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
16 created buffers. Similar fix for radius expansion condition.
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.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
56 JH/01 Avoid costly startup code when not strictly needed. This reduces time
57 for some exim process initialisations. It does mean that the logging
58 of TLS configuration problems is only done for the daemon startup.
60 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
62 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
63 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
66 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
67 DKIM signing is being done. Previously a transport_filter would always
68 disable CHUNKING, falling back to traditional DATA.
70 JH/05 Regard command-line receipients as tainted.
72 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
74 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
75 PAM library frees one of the arguments given to it, despite the
76 documentation. Therefore a plain malloc must be used.
78 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
79 on-stack buffers were used, resulting in a taint trap when DSN information
80 copied from a received message was written into the buffer.
82 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
83 the ordering of its ARC headers. This caused a crash.
85 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
86 a new record was being constructed with information from the peer, a trap
89 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
90 installation would get error messages from DMARC verify, when it hit the
91 nonexistent file indicated by the default. Distros wanting DMARC enabled
92 should both provide the file and set the option.
93 Also enforce no DMARC verification for command-line sourced messages.
95 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
96 could, depending on the platform, hang at the STARTTLS response.
98 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
99 message on a connection. Previously if one message had ARC headers and
100 the following one did not, a crash could result when adding an
101 Authentication-Results: header.
103 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
104 Exim main code and Exim-related utities. The introduction of taint
105 tracking also did many adjustments to string handling. Since then, eximon
106 frequently terminated with an assert failure.
108 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
109 check for 452 responses. This slightly helps the inefficieny of doing
110 a large alias-expansion into a recipient-limited target. The max_rcpt
111 transport option still applies (and at the current default, will override
112 the new feature). The check is done for either cause of synch, and forces
113 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
114 connection. The new facility is not tunable at this time.
116 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
117 library live data was being used, so the results became garbage. Make
118 copies while it is still usable.
120 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
121 on delivery deferred (==) and failed (**) lines (if a delivery was
122 attemtped). Previously it was only on completion (=>) lines.
124 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
125 for the expansion of the server_scram_iter and server_scram_salt options.
127 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
128 are now specifically given a NO_DATA response without hitting the system
129 resolver. The library goes on to do the now-standard TXT lookup.
130 Use of dnsdb lookups is not affected.
132 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
133 only retrieve the errormessage once. Previously two calls to dlerror()
134 were used, and the second one (for mainlog/paniclog) retrieved null
137 JH/20 Taint checking: disallow use of tainted data for
138 - the appendfile transport file and directory options
139 - the pipe transport command
140 - the autoreply transport file, log and once options
141 - file names used by the redirect router (including filter files)
143 - paths used by single-key lookups
144 Previously this was permitted.
146 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
147 adjusted the size of a major service buffer; this failed because the
148 buffer was in use at the time. Change to a compile-time increase in the
149 buffer size, when this authenticator is compiled into exim.
151 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
152 previous fast-mode was untenable in the face of glibs using mmap to
153 support larger malloc requests.
155 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
156 New values supported, if defined on system where compiled:
157 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
158 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
160 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
161 running a limited number of proceses in parallel, a benefit is gained. The
162 amount varies with the platform hardware and load. The use of the option
163 queue_run_in_order means we cannot do this, as ordering becomes
166 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
167 had introduced a string-copy (for ensuring NUL-termination) which was not
168 appropriate for that case, which can include embedded NUL bytes in the
169 block of data. Investigation showed the copy to actually be needless, the
170 data being length-specified.
172 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
173 done during a receiving connection, and both used TLS, global info was
174 used rather than per-connection info for tracking the state of data
175 queued for transmission. This could result in a connection hang.
177 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
178 Previously, when delivering serveral messages down a single connection
179 only the first would provide a SIZE. This was due to the size information
180 not being properly tracked.
182 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
183 TAI (at 37 seconds currently), pretend to be in UTC for time-related
184 expansion and logging. Previously, spurious values such as a future
185 minute could be seen.
187 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
188 it could crash from a null-deref. This could also affect the
189 ${addresses: } operator and ${readsock } item.
191 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
192 message following a mime one, the variable was not reset.
194 JH/30 When an pipelined-connect fails at the first response, assume incorrect
195 cached capability (perhaps the peer reneged?) and immediately retry in
198 JH/31 Fix spurious detection of timeout while writing to transport filter.
200 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
201 an attempt to copy the string was made before checking it.
203 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
204 taint of the lookup key was maintained; we now regard the presence in the
205 filesystem as sufficient validation.
207 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
208 argument is supplied.
210 JH/35 The dsearch lookup now requires that the directory is an absolute path.
211 Previously this was not checked, and nonempty relative paths made an
212 access under Exim's current working directory.
214 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
215 Previously no event was raised.
217 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
218 parameter supplied by the sender MAIL FROM command. Previously it was
219 ignored, and only the check_spool_space option value for the required
222 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
223 the size of the signing public-key. Previously it was instead giving
224 the size of the signature hash.
226 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
227 the default. See the (new) dkim_verify_min_keysizes option.
229 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
230 and an ACL use a lookup for checking either the local_part or domain,
231 stale data could be accessed. Ensure that variable references are
232 dropped between messages.
234 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
235 by the client was not checked as pointing within response data before
236 being used. A malicious client could thus cause an out-of-bounds read and
237 possibly gain authentication. Fix by adding the check.
239 JH/42 Internationalisation: change the default for downconversion in the smtp
240 transport to be "if needed". Previously it was "as previously set" for
241 the message, which usually meant "if needed" for message-submission but
242 "no" for everything else. However, MTAs have been seen using SMTPUTF8
243 even when the envelope addresses did not need it, resulting in forwarding
244 failures to non-supporting MTAs. A downconvert in such cases will be
245 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
246 transport. The change does mean that addresses needing conversion will
247 be converted when previously a delivery failure would occur.
249 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
250 response was received it would be used unchecked in a fail-DSN, violating
251 standards on line-length limits. Truncate if needed.
253 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
254 doesn't work. (Additionally add a new main config option to configure the
261 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
262 side, to decode a TLS 1.3 packet capture.
264 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
265 Previously the default library behaviour applied, sending two, each in
268 JH/03 Debug output for ACL now gives the config file name and line number for
271 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
273 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
275 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
276 buffer overrun for (non-chunking) other transports.
278 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
279 TLS1.3, means that a server rejecting a client certificate is not visible
280 to the client until the first read of encrypted data (typically the
281 response to EHLO). Add detection for that case and treat it as a failed
282 TLS connection attempt, so that the normal retry-in-clear can work (if
283 suitably configured).
285 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
286 and/or domain. Found and fixed by Jason Betts.
288 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
289 configuration). If a CNAME target was not a wellformed name pattern, a
292 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
293 the OS reports them interleaved with other addresses.
295 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
296 used both for input and for a verify callout, both encrypted, SMTP
297 responses being sent by the server could be lost. This resulted in
298 dropped connections and sometimes bounces generated by a peer sending
301 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
302 string. Previously it was possible to cause undefined behaviour in a
303 library routine (usually a crash). Found by "zerons".
305 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
308 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
309 API was removed, so update to use the newer ones.
311 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
312 any timeout set, is taking a long time. Previously we would hang on to a
313 rotated logfile "forever" if the input was arriving with long gaps
314 (a previous attempt to fix addressed lack, for a long time, of initial
317 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
318 shared (NFS) environment. The length of the tempfile name is now
319 4 + 16 ("hdr.$message_exim_id") which might break on file
320 systems which restrict the file name length to lower values.
323 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
324 shared (NFS) environment.
326 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
327 did for all versions <4.90). Notably -M, -m, --invert, -I may be
330 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
331 on some platforms for bit 31.
333 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
334 to changes apparently associated with TLS1.3 handling some of the APIs
335 previously used were either nonfunctional or inappropriate. Strings
336 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
337 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
338 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
339 This affects log line X= elements, the $tls_{in,out}_cipher variables,
340 and the use of specific cipher names in the encrypted= ACL condition.
342 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
344 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
345 verification result was not updated unless hosts_require_ocsp applied.
347 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
348 queue_list_requires_admin set to false, non-admin users were denied the
351 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
352 directory-of-certs mode. Previously they were advertised despite the
355 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
356 A single TCP connection by a client will now hold a TLS connection open
357 for multiple message deliveries, by default. Previously the default was to
360 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
361 default. If built with the facility, DANE will be used. The facility
362 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
364 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
365 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
366 must be defined and you must still, unless you define DISABLE_TLS, manage
367 the the include-dir and library-file requirements that go with that
368 choice. Non-TLS builds are still supported.
370 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
373 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
374 default. If the platform supports and has the facility enabled, it will
375 be requested on all coneections.
377 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
378 controlled by the build-time option SUPPORT_PIPE_CONNECT.
380 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
382 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
383 success-DSN messages. Previously the From: header was always the default
384 one for these; the option was ignored.
386 JH/28 Fix the timeout on smtp response to apply to the whole response.
387 Previously it was reset for every read, so a teergrubing peer sending
388 single bytes within the time limit could extend the connection for a
389 long time. Credit to Qualsys Security Advisory Team for the discovery.
391 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
392 delivery address, which leaked information of the results of local
393 forwarding. Change to the original envelope recipient address, per
396 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
397 requested. Previously not bounce was generated and a log entry of
398 error ignored was made.
400 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
402 JH/32 Introduce a general tainting mechanism for values read from the input
403 channel, and values derived from them. Refuse to expand any tainted
404 values, to catch one form of exploit.
