at current processor speeds, where it is
7373 unlikely to ever wait at all. But we try to future-proof.]
7375 5. The function that sleeps for a time interval that includes fractions of a
7376 second contained a race. It did not block SIGALRM between setting the
7377 timer, and suspending (a couple of lines later). If the interval was short
7378 and the sigsuspend() was delayed until after it had expired, the suspension
7379 never ended. On busy systems this could lead to processes getting stuck for
7382 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7383 process, before it forks any delivery processes. The open lookup caching
7384 mechanism meant that the open file or database connection was passed into
7385 the delivery process. The problem was that delivery processes always tidy
7386 up cached lookup data. This could cause a problem for the next delivery
7387 process started by the queue runner, because the external queue runner
7388 process does not know about the closure. So the next delivery process
7389 still has data in the lookup cache. In the case of a file lookup, there was
7390 no problem because closing a file descriptor in a subprocess doesn't affect
7391 the parent. However, if the lookup was caching a connection to a database,
7392 the connection was closed, and the second delivery process was likely to
7393 see errors such as "PGSQL: query failed: server closed the connection
7394 unexpectedly". The problem has been fixed by closing all cached lookups
7395 in a queue runner before running a delivery process.
7397 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7398 seem to have the "const" qualifier which it has on other OS. I've
7401 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7402 *of the same type* (client or server) with the same public name that an
7403 error should be diagnosed.
7405 9. When Exim looked up a host name for an IP address, but failed to find the
7406 original IP address when looking up the host name (a safety check), it
7407 output the message "<ip address> does not match any IP for NULL", which was
7408 confusing, to say the least. The bug was that the host name should have
7409 appeared instead of "NULL".
7411 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7412 user that is built into the binary, and the -C or -D options is used, root
7413 privilege is dropped before the configuration file is read. In addition,
7414 logging is switched to stderr instead of the normal log files. If the
7415 configuration then re-defines the Exim user, the unprivileged environment
7416 is probably not what is expected, so Exim logs a panic warning message (but
7419 However, if deliver_drop_privilege is set, the unprivileged state may well
7420 be exactly what is intended, so the warning has been cut out in that case,
7421 and Exim is allowed to try to write to its normal log files.
7427 1. smtp_return_error_details was not giving details for temporary sender
7428 or receiver verification errors.
7430 2. Diagnose a configuration error if two authenticators have the same public
7433 3. Exim used not to create the message log file for a message until the first
7434 delivery attempt. This could be confusing when incoming messages were held
7435 for policy or load reasons. The message log file is now created at the time
7436 the message is received, and an initial "Received" line is written to it.
7438 4. The automatically generated man page for command line options had a minor
7439 bug that caused no ill effects; however, a more serious problem was that
7440 the procedure for building the man page automatically didn't always
7441 operate. Consequently, release 4.20 contains an out-of-date version. This
7442 shouldn't happen again.
7444 5. When building Exim with embedded Perl support, the script that builds the
7445 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7446 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7448 6. The freeze_tell option was not being used for messages that were frozen on
7449 arrival, either by an ACL or by local_scan().
7451 7. Added the smtp_incomplete_transaction log selector.
7453 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7454 was accepting AUTH without a new EHLO.
7456 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7457 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7460 10. Logging of TCP/IP connections (when configured) now happens in the main
7461 daemon process instead of the child process, so that the TCP/IP connection
7462 count is more accurate (but it can never be perfect).
7464 11. The use of "drop" in a nested ACL was not being handled correctly in the
7465 outer ACL. Now, if condition failure induced by the nested "drop" causes
7466 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7467 or "require"), the connection is dropped.
7469 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7470 that yield "discard" can now be used with an "accept" or a "discard" verb,
7471 but an error is generated for any others (because I can't see a useful way
7472 to define what should happen).
7474 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7475 are now processed in the same way as lines in the Exim configuration file.
7476 In particular, continuation lines are supported.
7478 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7480 15. Added -ti meaning -t -i.
7482 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7483 domains, and warn by logging if others are found.
7485 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7486 returned from GIFCONF when Exim is trying to find the list of interfaces on
7487 a host. The code in os.c has been modified to copy each ifreq to an aligned
7488 structure in all cases.