406 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
407 was unused and the unexpanded text used for the test. Found and
408 fixed by Ruben Jenster.
410 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
411 an attempt to use a TLS library read routine dereffed a nul pointer,
414 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
415 them in our smtp response.
417 JH/36 Have the general router option retry_use_local_part default to true when
418 any of the restrictive preconditions are set (to anything). Previously it
419 was only for check_local user. The change removes one item of manual
420 configuration which is required for proper retries when a remote router
421 handles a subset of addresses for a domain.
423 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
424 link count into consideration.
426 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
427 caused the extension of big_buffer, the following lines were ignored.
429 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
430 accordance with RFC 2308. Previously there was no expiry, so a longlived
431 receive process (eg. due to ACL delays) versus a short SOA value could
434 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
436 JH/39 Promote DMARC support to mainline.
438 JH/40 Bug 2452: Add a References: header to DSNs.
440 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
441 parameters. The relevant library call is documented as "Deprecated: This
442 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
443 3.6.0, DH parameters are negotiated following RFC7919."
445 HS/06 Change the default of dnssec_request_domains to "*"
447 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
448 carried on and emitted a BDAT command, even when PIPELINING was not
451 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
452 buffer was used for the filename, resulting in a trap when tainted
453 arguments (eg. $domain) were used.
455 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
456 recommended to avoid a possible server-load attack. The feature can be
457 re-enabled via the openssl_options main cofiguration option.
459 JH/45 local_scan API: documented the current smtp_printf() call. This changed
460 for version 4.90 - adding a "more data" boolean to the arguments.
461 Bumped the ABI version number also, this having been missed previously;
462 release versions 4.90 to 4.92.3 inclusive were effectively broken in
463 respect of usage of smtp_printf() by either local_scan code or libraries
464 accessed via the ${dlfunc } expansion item. Both will need coding
465 adjustment for any calls to smtp_printf() to match the new function
466 signature; a FALSE value for the new argument is always safe.
468 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
469 the file-offset (which the Linux syscall does, and exim expects); this
470 resulted in an indefinite loop.
472 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
473 to do ARC verification. The Authentication-Results: header line added
474 by the configuration then had no ARC item.
480 JH/01 Remove code calling the customisable local_scan function, unless a new
481 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
483 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
484 non-signal-safe functions being used.
486 JH/03 Bug 2269: When presented with a received message having a stupidly large
487 number of DKIM-Signature headers, disable DKIM verification to avoid
488 a resource-consumption attack. The limit is set at twenty.
490 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
491 report of oldest_pass in ${authres } in consequence, and separate out
492 some descriptions of reasons for verification fail.
494 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
495 files in the spool were present and unlocked. A queue-runner could spot
496 them, resulting in a duplicate delivery. Fix that by doing the unlock
497 after the unlink. Investigation by Tim Stewart. Take the opportunity to
498 add more error-checking on spoolfile handling while that code is being
501 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
502 No known attacks, no CVE, this is defensive hardening.
504 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
505 a queue-runner could start a delivery while other operations were ongoing.
506 Cutthrough delivery was a common victim, resulting in duplicate delivery.
507 Found and investigated by Tim Stewart. Fix by using the open message data
508 file handle rather than opening another, and not locally closing it (which
509 releases a lock) for that case, while creating the temporary .eml format
510 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
512 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
513 $sender_verify_failure/$recipient_verify_failure to "random".
515 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
518 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
519 Previously this would segfault.
521 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
524 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
525 like zero, since the resolver should be doing this for us, But we need one
526 as a CNAME but no MX presence gets the CNAME returned; we need to check
527 that doesn't point to an MX to declare it "no MX returned" rather than
528 "error, loop". A new main option is added so the older capability of
529 following some limited number of chain links is maintained.
531 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
533 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
534 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
535 now-deprecated earlier definitions used only specified the range up to TLS
536 1.2 (in the older-version library docs).
538 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
540 JH/15 Rework TLS client-side context management. Stop using a global, and
541 explicitly pass a context around. This enables future use of TLS for
542 connections to service-daemons (eg. malware scanning) while a client smtp
543 connection is using TLS; with cutthrough connections this is quite likely.
545 JH/16 Fix ARC verification to do AS checks in reverse order.
547 JH/17 Support a "tls" option on the ${readsocket } expansion item.
549 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
550 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
551 Previously the "utf8" would be re-prepended for every additional message.
553 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
554 Previously thery were accepted, resulting in issues when attempting to
555 forward messages to a non-supporting MTA.
557 PP/02 Let -n work with printing macros too, not just options.
559 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
560 one parent address was copied, and bogus data was used at delivery-logging
561 time. Either a crash (after delivery) or bogus log data could result.
562 Discovery and analysis by Tim Stewart.
564 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
565 Previously if the string ended mid-character, we did not insert the
566 promised '?' replacement.
568 PP/04 Documentation: current string operators work on bytes, not codepoints.
570 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
571 should pack well giving a smaller memory footprint so better caching and
572 therefore performance. Group the declarations where this can't be done so
573 that the byte-sized flag variables are not interspersed among pointer
574 variables, giving a better chance of good packing by the compiler.
576 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
577 non-null, to avoid issues with sites running BATV. Previously reports were
578 sent with an empty envelope sender so looked like bounces.
580 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
581 The ignore_error flag wasn't being returned from the filter subprocess so
582 was not set for later routers. Investigation and fix by Matthias Kurz.
584 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
585 and a msg:complete for the whole, when a message is manually removed using
586 -Mrm. Developement by Matthias Kurz, hacked on by JH.
588 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
589 a "Gnu special" function, asprintf() in the DB utility binary builds; I
590 hope that is portable enough.
592 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
593 requiring a known-CA anchor certificate; make it now rely entirely on the
594 TLSA as an anchor. Checking the name on the leaf cert against the name
595 on the A-record for the host is still done for TA (but not for EE mode).
597 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
598 included in delivery lines for non-proxied connections, when compiled with
599 SUPPORT_SOCKS and running with proxy logging enabled.
601 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
602 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
603 move the existing event to fire before the normal logging of message
604 failure so that custom logging is bracketed by normal logging.
606 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
607 msg:fail:internal event. Developement by Matthias Kurz.
609 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
610 far too small for todays use of crypto signatures stored there. Go all
611 the way to the max DNS message size of 64kB, even though this might be
612 overmuch for IOT constrained device use.
614 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
615 copy a string over itself. The library routine is documented as not
616 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
618 JH/32 For main options check_spool_space and check_inode_space, where the
619 platform supports 64b integers, support more than the previous 2^31 kB
620 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
621 the previous G, M, k.
623 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
624 $authenticated_fail_id variable on authentication failure. Previously
627 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
628 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
629 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
630 GNUTLS_SEC_PARAM_MEDIUM.
632 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
633 side. Previously we would continue as if no SNI had been received.
635 JH/36 Harden the handling of string-lists. When a list consisted of a sole
636 "<" character, which should be a list-separator specification, we walked
637 off past the nul-terimation.
639 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
640 causes) even when the retry time is not yet met. Previously they were
641 not, meaning that when (say) an account was over-quota and temp-rejecting,
642 and multiple senders' messages were queued, only one sender would get
643 notified on each configured delay_warning cycle.
645 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
647 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
648 cipher-suites, an error can be left on the stack even for a succeeding
649 accept; this results in impossible error messages when a later operation
652 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
653 return error codes indicating retry. Under TLS1.3 this becomes required.
655 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
656 it only wrote the new authenticators, resulting in a lack of tracking of
657 peer changes of ESMTP extensions until the next cache flush.
659 JH/41 Fix the loop reading a message header line to check for integer overflow,
660 and more-often against header_maxsize. Previously a crafted message could
661 induce a crash of the recive process; now the message is cleanly rejected.
663 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
664 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
670 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
671 When redis_servers is set to a list of > 1 element, and the Redis servers
672 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
673 case of MOVED into a DEFER case instead, thus moving the query onto the
674 next server in the list. For a cluster of N elements, all N servers must
675 be defined in redis_servers.
677 GF/02 Catch and remove uninitialized value warning in exiqsumm
678 Check for existence of @ARGV before looking at $ARGV[0]
680 JH/01 Replace the store_release() internal interface with store_newblock(),
681 which internalises the check required to safely use the old one, plus
682 the allocate and data copy operations duplicated in both (!) of the
683 extant use locations.
685 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
686 modifier. This matches the restriction on the commandline.
688 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
689 Previously only the last row was returned.
691 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
692 we assumed that tags in the header were well-formed, and parsed the
693 element content after inspecting only the first char of the tag.
694 Assumptions at that stage could crash the receive process on malformed
697 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
698 While running the DKIM ACL we operate on the Permanent memory pool so that
699 variables created with "set" persist to the DATA ACL. Also (at any time)
700 DNS lookups that fail create cache records using the Permanent pool. But
701 expansions release any allocations made on the current pool - so a dnsdb
702 lookup expansion done in the DKIM ACL releases the memory used for the
703 DNS negative-cache, and bad things result. Solution is to switch to the
704 Main pool for expansions.
705 While we're in that code, add checks on the DNS cache during store_reset,
706 active in the testsuite.
707 Problem spotted, and debugging aided, by Wolfgang Breyha.
709 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
710 When none of the hosts presented to a transport match an already-open
711 connection, close it and proceed with the list. Previously we would
712 queue the message. Spotted by Lena with Yahoo, probably involving
715 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
716 Previously a spurious "250 OK id=" response was appended to the proper
719 JH/08 The "support for" informational output now, which built with Content
720 Scanning support, has a line for the malware scanner interfaces compiled
721 in. Interface can be individually included or not at build time.
723 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
724 by the template makefile "src/EDITME". The "STREAM" support for an older
725 ClamAV interface method is removed.
727 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
728 rows affected is given instead).
730 JH/11 The runtime Berkeley DB library version is now additionally output by
731 "exim -d -bV". Previously only the compile-time version was shown.
733 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
734 SMTP connection. Previously, when one had more receipients than the
735 first, an abortive onward connection was made. Move to full support for
736 multiple onward connections in sequence, handling cutthrough connection
737 for all multi-message initiating connections.
739 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
740 routers. Previously, a multi-recipient message would fail to match the
741 onward-connection opened for the first recipient, and cause its closure.
743 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
744 a timeout on read on a GnuTLS initiating connection, resulting in the
745 initiating connection being dropped. This mattered most when the callout
746 was marked defer_ok. Fix to keep the two timeout-detection methods
749 JH/15 Relax results from ACL control request to enable cutthrough, in
750 unsupported situations, from error to silently (except under debug)
751 ignoring. This covers use with PRDR, frozen messages, queue-only and
754 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
756 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
757 metadata, resulting in a crash in free().
759 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
760 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
761 Broken also in d185889f4, with init system revamp.
763 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
764 Previously we abruptly closed the connection after reading a malware-
765 found indication; now we go on to read the "scan ok" response line,
768 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
769 ACL. Previously, a crash would result.
771 JH/19 Speed up macro lookups during configuration file read, by skipping non-
772 macro text after a replacement (previously it was only once per line) and
773 by skipping builtin macros when searching for an uppercase lead character.
775 JH/20 DANE support moved from Experimental to mainline. The Makefile control
776 for the build is renamed.
778 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
779 was allocated for every new TLS startup, meaning one per message. Fix
780 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
782 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
783 reported the original. Fix to report (as far as possible) the ACL
784 result replacing the original.
786 JH/23 Fix memory leak during multi-message connections using STARTTLS under
787 OpenSSL. Certificate information is loaded for every new TLS startup,
788 and the resources needed to be freed.
790 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
792 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
795 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
796 DATA response info to the (existing) per-recipient response info for
797 the "C=" log element. It can have useful tracking info from the
798 destination system. Patch from Simon Arlott.
800 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
801 length value. Previously this would segfault.
803 HS/02 Support Avast multiline protoocol, this allows passing flags to
804 newer versions of the scanner.
806 JH/28 Ensure that variables possibly set during message acceptance are marked
807 dead before release of memory in the daemon loop. This stops complaints
808 about them when the debug_store option is enabled. Discovered specifically
809 for sender_rate_period, but applies to a whole set of variables.
810 Do the same for the queue-runner and queue-list loops, for variables set
811 from spool message files. Do the same for the SMTP per-message loop, for
812 certain variables indirectly set in ACL operations.
814 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
815 as a multi-recipient message from a mailinglist manager). The coding had
816 an arbitrary cutoff number of characters while checking for more input;
817 enforced by writing a NUL into the buffer. This corrupted long / fast
818 input. The problem was exposed more widely when more pipelineing of SMTP
819 responses was introduced, and one Exim system was feeding another.
820 The symptom is log complaints of SMTP syntax error (NUL chars) on the
821 receiving system, and refused recipients seen by the sending system
822 (propating to people being dropped from mailing lists).
823 Discovered and pinpointed by David Carter.
825 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
826 replaced by the ${authresults } expansion.
828 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
830 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
831 allows proper process termination in container environments.
833 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
834 Previously the "final dot" had a newline after it; ensure it is CR,LF.
836 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
837 and "err_perm", deprecated since 4.83 when the RFC-defined words
838 "temperror" and "permerror" were introduced.
840 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
841 transport-filters or DKIM-signing. The restriction was lost in the
842 consolidation of verify-callout and delivery SMTP handling.
843 Extend the restriction to also cover ARC-signing.
845 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
846 in defer=pass mode supply a 450 to the initiator. Previously the message
849 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
850 tls_require_ciphers is used as before.
852 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
853 "pass_unscanned". Only tmpfails from the scanner are written to
854 the paniclog, as they may require admin intervention (permission
855 denied, license issues). Other scanner errors (like decompression
856 bombs) do not cause a paniclog entry.
858 JH/36 Fix reinitialisation of DKIM logging variable between messages.
859 Previously it was possible to log spurious information in receive log
862 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
863 triggered odd behaviour from Outlook Express clients.
865 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
868 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
869 since the IETF WG has not yet settled on that versus the original
870 "bare" representation.
872 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
873 Previously the millisecond value corrupted the output.
874 Fix also for syslog_pid=no and log_selector +pid, for which the pid
875 corrupted the output.
881 JH/01 Rework error string handling in TLS interface so that the caller in
882 more cases is responsible for logging. This permits library-sourced
883 string to be attached to addresses during delivery, and collapses
884 pairs of long lines into single ones.
886 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
887 during configuration. Wildcards are allowed and expanded.
889 JH/02 Rework error string handling in DKIM to pass more info back to callers.
890 This permits better logging.
892 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
893 do not close it down and have the child transport start it up again on
894 the passed-on TCP connection. Instead, proxy the child (and any
895 subsequent ones) for TLS via a unix-domain socket channel. Logging is
896 affected: the continued delivery log lines do not have any DNSSEC, TLS
897 Certificate or OCSP information. TLS cipher information is still logged.
899 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
900 identical IP addresses on different listening ports. Will also affect
903 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
904 add noisy ifdef guards to special-case this sillyness.
905 Patch from Bernd Kuhls.
907 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
908 than 255 are no longer allowed.
910 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
911 Disable the session-cache too, which might reduce our load. Since we
912 currrectly use a new context for every connection, both as server and
913 client, there is no benefit for these.
914 GnuTLS appears to not support tickets server-side by default (we don't
915 call gnutls_session_ticket_enable_server()) but client side is enabled
916 by default on recent versions (3.1.3 +) unless the PFS priority string
919 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
920 <https://reproducible-builds.org/specs/source-date-epoch/>.
922 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
923 the check for any unsuccessful recipients did not notice the limit, and
924 erroneously found still-pending ones.
926 JH/08 Pipeline CHUNKING command and data together, on kernels that support
927 MSG_MORE. Only in-clear (not on TLS connections).
929 JH/09 Avoid using a temporary file during transport using dkim. Unless a
930 transport-filter is involved we can buffer the headers in memory for
931 creating the signature, and read the spool data file once for the
932 signature and again for transmission.
934 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
935 4.77 as the kernel support then wasn't solid, having issues in 64bit
936 mode. Now, it's been long enough. Add support for FreeBSD also.
938 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
939 case where the routing stage had gathered several addresses to send to
940 a host before calling the transport for the first, we previously failed
941 to close down TLS in the old transport process before passing the TCP
942 connection to the new process. The new one sent a STARTTLS command
943 which naturally failed, giving a failed delivery and bloating the retry
944 database. Investigation and fix prototype from Wolfgang Breyha.
946 JH/12 Fix check on SMTP command input synchronisation. Previously there were
947 false-negatives in the check that the sender had not preempted a response
948 or prompt from Exim (running as a server), due to that code's lack of
949 awareness of the SMTP input buffering.
951 PP/04 Add commandline_checks_require_admin option.
952 Exim drops privileges sanely, various checks such as -be aren't a
953 security problem, as long as you trust local users with access to their
954 own account. When invoked by services which pass untrusted data to
955 Exim, this might be an issue. Set this option in main configuration
956 AND make fixes to the calling application, such as using `--` to stop
959 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
960 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
961 responses to those, into a single TLS record each way (this usually means
962 a single packet). As a side issue, smtp_enforce_sync now works on TLS
965 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
966 affects you only if you're dancing at the edge of the param size limits.
967 If you are, and this message makes sense to you, then: raise the
968 configured limit or use OpenSSL 1.1. Nothing we can do for older
971 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
972 cmdline element to get the documented default one. Previously it was
975 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
976 are unsuable for non-message-specific reasons - eg. connection timeout,
977 banner-time rejection.
979 JH/16 Fix logging of delivery remote port, when specified by router, under
982 PP/06 Repair manualroute's ability to take options in any order, even if one
983 is the name of a transport.
986 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
988 JH/17 Change the list-building routines interface to use the expanding-string
989 triplet model, for better allocation and copying behaviour.
991 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
992 Previously it was constructed the first time a possibly-matching string
993 was met in the configuration file input during startup; now it is done
996 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
997 compatible one, to avoid the (poorly documented) possibility of a config
998 file in the working directory redirecting the DB files, possibly correpting
999 some existing file. CVE-2017-10140 assigned for BDB.
1001 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1002 cache-hot. Previously, although the result was properly cached, the
1003 initial verify call returned a defer.
1005 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1006 the main verify for receipient in uncached-mode.
1008 JH/22 Retire historical build files to an "unsupported" subdir. These are
1009 defined as "ones for which we have no current evidence of testing".
1011 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1012 if present. Previously it was ignored.
1014 JH/24 Start using specified-initialisers in C structure init coding. This is
1015 a C99 feature (it's 2017, so now considered safe).
1017 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1018 if was a fixed-sized field and bitmask ops via macros; it is now more
1021 PP/07 GitHub PR 56: Apply MariaDB build fix.
1022 Patch provided by Jaroslav Škarvada.
1024 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1025 during Coverity cleanups [4.87 JH/47]
1026 Diagnosis and fix provided by Michael Fischer v. Mollard.
1028 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1029 the right size to place the terminating semicolon on its own folded
1030 line, the header hash was calculated to an incorrect value thanks to
1031 the (relaxed) space the fold became.