7490 Also, in some cases, the returned ifreq's were being copied to a 'struct
7491 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7492 means the last couple of bytes of an IPv6 address could be chopped if the
7493 ifreq contained only a normal sockaddr (14 bytes storage).
7495 18. Named domain lists were not supported in the hosts_treat_as_local option.
7496 An entry such as +xxxx was not recognized, and was treated as a literal
7499 19. Ensure that header lines added by a DATA ACL are included in the reject log
7500 if the ACL subsequently rejects the message.
7502 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7503 MD5 (which is deprecated).
7505 21. When testing a filter file using -bf, Exim was writing a message when it
7506 took the sender from a "From " line in the message, but it was not doing so
7507 when it took $return_path from a Return-Path: header line. It now does.
7509 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7510 with a valid header line field name (a series of printing characters
7511 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7513 23. Changed "disc" in the source to "disk" to conform to the documentation and
7514 the book and for uniformity.
7516 24. Ignore Sendmail's -Ooption=value command line item.
7518 25. When execve() failed while trying to run a command in a pipe transport,
7519 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7520 could be confused with a return value of 69 from the command itself. This
7521 has been changed to 127, the value the shell returns if it is asked to run
7522 a non-existent command. The wording for the related log line suggests a
7523 non-existent command as the problem.
7525 26. If received_header_text expands to an empty string, do not add a Received:
7526 header line to the message. (Well, it adds a token one on the spool, but
7527 marks it "old" so that it doesn't get used or transmitted.)
7529 27. Installed eximstats 1.28 (addition of -nt option).
7531 28. There was no check for failure on the call to getsockname() in the daemon
7532 code. This can fail if there is a shortage of resources on the system, with
7533 ENOMEM, for example. A temporary error is now given on failure.
7535 29. Contrary to the C standard, it seems that in some environments, the
7536 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7537 program. Exim now does this explicitly; it affects the formatting of
7538 timestamps using strftime().
7540 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7541 complaints. I've now initialized all the variables, to avoid this.
7543 32. Header lines added by a system filter were not being "seen" during
7544 transport-time rewrites.
7546 33. The info_callback() function passed to OpenSSL is set up with type void
7547 (*)(SSL *, int, int), as described somewhere. However, when calling the
7548 function (actually a macro) that sets it up, the type void(*)() is
7549 expected. I've put in a cast to prevent warnings from picky compilers.
7551 34. If a DNS black list lookup found a CNAME record, but there were no A
7552 records associated with the domain it pointed at, Exim crashed.
7554 35. If a DNS black list lookup returned more than one A record, Exim ignored
7555 all but the first. It now scans all returned addresses if a particular IP
7556 value is being sought. In this situation, the contents of the
7557 $dnslist_value variable are a list of all the addresses, separated by a
7560 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7561 to accept a host name and all its aliases if the forward lookup for any of
7562 them yielded the IP address of the incoming connection. Now it accepts only
7563 those names whose forward lookup yields the correct IP address. Any other
7564 names are discarded. This closes a loophole whereby a rogue DNS
7565 administrator could create reverse DNS records to break through a
7566 wildcarded host restriction in an ACL.
7568 37. If a user filter or a system filter that ran in a subprocess used any of
7569 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7570 the wrong values were passed to the pipe command ($thisaddress had the
7571 value of $0, $0 had the value of $1, etc). This bug was introduced by
7572 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7574 38. Improved the line breaking for long SMTP error messages from ACLs.
7575 Previously, if there was no break point between 40 and 75 characters, Exim
7576 left the rest of the message alone. Two changes have been made: (a) I've
7577 reduced the minimum length to 35 characters; (b) if it can't find a break
7578 point between 35 and 75 characters, it looks ahead and uses the first one
7579 that it finds. This may give the occasional overlong line, but at least the
7580 remaining text gets split now.
7582 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7583 file descriptors might be low, and that setting 1000 would always raise it.
7584 It turns out that in some environments, the limit is already over 1000 and
7585 that lowering it causes trouble. So now Exim takes care not to decrease it.
7587 40. When delivering a message, the value of $return_path is set to $sender_
7588 address at the start of routing (routers may change the value). By an
7589 oversight, this default was not being set up when an address was tested by
7590 -bt or -bv, which affected the outcome if any router or filter referred to
7593 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7594 log only when "message" or "log_message" are set. However, if one of the
7595 conditions was an address verification, or a call to a nested ACL, the
7596 messages generated by the underlying test were being passed through. This
7597 no longer happens. The underlying message is available in $acl_verify_
7598 message for both "message" and "log_message" expansions, so it can be
7599 passed through if needed.