1033 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1034 and confused the parent.
1036 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1037 which could crash as a result. This could lead to undeliverable messages.
1039 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1042 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1043 matters on fast-turnover and PID-randomising systems, which were getting
1044 out-of-order delivery.
1046 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1047 a possibly-overlapping copy. The symptom was that "Remote host closed
1048 connection in response to HELO" was logged instead of the actual 4xx
1051 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1052 Previously only that bufferd was discarded, resulting in SYMTP command
1055 JH/32 DKIM: when a message has multiple signatures matching an identity given
1056 in dkim_verify_signers, run the dkim acl once for each. Previously only
1057 one run was done. Bug 2189.
1059 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1060 "panic the current process" to "deliberately defer". The panic log is
1061 still written with the problem list name; the mail and reject logs now
1062 get a temp-reject line for the message that was being handled, saying
1063 something like "domains check lookup or other defer". The SMTP 451
1064 message is still "Temporary local problem".
1066 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1067 A crafted sequence of BDAT commands could result in in-use memory beeing
1068 freed. CVE-2017-16943.
1070 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1071 from SMTP input. Previously it was always done; now only done for DATA
1072 and not BDAT commands. CVE-2017-16944.
1074 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1075 to the message (such as an overlong header line). Previously this was
1076 not done and we did not exit BDAT mode. Followon from the previous item
1077 though a different problem.
1083 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1084 than -2003 did; needs libidn2 in addition to libidn.
1086 JH/02 The path option on a pipe transport is now expanded before use.
1088 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1089 Patch provided by "Björn", documentation fix added too.
1091 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1092 missing a wire-to-host endian conversion.
1094 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1095 close after a BDAT command line could be taken as a following command,
1096 giving a synch failure. Fix by only checking for synch immediately
1097 before acknowledging the chunk.
1099 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1100 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1101 macro. Patches provided by Josh Soref.
1103 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1104 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1105 by IANA as a service extension. However, John Klensin suggests that we
1108 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1109 the dkim code may be unix-mode line endings rather than smtp wire-format
1110 CRLF, so prepend a CR to any bare LF.
1112 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1113 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1115 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1116 the first were themselves being wrongly included in the feed into dkim
1117 processing; with most chunk sizes in use this resulted in an incorrect
1118 body hash calculated value.
1120 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1121 DKIM signature block, for verification. Although advised against by
1122 standards it is specifically not ruled illegal.
1124 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1126 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1127 missing a body hash (the bh= tag).
1129 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1130 It seems that HAProxy sends the Proxy Protocol information in clear and
1131 only then does a TLS startup, so do the same.
1133 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1134 TCP connections (such as for Spamd) unless the daemon successfully set
1135 Fast Open mode on its listening sockets. This fixes breakage seen on
1136 too-old kernels or those not configured for Fast Open, at the cost of
1137 requiring both directions being enabled for TFO, and TFO never being used
1138 by non-daemon-related Exim processes.
1140 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1141 endings, at least on the first header line. Try to canonify any that get
1142 past that check, despite the cost.
1144 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1145 now limited to an arbitrary five deep, while parsing addresses with the
1146 strip_excess_angle_brackets option enabled.
1148 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1149 instead leave the unprompted TLS handshake in socket buffer for the
1150 TLS library to consume.
1152 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1154 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1156 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1157 region in one operation, for speed, and this leaves assigned pointers
1158 dangling. Add checks run only under the testsuite which checks all
1159 variables at a store-reset and panics on a dangling pointer; add code
1160 explicitly nulling out all the variables discovered. Fixes one known
1161 bug: a transport crash, where a dangling pointer for $sending_ip_address
1162 originally assigned in a verify callout, is re-used.
1164 PP/06 Drop '.' from @INC in various Perl scripts.
1166 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1168 PP/08 Reduce a number of compilation warnings under clang; building with
1169 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1170 should be warning-free.
1172 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1174 HS/01 Fix portability problems introduced by PP/08 for platforms where
1175 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1177 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1178 chunk. This allows us to accept broken chunked messages. We need a more
1179 general solution here.
1181 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1182 already-broken messages in the queue.
1184 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1186 JH/19 Fix reference counting bug in routing-generated-address tracking.
1192 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1193 supports it and a size is available (ie. the sending peer gave us one).
1195 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1196 years of being deprecated). The replacements are the ACLs
1197 acl_smtp_mime and acl_not_smtp_mime.
1199 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1200 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1201 or even an in-clear connection were permitted. Now, if the host lookup
1202 was dnssec and dane was requested then the host is only used if the
1203 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1204 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1205 if one fails this test.
1206 This means that a poorly-configured remote DNS will make it incommunicado;
1207 but it protects against a DNS-interception attack on it.
1209 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1210 non-noisy when a race steals the message being considered.
1212 JH/05 If main configuration option tls_certificate is unset, generate a
1213 self-signed certificate for inbound TLS connections.
1215 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1216 in rewrites and routers.
1218 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1219 and logged a warning sing 4.83; now they are a configuration file error.
1221 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1222 (lacking @domain). Apply the same qualification processing as RCPT.
1224 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1226 JH/10 Support ${sha256:} applied to a string (as well as the previous
1229 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1230 a cutthrough deliver is pending, as we always want to make a connection.
1231 This also avoids re-routing the message when later placing the cutthrough
1232 connection after a verify cache hit.
1233 Do not update it with the verify result either.
1235 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1236 when routing results in more than one destination address.
1238 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1239 signing (which inhibits the cutthrough capability). Previously only
1240 the presence of an option was tested; now an expansion evaluating as
1241 empty is permissible (obviously it should depend only on data available
1242 when the cutthrough connection is made).
1244 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1245 the relevant preceding SMTP command did not note the pipelining mode.
1247 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1248 Previously they were not counted.
1250 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1251 as one having no matching records. Previously we deferred the message
1252 that needed the lookup.
1254 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1255 distinguished as "(=".
1257 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1258 for missing MX records. Previously it only worked for missing A records.
1260 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1262 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1263 after the data-go-ahead and data-ack. Patch from Jason Betts.
1265 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1266 even for a "none" policy. Patch from Tony Meyer.
1268 JH/22 Fix continued use of a connection for further deliveries. If a port was
1269 specified by a router, it must also match for the delivery to be
1272 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1273 When one of the recipients of a message was unsuitable for the connection
1274 (has no matching addresses), we lost track of needing to mark it
1275 deferred. As a result mail would be lost.
1277 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1279 JH/25 Decoding ACL controls is now done using a binary search; the source code
1280 takes up less space and should be simpler to maintain. Merge the ACL
1281 condition decode tables also, with similar effect.
1283 JH/26 Fix problem with one_time used on a redirect router which returned the
1284 parent address unchanged. A retry would see the parent address marked as
1285 delivered, so not attempt the (identical) child. As a result mail would
1288 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1289 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1290 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1293 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1294 Main config options check_{log,spool}_{inodes,space} are now
1295 100 inodes, 10MB unless set otherwise in the configuration.
1297 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1298 Previously it only applied to the main-section connection policy
1301 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1303 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1304 by me. Added RFC7919 DH primes as an alternative.
1306 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1307 are not in the system include path.
1309 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1310 GnuTLS, when a session startup failed (eg because the client disconnected)
1311 Exim did stdio operations after fclose. This was exposed by a recent
1312 change which nulled out the file handle after the fclose.
1314 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1315 signed directly by the cert-signing cert, rather than an intermediate
1316 OCSP-signing cert. This is the model used by LetsEncrypt.
1318 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1320 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1321 an incoming connection.
1323 HS/02 Bug 1802: Do not half-close the connection after sending a request
1326 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1327 fallback to "prime256v1".
1329 JH/34 SECURITY: Use proper copy of DATA command in error message.
1330 Could leak key material. Remotely exploitable. CVE-2016-9963.
1336 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1337 and 3.4.4 - once the server is enabled to respond to an OCSP request
1338 it does even when not requested, resulting in a stapling non-aware
1339 client dropping the TLS connection.
1341 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1342 support variable-length bit vectors. No functional change.
1344 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1345 The I= interface field on outgoing lines is now after the H= remote
1346 host field, same as incoming lines. There is a separate
1347 outgoing_interface log selector which allows you to disable the
1350 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1351 If not running log_selector +smtp_connection the mainlog would be held
1352 open indefinitely after a "too many connections" event, including to a
1353 deleted file after a log rotate. Leave the per net connection logging
1354 leaving it open for efficiency as that will be quickly detected by the
1355 check on the next write.
1357 HS/01 Bug 1671: Fix post transport crash.
1358 Processing the wait-<transport> messages could crash the delivery
1359 process if the message IDs didn't exist for some reason. When
1360 using 'split_spool_directory=yes' the construction of the spool
1361 file name failed already, exposing the same netto behaviour.
1363 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1364 mime_regex ACL conditions.
1366 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1367 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1368 to HELO, local diagnostic string.
1370 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1371 log line to debug. Even when configured with a tls authenticator many
1372 client connections are expected to not authenticate in this way, so
1373 an authenticate fail is not an error.
1375 HS/02 Add the Exim version string to the process info. This way exiwhat
1376 gives some more detail about the running daemon.
1378 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1379 matter for fast-change records such as DNSBLs.
1381 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1382 retry record, even if constant. There may be multiple transports with
1383 different interface settings and the retry behaviour needs to be kept
1386 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1388 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1390 JH/10 Bug 840: fix log_defer_output option of pipe transport
1392 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1393 hosts_randomize. This matters a lot when combined with mua_wrapper.
1395 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1396 ${quote_pgsql:<string>} operator.
1398 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1400 JH/14 Change header file name for freeradius-client. Relevant if compiling
1401 with Radius support; from the Gentoo tree and checked under Fedora.
1403 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1405 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1406 option transport_filter, meaning no filtering.
1408 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1410 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1411 defaults to "*" (all hosts). The variable is now available when not built
1412 with TLS, default unset, mainly to enable keeping the testsuite sane.
1413 If a server certificate is not supplied (via tls_certificate) an error is
1414 logged, and clients will find TLS connections fail on startup. Presumably
1415 they will retry in-clear.
1416 Packagers of Exim are strongly encouraged to create a server certificate
1417 at installation time.
1419 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1420 with the $config_file variable.
1422 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1423 in transport context, after the attempt, and per-recipient. The latter type
1424 is per host attempted. The event data is the error message, and the errno
1425 information encodes the lookup type (A vs. MX) used for the (first) host,
1426 and the trailing two digits of the smtp 4xx response.
1428 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1429 to write to mainlog (or rejectlog, paniclog) in the window between file
1430 creation and permissions/ownership being changed. Particularly affects
1431 installations where exicyclog is run as root, rather than exim user;
1432 result is that the running daemon panics and dies.
1434 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1436 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1437 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1438 "pri" and "weight". Note that the previous implicit priority given by the
1439 list order is no longer honoured.
1441 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1442 for DKIM processing.
1444 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1445 by defining SUPPORT_SOCKS.
1447 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1448 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1449 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1450 variables are renamed to proxy_{local,external}_{address,port}.
1452 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1453 for a build by defining SUPPORT_I18N
1455 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1456 of the query string, and make ${quote_redis:} do that quoting.
1458 JH/29 Move Events support from Experimental to mainline, enabled by default
1459 and removable for a build by defining DISABLE_EVENT.
1461 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1463 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1464 cached by the daemon.
1466 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1467 by defining LOOKUP_REDIS. The libhiredis library is required.
1469 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1470 keys are given for lookup.
1472 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1473 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1474 only supported when built with TLS support. The PolarSSL SHA routines
1475 are still used when the TLS library is too old for convenient support.
1477 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1478 openssl_options), for security. OpenSSL forces this from version 1.1.0
1479 server-side so match that on older versions.
1481 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1482 allocation for $value could be released as the expansion processing
1483 concluded, but leaving the global pointer active for it.
1485 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1486 and to use the domains and local_parts ACL conditions.
1488 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1489 incorrectly not doubled on cutthrough transmission, hence seen as a
1490 body-termination at the receiving system - resulting in truncated mails.
1491 Commonly the sender saw a TCP-level error, and retransmitted the message
1492 via the normal store-and-forward channel. This could result in duplicates
1493 received - but deduplicating mailstores were liable to retain only the
1494 initial truncated version.
1496 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1498 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1500 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1501 we're in there, support oversigning also; bug 1309.
1503 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1505 HS/04 Add support for keep_environment and add_environment options.
1507 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1508 either intentional arithmetic overflow during PRNG, or testing config-
1511 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1512 delivery resulted in actual delivery. Cancel cutthrough before DATA
1515 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1516 rejects a recipient: pass the reject to the originator.
1518 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1519 Many were false-positives and ignorable, but it's worth fixing the
1522 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1523 for the new environment-manipulation done at startup. Move the routines
1524 from being local to tls.c to being global via the os.c file.
1526 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1527 an extract embedded as result-arg for a map, the first arg for extract
1528 is unavailable so we cannot tell if this is a numbered or keyed
1529 extraction. Accept either.
1535 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1538 JH/02 The smtp transport option "multi_domain" is now expanded.
1540 JH/03 The smtp transport now requests PRDR by default, if the server offers
1543 JH/04 Certificate name checking on server certificates, when exim is a client,
1544 is now done by default. The transport option tls_verify_cert_hostnames
1545 can be used to disable this per-host. The build option
1546 EXPERIMENTAL_CERTNAMES is withdrawn.
1548 JH/05 The value of the tls_verify_certificates smtp transport and main options
1549 default to the word "system" to access the system default CA bundle.
1550 For GnuTLS, only version 3.0.20 or later.
1552 JH/06 Verification of the server certificate for a TLS connection is now tried
1553 (but not required) by default. The verification status is now logged by
1554 default, for both outbound TLS and client-certificate supplying inbound
1557 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1560 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1561 Status Notification (bounce) messages are now MIME format per RFC 3464.
1562 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1563 under the control of the dsn_advertise_hosts option, and routers may
1564 have a dsn_lasthop option.
1566 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1567 default, modifiable by a malware= option. The list separator for
1568 the options can now be changed in the usual way. Bug 68.
1570 JH/10 The smtp_receive_timeout main option is now expanded before use.
1572 JH/11 The incoming_interface log option now also enables logging of the
1573 local interface on delivery outgoing connections.
1575 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1576 if the interface and destination host and port all match.
1578 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1581 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1582 Patch from Andrew Lewis.
1584 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1585 now supports optional time-restrictions, weighting, and priority
1586 modifiers per server. Patch originally by <rommer@active.by>.
1588 JH/16 The spamd_address main option now supports a mixed list of local
1589 and remote servers. Remote servers can be IPv6 addresses, and
1590 specify a port-range.
1592 JH/17 Bug 68: The spamd_address main option now supports an optional
1593 timeout value per server.
1595 JH/18 Bug 1581: Router and transport options headers_add/remove can
1596 now have the list separator specified.
1598 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1601 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1604 JH/21 Support for the A6 type of dns record is withdrawn.
1606 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1607 rather than the verbs used.
1609 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1610 from 255 to 1024 chars.
1612 JH/24 Verification callouts now attempt to use TLS by default.
1614 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1615 are generic router options now. The defaults didn't change.
1617 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1618 Original patch from Alexander Shikoff, worked over by JH.
1620 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1621 files if ZCAT_COMMAND is not executable.
1623 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1625 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1627 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1628 Normally benign, it bites when the pair was led to by a CNAME;
1629 modern usage is to not canonicalize the domain to a CNAME target
1630 (and we were inconsistent anyway for A-only vs AAAA+A).
1632 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1634 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1635 when evaluating $sender_host_dnssec.
1637 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1638 $sender_helo_dnssec variable.
1640 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1642 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1644 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1646 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1647 documented as working, but never had. Support all but $spam_report.
1649 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1650 added for tls authenticator.
1652 HS/03 Add perl_taintmode main config option
1658 TL/01 When running the test suite, the README says that variables such as
1659 no_msglog_check are global and can be placed anywhere in a specific
1660 test's script, however it was observed that placement needed to be near
1661 the beginning for it to behave that way. Changed the runtest perl
1662 script to read through the entire script once to detect and set these
1663 variables, reset to the beginning of the script, and then run through
1664 the script parsing/test process like normal.
1666 TL/02 The BSD's have an arc4random API. One of the functions to induce
1667 adding randomness was arc4random_stir(), but it has been removed in
1668 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1669 function when detected.
1671 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1672 cause callback expansion.
1674 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1675 syntax errors in an expansion can be treated as a string instead of
1676 logging or causing an error, due to the internal use of bool_lax
1677 instead of bool when processing it.
1679 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1680 server certificates when making smtp deliveries.
1682 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1684 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1686 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1688 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1689 Merged patch from Sebastian Wiedenroth.
1691 JH/05 Fix results-pipe from transport process. Several recipients, combined
1692 with certificate use, exposed issues where response data items split
1693 over buffer boundaries were not parsed properly. This eventually
1694 resulted in duplicates being sent. This issue only became common enough
1695 to notice due to the introduction of connection certificate information,
1696 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1698 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1699 size buffer was used, resulting in syntax errors when an expansion
1702 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1703 version 3.3.6 or later.
1705 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1706 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1707 both become "event_action", the variables become $event_name, $event_data
1708 and $event_defer_errno. There is a new variable $verify_mode, usable in
1709 routers, transports and related events. The tls:cert event is now also
1710 raised for inbound connections, if the main configuration event_action
1713 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1714 early OCSP support, but no stapling (appears to be less than 1.0.0).
1716 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1717 server certificate names available under the smtp transport option
1718 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1721 JH/10 Time-related extraction expansions from certificates now use the main
1722 option "timezone" setting for output formatting, and are consistent
1723 between OpenSSL and GnuTLS compilations. Bug 1541.
1725 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1726 encoded parameter in the incoming message. Bug 1558.
1728 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1729 include certificate info, eximon was claiming there were spoolfile
1732 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1734 JH/14 Log delivery-related information more consistently, using the sequence
1735 "H=<name> [<ip>]" wherever possible.
1737 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1738 are problematic for Debian distribution, omit them from the release
1741 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1743 JH/16 Fix string representation of time values on 64bit time_t architectures.
1746 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1747 output list separator was used.
1752 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1753 checkers that were complaining about end of non-void function with no
1756 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1757 This was a regression introduced in 4.83 by another bugfix.
1759 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1761 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1762 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1768 TF/01 Correctly close the server side of TLS when forking for delivery.
1770 When a message was received over SMTP with TLS, Exim failed to clear up
1771 the incoming connection properly after forking off the child process to
1772 deliver the message. In some situations the subsequent outgoing
1773 delivery connection happened to have the same fd number as the incoming
1774 connection previously had. Exim would try to use TLS and fail, logging
1775 a "Bad file descriptor" error.
1777 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1778 utilities have not been installed.
1780 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1781 temporary space as the ACL may create new global variables.
1783 TL/01 LDAP support uses per connection or global context settings, depending
1784 upon the detected version of the libraries at build time.
1786 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1787 to extract and use the src ip:port in logging and expansions as if it
1788 were a direct connection from the outside internet. PPv2 support was
1789 updated based on HAProxy spec change in May 2014.
1791 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1793 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1794 Properly escape header and check for NULL return.
1796 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1799 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1801 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1802 characters in header names, implemented as a verify condition.
1803 Contributed by Michael Fischer v. Mollard.
1805 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1806 results permerror and temperror. Previous values are deprecated but
1807 still accepted. In a future release, err_perm and err_temp will be
1808 completely removed, which will be a backward incompatibility if the
1809 ACL tests for either of these two old results. Patch contributed by
1810 user bes-internal on the mailing list.
1812 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1814 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1815 selectors, in both main and reject logs.
1817 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1820 JH/07 Add malware type "sock" for talking to simple daemon.
1822 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1824 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1825 routers/transports under cutthrough routing.
1827 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1828 numbers. Touch up "bool" conditional to keep the same definition.
1830 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1832 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1834 JH/12 Expand items in router/transport headers_add or headers_remove lists
1835 individually rather than the list as a whole. Bug 1452.
1837 Required for reasonable handling of multiple headers_ options when
1838 they may be empty; requires that headers_remove items with embedded
1839 colons must have them doubled (or the list-separator changed).
1841 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1842 view the policy declared in the DMARC record. Currently, $dmarc_status
1843 is a combined value of both the record presence and the result of the
1846 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1848 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1849 dnslookup router and the smtp transport (applying to the forward
1852 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1853 of ldap servers used for a specific lookup. Patch provided by Heiko
1856 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1857 New variable $lookup_dnssec_authenticated for observability.
1859 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1860 Patch submitted by Lars Timman.
1862 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1864 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1865 Requires trusted mode and valid format message id, aborts otherwise.
1866 Patch contributed by Heiko Schlichting.
1868 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1869 certextract with support for various fields. Bug 1358.
1871 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1872 is requested by default, modifiable by smtp transport option
1875 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1876 operate on certificate variables to give certificate fingerprints
1877 Also new ${sha256:cert_variable}.
1879 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1881 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1882 Christian Aistleitner.
1884 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1886 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1887 file. Patch from Wolfgang Breyha.
1889 JH/25 Expand the coverage of the delivery $host and $host_address to
1890 client authenticators run in verify callout. Bug 1476.
1892 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1893 align with daemon_smtp_ports. Bug 72.
1895 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1896 support and error reporting did not work properly.
1898 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1899 and is readable. Patch from Andrew Colin Kissa.
1901 TL/14 Enhance documentation of ${run expansion and how it parses the
1902 commandline after expansion, particularly in the case when an
1903 unquoted variable expansion results in an empty value.
1905 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1907 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1908 by renaming away from it. Addresses GH issue 15, reported by
1911 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1913 TL/15 SECURITY: prevent double expansion in math comparison functions
1914 (can expand unsanitized data). Not remotely exploitable.
1921 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1923 PP/02 Make -n do something, by making it not do something.
1924 When combined with -bP, the name of an option is not output.
1926 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1929 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1930 $sender_host_name and config options to manage this, and basic check
1933 PP/05 DSCP support for outbound connections and control modifier for inbound.
1935 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1936 (Only plugin which currently uses this is kerberos4, which nobody should
1937 be using, but we should make it available and other future plugins might
1938 conceivably use it, even though it would break NAT; stuff *should* be
1939 using channel bindings instead).
1941 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1942 name; added for Sendmail compatibility; requires admin caller.
1943 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1944 just ignore it); requires trusted caller.
1945 Also parse but ignore: -Ac -Am -X<logfile>
1948 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1950 TL/02 Add +smtp_confirmation as a default logging option.
1952 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1953 Patch by Magnus Holmgren from 2007-02-20.
1955 TL/04 Bugzilla 1281 - Spec typo.
1956 Bugzilla 1283 - Spec typo.
1957 Bugzilla 1290 - Spec grammar fixes.
1959 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1961 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1963 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1964 mailing list by by Dmitry Isaikin.
1966 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1968 JH/02 Support "G" suffix to numbers in ${if comparisons.
1970 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1972 NM/01 Bugzilla 1197 - Spec typo
1973 Bugzilla 1196 - Spec examples corrections
1975 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1977 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1978 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1981 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1982 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1984 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1985 "acl {{name}{arg}...}", and optional args on acl condition
1988 JH/05 Permit multiple router/transport headers_add/remove lines.
1990 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1992 JH/07 Avoid using a waiting database for a single-message-only transport.
1993 Performance patch from Paul Fisher. Bugzilla 1262.
1995 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1998 JH/09 Add $headers_added variable, with content from use of ACL modifier
1999 add_header (but not yet added to the message). Bugzilla 199.
2001 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2002 Pulled from Bugzilla 817 by Wolfgang Breyha.
2004 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2006 (nb: this is the same fix as in Exim 4.80.1)
2008 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2011 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2014 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2016 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2017 advertise SMTP AUTH mechanism to us, instead of a generic
2018 protocol violation error. Also, make Exim more robust to bad
2019 data from the Dovecot auth socket.
2021 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2023 When a queue runner is handling a message, Exim first routes the
2024 recipient addresses, during which it prunes them based on the retry
2025 hints database. After that it attempts to deliver the message to
2026 any remaining recipients. It then updates the hints database using
2029 So if a recipient address works intermittently, it can get repeatedly
2030 deferred at routing time. The retry hints record remains fresh so the
2031 address never reaches the final cutoff time.
2033 This is a fairly common occurrence when a user is bumping up against
2034 their storage quota. Exim had some logic in its local delivery code
2035 to deal with this. However it did not apply to per-recipient defers
2036 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2038 This change adds a proper retry rule check during routing so that the
2039 final cutoff time is checked against the message's age. We only do
2040 this check if there is an address retry record and there is not a
2041 domain retry record; this implies that previous attempts to handle
2042 the address had the retry_use_local_parts option turned on. We use
2043 this as an approximation for the destination being like a local
2044 delivery, as in LMTP.
2046 I suspect this new check makes the old local delivery cutoff check
2047 redundant, but I have not verified this so I left the code in place.
2049 TF/02 Correct gecos expansion when From: is a prefix of the username.
2051 Test 0254 submits a message to Exim with the header
2055 When I ran the test suite under the user fanf2, Exim expanded
2056 the header to contain my full name, whereas it should have added
2057 a Resent-Sender: header. It erroneously treats any prefix of the
2058 username as equal to the username.
2060 This change corrects that bug.
2062 GF/01 DCC debug and logging tidyup
2063 Error conditions log to paniclog rather than rejectlog.
2064 Debug lines prefixed by "DCC: " to remove any ambiguity.
2066 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2068 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2069 Bug spotted by Jeremy Harris; was flawed since initial commit.
2070 Would have resulted in OCSP responses post-SNI triggering an Exim
2071 NULL dereference and crash.
2073 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2075 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2076 Bug detection, analysis and fix by Samuel Thibault.
2077 Bugzilla 1331, Debian bug #698092.
2079 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2081 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2082 Server implementation by Todd Lyons, client by JH.
2083 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2084 config variable "prdr_enable" controls whether the server
2085 advertises the facility. If the client requests PRDR a new
2086 acl_data_smtp_prdr ACL is called once for each recipient, after
2087 the body content is received and before the acl_smtp_data ACL.
2088 The client is controlled by both of: a hosts_try_prdr option
2089 on the smtp transport, and the server advertisement.
2090 Default client logging of deliveries and rejections involving
2091 PRDR are flagged with the string "PRDR".
2093 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2094 fclose(). Diagnosis by Todd Lyons.
2096 PP/17 Update configure.default to handle IPv6 localhost better.
2097 Patch by Alain Williams (plus minor tweaks).
2100 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2101 This is now consistent with GnuTLS, and is now documented: the
2102 previous undocumented portable approach to treating the option as
2103 unset was to force an expansion failure. That still works, and
2104 an empty string is now equivalent.
2106 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2107 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2108 not performing validation itself.
2110 PP/20 Added force_command boolean option to pipe transport.
2111 Patch from Nick Koston, of cPanel Inc.
2113 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2116 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2118 PP/21 Fix eximon continuous updating with timestamped log-files.
2119 Broken in a format-string cleanup in 4.80, missed when I repaired the
2120 other false fix of the same issue.
2121 Report and fix from Heiko Schlichting.
2124 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2125 Report from Prashanth Katuri.
2127 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2128 It's SecureTransport, so affects any MacOS clients which use the
2129 system-integrated TLS libraries, including email clients.
2131 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2132 using a MIME ACL for non-SMTP local injection.
2133 Report and assistance in diagnosis by Warren Baker.
2135 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2137 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2139 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2140 last id that failed so it may be referenced in subsequent ACL's.
2142 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2145 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2146 ldap library initialization, allowing self-signed CA's to be
2147 used. Also properly sets require_cert option later in code by
2148 using NULL (global ldap config) instead of ldap handle (per
2149 session). Bug diagnosis and testing by alxgomz.
2151 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2152 the src/util/ subdirectory.
2154 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2155 renamed to Transport Post Delivery Action by Jeremy Harris, as
2158 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2159 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2160 redis_servers = needs to be configured which will be used by the redis
2161 lookup. Patch from Warren Baker, of The Packet Hub.
2163 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2165 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2166 hostname or reverse DNS when processing a host list. Used suggestions
2167 from multiple comments on this bug.
2169 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2171 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2172 Missed a few lines, added it to make the runtest require no keyboard
2175 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2176 contains upper case chars. Make router use caseful_local_part.
2178 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2179 support when GnuTLS has been built with p11-kit.
2185 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2187 This, or similar/improved, will also be change PP/11 of 4.82.
2193 PP/01 Handle short writes when writing local log-files.
2194 In practice, only affects FreeBSD (8 onwards).
2195 Bugzilla 1053, with thanks to Dmitry Isaikin.
2197 NM/01 Bugzilla 949 - Documentation tweak
2199 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2202 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2204 PP/02 Implemented gsasl authenticator.
2206 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2208 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2209 `pkg-config foo` for cflags/libs.
2211 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2212 with rest of GSASL and with heimdal_gssapi.
2214 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2215 `pkg-config foo` for cflags/libs for the TLS implementation.
2217 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2218 properties get this fed in as external SSF. A number of robustness
2219 and debugging improvements to the cyrus_sasl authenticator.
2221 PP/08 cyrus_sasl server now expands the server_realm option.
2223 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2224 Patch by Jeremy Harris.
2226 PP/10 Added dbmjz lookup type.
2228 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2230 PP/12 MAIL args handles TAB as well as SP, for better interop with
2231 non-compliant senders.
2232 Analysis and variant patch by Todd Lyons.
2234 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2235 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2236 Patch from Dirk Mueller <dmueller@suse.com>
2238 PP/13 tls_peerdn now print-escaped for spool files.
2239 Observed some $tls_peerdn in wild which contained \n, which resulted
2240 in spool file corruption.
2242 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2243 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2244 or write after TLS renegotiation, which otherwise led to messages
2247 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2248 as a tracking header (ie: a signed header comes before the signature).
2249 Patch from Wolfgang Breyha.
2251 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2252 comma-sep list; embedded commas doubled.
2254 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2256 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2258 Report and patch from Dmitry Banschikov.
2260 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2261 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2262 failures. We appear to now support TLS1.1+ with Exim.
2264 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2265 lets Exim select keys and certificates based upon TLS SNI from client.
2266 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2267 before an outbound SMTP session. New log_selector, +tls_sni.
2269 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2270 NULL dereference. Report and patch from Alun Jones.
2272 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2273 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2274 Not seeing resolver debug output on NetBSD, but suspect this is a
2275 resolver implementation change.
2277 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2278 Left warnings. Added "eximon gdb" invocation mode.
2280 PP/21 Defaulting "accept_8bitmime" to true, not false.
2282 PP/22 Added -bw for inetd wait mode support.
2284 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2285 locate the relevant includes and libraries. Made this the default.
2287 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2288 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2290 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2291 This may cause build issues on older platforms.
2293 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2294 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2295 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2296 Added SNI support via GnuTLS too.
2297 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2299 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2301 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2302 Applied second patch from Janne, implementing suggestion to default
2303 multiple-strings-in-record handling to match SPF spec.
2305 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2307 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2308 read-only, out of scope).
2309 Patch from Wolfgang Breyha, report from Stuart Northfield.
2311 PP/29 Fix three issues highlighted by clang analyser static analysis.
2312 Only crash-plausible issue would require the Cambridge-specific
2313 iplookup router and a misconfiguration.
2314 Report from Marcin Mirosław.
2316 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2318 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2319 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2320 As part of this, removing so much warning spew let me fix some minor
2321 real issues in debug logging.
2323 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2324 assignment on my part. Fixed.
2326 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2327 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2328 Janne Snabb (who went above and beyond: thank you).
2330 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2331 string otherwise requires a connection and a bunch more work and it's
2332 relatively easy to get wrong. Should also expose TLS library linkage
2335 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2336 64-bit ${eval} (JH/03).
2338 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2339 GNU libc to support some of the 64-bit stuff, should not lead to
2340 conflicts. Defined before os.h is pulled in, so if a given platform
2341 needs to override this, it can.
2343 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2344 protection layer was required, which is not implemented.
2345 Bugzilla 1254, patch from Wolfgang Breyha.
2347 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2348 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2349 tls_dhparam take prime identifiers. Also unbreak combination of
2350 OpenSSL+DH_params+TLSSNI.
2352 PP/39 Disable SSLv2 by default in OpenSSL support.
2358 PP/01 Solaris build fix for Oracle's LDAP libraries.
2359 Bugzilla 1109, patch from Stephen Usher.
2361 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2363 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2366 TF/02 Fix a couple more cases where we did not log the error message
2367 when unlink() failed. See also change 4.74-TF/03.
2369 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2370 lock up or crash if it happened to be inside a call to libc when it
2371 got a SIGUSR1 from exiwhat.
2373 The SIGUSR1 handler appends the current process status to the process
2374 log which is later printed by exiwhat. It used to use the general
2375 purpose logging code to do this, but several functions it calls are
2376 not safe for signals.
2378 The new output code in the SIGUSR1 handler is specific to the process
2379 log, and simple enough that it's easy to inspect for signal safety.
2380 Removing some special cases also simplifies the general logging code.
2381 Removing the spurious timestamps from the process log simplifies
2384 TF/04 Improved ratelimit ACL condition.
2386 The /noupdate option has been deprecated in favour of /readonly which
2387 has clearer semantics. The /leaky, /strict, and /readonly update modes
2388 are mutually exclusive. The update mode is no longer included in the
2389 database key; it just determines when the database is updated. (This
2390 means that when you upgrade Exim will forget old rate measurements.)
2392 Exim now checks that the per_* options are used with an update mode that
2393 makes sense for the current ACL. For example, when Exim is processing a
2394 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2395 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2396 must specify per_mail/readonly. If you omit the update mode it defaults to
2397 /leaky where that makes sense (as before) or /readonly where required.
2399 The /noupdate option is now undocumented but still supported for
2400 backwards compatibility. It is equivalent to /readonly except that in
2401 ACLs where /readonly is required you may specify /leaky/noupdate or
2402 /strict/noupdate which are treated the same as /readonly.
2404 A useful new feature is the /count= option. This is a generalization
2405 of the per_byte option, so that you can measure the throughput of other
2406 aggregate values. For example, the per_byte option is now equivalent
2407 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2409 The per_rcpt option has been generalized using the /count= mechanism
2410 (though it's more complicated than the per_byte equivalence). When it is
2411 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2412 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2413 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2414 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2415 Note that using per_rcpt with a non-readonly update mode in more than
2416 one ACL will cause the recipients to be double-counted. (The per_mail
2417 and per_byte options don't have this problem.)
2419 The handling of very low rates has changed slightly. If the computed rate
2420 is less than the event's count (usually one) then this event is the first
2421 after a long gap. In this case the rate is set to the same as this event's
2422 count, so that the first message of a spam run is counted properly.
2424 The major new feature is a mechanism for counting the rate of unique
2425 events. The new per_addr option counts the number of different
2426 recipients that someone has sent messages to in the last time period. It
2427 behaves like per_rcpt if all the recipient addresses are different, but
2428 duplicate recipient addresses do not increase the measured rate. Like
2429 the /count= option this is a general mechanism, so the per_addr option
2430 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2431 example, measure the rate that a client uses different sender addresses
2432 with the options per_mail/unique=$sender_address. There are further
2433 details in the main documentation.
2435 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2437 TF/06 Removed a few PCRE remnants.
2439 TF/07 Automatically extract Exim's version number from tags in the git
2440 repository when doing development or release builds.
2442 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2443 Bugzilla 879. Patch from Paul Fisher.
2445 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2446 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2449 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2451 PP/05 Variable $av_failed, true if the AV scanner deferred.
2452 Bugzilla 1078. Patch from John Horne.
2454 PP/06 Stop make process more reliably on build failure.
2455 Bugzilla 1087. Patch from Heiko Schlittermann.
2457 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2458 Bugzilla 1089. Patch from Heiko Schlittermann.
2460 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2461 Bugzilla 1131. Patch from Holger Weiß.
2463 PP/09 Handle IPv6 addresses with SPF.
2464 Bugzilla 860. Patch from Wolfgang Breyha.
2466 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2468 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2471 PP/11 match_* no longer expand right-hand-side by default.
2472 New compile-time build option, EXPAND_LISTMATCH_RHS.
2473 New expansion conditions, "inlist", "inlisti".
2475 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2477 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2479 PP/14 fix log_write() format string regression from TF/03.
2480 Bugzilla 1152. Patch from Dmitry Isaikin.
2486 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2488 PP/02 Harmonised TLS library version reporting; only show if debugging.
2489 Layout now matches that introduced for other libraries in 4.74 PP/03.
2491 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2493 PP/04 New "dns_use_edns0" global option.
2495 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2498 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2499 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2501 TK/01 Updated PolarSSL code to 0.14.2.
2502 Bugzilla 1097. Patch from Andreas Metzler.
2504 PP/07 Catch divide-by-zero in ${eval:...}.
2505 Fixes bugzilla 1102.
2507 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2510 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2511 format-string attack -- SECURITY: remote arbitrary code execution.
2513 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2514 time unintentionally subject to list matching rules, letting the header
2515 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2516 arbitrary string expansion). This allowed for information disclosure.
2518 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2519 INT_MIN/-1 -- value coerced to INT_MAX.
2525 NM/01 Workaround for PCRE version dependency in version reporting
2528 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2529 This fixes portability to compilers other than gcc, notably
2530 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2532 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2533 makefiles for portability to HP-UX and POSIX correctness.
2535 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2536 Also via indented variable definition in the Makefile.
2537 (Debugging by Oliver Heesakkers).
2539 PP/02 Restore caching of spamd results with expanded spamd_address.
2540 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2542 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2543 Improves build reliability. Fix from: Frank Elsner
2545 NM/02 Fix wide character breakage in the rfc2047 coding
2546 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2548 NM/03 Allow underscore in dnslist lookups
2549 Fixes bug 1026. Patch from Graeme Fowler
2551 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2552 Code patches from Adam Ciarcinski of NetBSD.
2554 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2557 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2558 is logged, to avoid truncation. Patch from John Horne.
2560 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2561 Patch from Jakob Hirsch.
2563 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2564 SQL string expansion failure details.
2565 Patch from Andrey Oktyabrski.
2567 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2568 Patch from Simon Arlott.
2570 PP/09 New lookups functionality failed to compile on old gcc which rejects
2571 extern declarations in function scope.
2572 Patch from Oliver Fleischmann
2574 PP/10 Use sig_atomic_t for flags set from signal handlers.
2575 Check getgroups() return and improve debugging.
2576 Fixed developed for diagnosis in bug 927 (which turned out to be
2579 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2580 Patch from Mark Zealey.
2582 PP/12 Bugzilla 1056: Improved spamd server selection.
2583 Patch from Mark Zealey.
2585 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2586 Based on patch from Heiko Schlittermann.
2588 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2589 Patch from Uwe Doering, sign-off by Michael Haardt.
2591 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2592 variable declaration deep within a block. Bug and patch from
2595 PP/15 lookups-Makefile IRIX compatibility coercion.
2597 PP/16 Make DISABLE_DKIM build knob functional.
2599 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2600 Patch by Simon Arlott
2602 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2603 variable argument macros. Our copy now differs from upstream.
2609 TF/01 Failure to get a lock on a hints database can have serious
2610 consequences so log it to the panic log.
2612 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2613 controlled using the smtp_confirmation log selector.
2615 TF/03 Include the error message when we fail to unlink a spool file.
2617 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2618 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2619 for maintaining out-of-tree patches for some time.
2621 PP/01 Bugzilla 139: Documentation and portability issues.
2622 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2623 Handle per-OS dynamic-module compilation flags.
2625 PP/02 Let /dev/null have normal permissions.
2626 The 4.73 fixes were a little too stringent and complained about the
2627 permissions on /dev/null. Exempt it from some checks.
2628 Reported by Andreas M. Kirchwitz.
2630 PP/03 Report version information for many libraries, including
2631 Exim version information for dynamically loaded libraries. Created
2632 version.h, now support a version extension string for distributors
2633 who patch heavily. Dynamic module ABI change.
2635 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2636 privilege escalation vulnerability whereby the Exim run-time user
2637 can cause root to append content of the attacker's choosing to
2640 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2643 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2644 If dropping privileges for untrusted macros, we disabled normal logging
2645 on the basis that it would fail; for the Exim run-time user, this is not
2646 the case, and it resulted in successful deliveries going unlogged.
2647 Fixed. Reported by Andreas Metzler.
2653 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2654 only prepend for the Resent-* case. Fixes regression introduced in
2655 Exim 4.70 by NM/22 for Bugzilla 607.
2657 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2658 increasing numbers of administrators be bitten by this.
2660 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2662 PP/03 Bugzilla 994: added openssl_options main configuration option.
2664 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2666 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2668 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2670 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2671 without a peer certificate, leading to a segfault because of an
2672 assumption that peers always have certificates. Be a little more
2673 paranoid. Problem reported by Martin Tscholak.
2675 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2676 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2677 NB: ClamAV planning to remove STREAM in "middle of 2010".
2678 CL also introduces -bmalware, various -d+acl logging additions and
2679 more caution in buffer sizes.
2681 PP/09 Implemented reverse_ip expansion operator.
2683 PP/10 Bugzilla 937: provide a "debug" ACL control.
2685 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2687 PP/12 Bugzilla 973: Implement --version.
2689 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2691 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2693 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2695 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2696 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2697 ignore trailing whitespace.
2699 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2701 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2704 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2705 Notification from Dr Andrew Aitchison.
2707 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2708 ExtendedDetectionInfo response format.
2709 Notification from John Horne.
2711 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2714 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2715 XSL and documented dependency on system catalogs, with examples of how
2718 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2721 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2722 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2723 configuration file which is writeable by the Exim user or group.
2725 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2726 of configuration files to cover files specified with the -C option if
2727 they are going to be used with root privileges, not just the default
2730 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2731 option (effectively making it always true).
2733 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2734 files to be used while preserving root privileges.
2736 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2737 that rogue child processes cannot use them.
2739 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2740 run-time user, instead of root.
2742 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2743 Exim run-time user without dropping privileges.
2745 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2746 result string, instead of calling string_vformat() twice with the same
2749 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2750 for other users. Others should always drop root privileges if they use
2751 -C on the command line, even for a whitelisted configure file.
2753 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2755 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2761 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2762 $data_path, and $header_path variables; fixed documentation bugs and
2765 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2766 exipick to access non-standard spools, including the "frozen" queue
2769 NM/01 Bugzilla 965: Support mysql stored procedures.
2770 Patch from Alain Williams
2772 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2774 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2775 Patch from Andreas Metzler
2777 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2778 Patch from Kirill Miazine
2780 NM/05 Bugzilla 671: Added umask to procmail example.
2782 JJ/03 installed exipick 20100323.0, fixing doc bug
2784 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2785 directory. Notification and patch from Dan Rosenberg.
2787 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2789 TK/02 Improve log output when DKIM signing operation fails.
2791 MH/01 Treat the transport option dkim_domain as a colon separated
2792 list, not as a single string, and sign the message with each element,
2793 omitting multiple occurences of the same signer.
2795 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2796 Bugzilla 985, 986. Patch by Simon Arlott
2798 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2799 Patch by Simon Arlott
2801 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2802 MBX locking. Notification from Dan Rosenberg.
2808 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2810 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2812 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2814 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2816 NM/04 Bugzilla 915: Flag broken perl installation during build.
2822 TK/01 Added patch by Johannes Berg that expands the main option
2823 "spamd_address" if it starts with a dollar sign.
2825 TK/02 Write list of recipients to X-Envelope-Sender header when building
2826 the mbox-format spool file for content scanning (suggested by Jakob
2829 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2830 (http://www.dcc-servers.net/) support via dccifd. Activated by
2831 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2833 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2834 by Mark Daniel Reidel <mr@df.eu>.
2836 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2837 When building exim an external PCRE library is now needed -
2838 PCRE is a system library on the majority of modern systems.
2839 See entry on PCRE_LIBS in EDITME file.
2841 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2842 conversation. Added nologin parameter to request.
2843 Patch contributed by Kirill Miazine.
2845 TF/01 Do not log submission mode rewrites if they do not change the address.
2847 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2849 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2850 log files in place. Contributed by Roberto Lima.
2852 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2854 TF/03 Bugzilla 615: When checking the local_parts router precondition
2855 after a local_part_suffix or local_part_prefix option, Exim now
2856 does not use the address's named list lookup cache, since this
2857 contains cached lookups for the whole local part.
2859 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2860 Robert Millan. Documentation is in experimental-spec.txt.
2862 TF/04 Bugzilla 668: Fix parallel build (make -j).
2864 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2866 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2867 Patch provided by Jan Srzednicki.
2869 TF/05 Leading white space used to be stripped from $spam_report which
2870 wrecked the formatting. Now it is preserved.
2872 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2873 that they are available at delivery time.
2875 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2877 TF/08 TLS error reporting now respects the incoming_interface and
2878 incoming_port log selectors.
2880 TF/09 Produce a more useful error message if an SMTP transport's hosts
2881 setting expands to an empty string.
2883 NM/06 Bugzilla 744: EXPN did not work under TLS.
2884 Patch provided by Phil Pennock.
2886 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2887 Patch provided by Richard Godbee.
2889 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2890 acl_smtp_notquit, added index entry.
2892 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2893 Patch provided by Eugene Bujak.
2895 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2896 accept(). Patch provided by Maxim Dounin.
2898 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2899 Patch provided by Phil Pennock.
2901 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2903 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2904 Patch provided by Brad "anomie" Jorsch.
2906 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2907 Patch provided by Dean Brooks.
2909 TK/05 Add native DKIM support (does not depend on external libraries).
2911 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2912 Patch provided by Graeme Fowler.
2914 NM/16 Bugzilla 851: Documentation example syntax fix.
2916 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2918 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2921 NM/19 Bugzilla 745: TLS version reporting.
2922 Patch provided by Phil Pennock.
2924 NM/20 Bugzilla 167: bool: condition support.
2925 Patch provided by Phil Pennock.
2927 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2928 clients. Patch provided by Phil Pennock.
2930 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2931 Patch provided by Brad "anomie" Jorsch.
2933 NM/23 Bugzilla 687: Fix misparses in eximstats.
2934 Patch provided by Heiko Schlittermann.
2936 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2937 Patch provided by Heiko Schlittermann.
2939 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2940 plus update to original patch.
2942 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2944 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2945 Patch provided by David Brownlee.
2947 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2949 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2951 NM/30 Bugzilla 888: TLS documentation bugfixes.
2953 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2955 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2956 Unlike the original bugzilla I have changed all shell scripts in src tree.
2958 NM/33 Bugzilla 898: Transport filter timeout fix.
2959 Patch by Todd Rinaldo.