7601 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7602 new expansion $bh_ to give the encoded byte string without charset
7603 translation. Translation happens only if iconv() is available; HAVE_ICONV
7604 indicates this at build time. HEADERS_CHARSET gives the charset to
7605 translate to; headers_charset can change it in the configuration, and
7606 "headers charset" can change it in an individual filter file.
7608 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7609 that creates RFC 2047 encoded "words" labels them as that charset instead
7610 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7611 expansion operator; (ii) when Exim creates a From: line for a local
7612 message; (iii) when a header line is rewritten to include a "phrase" part.
7614 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7615 buggy, causing it to skip the first lines of messages whose message ID
7616 ended in 'D'. This would not have bitten before Exim release 4.14, because
7617 message IDs were unlikely to end in 'D' before then. The effect was to have
7618 incorrect size information for certain domains.
7620 45. #include "config.h" was missing at the start of the crypt16.c module. This
7621 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7624 46. If there was a timeout during a "random" callout check, Exim treated it as
7625 a failure of the random address, and carried on sending RSET and the real
7626 address. If the delay was just some slowness somewhere, the response to the
7627 original RCPT would be taken as a response to RSET and so on, causing
7628 mayhem of various kinds.
7630 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7631 when I implemented it. It didn't allow for the fact that some option values
7632 may legitimately be negative (e.g. size_addition), and it didn't even do
7633 the right test for positive values.
7635 48. Domain names in DNS records are case-independent. Exim always looks them up
7636 in lower case. Some resolvers return domain names in exactly the case they
7637 appear in the zone file, that is, they may contain uppercase letters. Not
7638 all resolvers do this - some return always lower case. Exim was treating a
7639 change of case by a resolver as a change of domain, similar to a widening
7640 of a domain abbreviation. This triggered its re-routing code and so it was
7641 trying to route what was effectively the same domain again. This normally
7642 caused routing to fail (because the router wouldn't handle the domain
7643 twice). Now Exim checks for this case specially, and just changes the
7644 casing of the domain that it ultimately uses when it transmits the message
7647 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7650 50. If a filter generated a file delivery with a non-absolute name (possible if
7651 no home directory exists for the router), the forbid_file option was not
7654 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7655 the existing equality matching.
7657 52. Exim was using ints instead of ino_t variables in some places where it was
7658 dealing with inode numbers.
7660 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7661 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7662 called TMPDIR, and if it finds it is different, it changes its value.
7664 54. The smtp_printf() function is now made available to local_scan() so
7665 additional output lines can be written before returning. There is also an
7666 smtp_fflush() function to enable the detection of a dropped connection.
7667 The variables smtp_input and smtp_batched_input are exported to
7670 55. Changed the default runtime configuration: the message "Unknown user"
7671 has been removed from the ACL, and instead placed on the localuser router,
7672 using the cannot_route_message feature. This means that any verification
7673 failures that generate their own messages won't get overridden. Similarly,
7674 the "Unrouteable address" message that was in the ACL for unverifiable
7675 relay addresses has also been removed.
7677 56. Added hosts_avoid_esmtp to the smtp transport.
7679 57. The exicyclog script was not checking for the esoteric option
7680 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7681 will work only if exicyclog is run under the appropriate euid.
7683 58. Following a discussion on the list, the rules by which Exim recognises line
7684 endings on incoming messages have been changed. The -dropcr and drop_cr
7685 options are now no-ops, retained only for backwards compatibility. The
7686 following line terminators are recognized: LF CRLF CR. However, special
7687 processing applies to CR:
7689 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7690 nor a local message in the state where . is a terminator.
7692 (ii) If a bare CR is encountered in a header line, an extra space is added
7693 after the line terminator so as not to end the header. The reasoning
7694 behind this is that bare CRs in header lines are most likely either
7695 to be mistakes, or people trying to play silly games.
7697 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7698 was being incorrectly given as 18 bytes larger than it should have been.
7699 This is a VOB (very old bug).
7701 60. This may never have affected anything current, but just in case it has:
7702 When the local host is found other than at the start of a list of hosts,
7703 the local host, those with the same MX, and any that follow, are discarded.
7704 When the list in question was part of a longer list of hosts, the following
7705 hosts (not currently being processed) were also being discarded. This no
7706 longer happens. I'm not sure if this situation could ever has previously
7709 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7712 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7713 header. This is something that is recommended in a new Internet Draft, and
7714 is something that is documented as being done by Sendmail. There are two
7715 possible values. For messages generated by the autoreply transport, Exim
7718 Auto-Submitted: auto-replied
7720 whereas for all other generated messages (e.g. bounces) it adds
7722 Auto-Submitted: auto-generated
7724 63. The "personal" condition in filters now includes a test for the
7725 Auto-Submitted: header. If it contains the string "auto-" the message it
7726 not considered personal.
7728 64. Added rcpt_include_affixes as a generic transport option.
7730 65. Added queue_only_override (default true).
7732 66. Added the syslog_duplication option.
7734 67. If what should have been the first header line of a message consisted of
7735 a space followed by a colon, Exim was mis-interpreting it as a header line.
7736 It isn't of course - it is syntactically invalid and should therefore be
7737 treated as the start of the message body. The misbehaviour could have
7738 caused a number of strange effects, including loss of data in subsequent
7739 header lines, and spool format errors.
7741 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7742 client host had authenticated. This control can now be exercised by an ACL
7743 for more flexibility.
7745 69. By default, callouts do not happen when testing with -bh. There is now a
7746 variant, -bhc, which does actually run the callout code, including
7747 consulting and updating the callout cache.
7749 70. Added support for saslauthd authentication, courtesy of Alexander
7752 71. If statvfs() failed on the spool or log directories while checking their
7753 size for availability, Exim confusingly gave the error "space shortage".
7754 Furthermore, in debugging mode it crashed with a floating point exception.
7755 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7756 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7757 serious problem, Exim now writes to the main and panic logs when this
7758 happens, with details of the failure. It then refuses to accept the
7759 incoming message, giving the message "spool directory problem" or "log
7760 directory problem" with a 421 code for SMTP messages.
7762 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7763 0, 1, and 2 exist, because some OS complain for execs without them (see
7764 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7765 descriptors. However, the code omitted to check that the open succeeded,
7766 causing mysterious errors if for some reason the permissions on /dev/null
7767 got screwed. Now Exim writes a message to the main and panic logs, and
7768 bombs out if it can't open /dev/null.
7770 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7771 interact so that it is all more flexible. It is supposed to remain
7772 backwards compatible. Also added extra_local_interfaces.
7774 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7775 to bomb out with an assertion failure - to the client this appears as a
7776 connection drop. This problem occurs in the part of the code that was taken
7777 from the Samba project. Fortunately, the assertion is in a very simple
7778 function, so I have fixed this by reproducing the function inline in the
7779 one place where it is called, and arranging for authentication to fail
7780 instead of killing the process with assert().
7782 75. The SPA client code was not working when the server requested OEM rather
7783 than Unicode encoding.
7785 76. Added code to make require_files with a specific uid setting more usable in
7786 the case where statting the file as root fails - usually a non-root-mounted
7787 NFS file system. When this happens and the failure is EACCES, Exim now
7788 forks a subprocess and does the per-uid checking as the relevant uid.
7790 77. Added process_log_path.
7792 78. If log_file_path was not explicitly set, a setting of check_log_space or
7793 check_log_inodes was ignored.
7795 79. If a space check for the spool or log partitions fails, the incident is now
7796 logged. Of course, in the latter case the data may get lost...
7798 80. Added the %p formatting code to string_format() so that it can be used to
7799 print addresses in debug_print(). Adjusted all the address printing in the
7800 debugging in store.c to use %p rather than %d.
7802 81. There was a concern that a line of code in smtp_in.c could overflow a
7803 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7804 initially expressed, the concern was not well-founded, because trailing
7805 spaces are removed early. However, if the trailing spaces were followed by
7806 a NULL, they did not get removed, so the overflow was possible. Two fixes
7809 (a) I re-wrote the offending code in a cleaner fashion.
7810 (b) If an incoming SMTP command contains a NULL character, it is rejected
7813 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7814 runs initgroups(), so that if the Exim user is in any additional groups,
7815 they will be used during message reception.
7821 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob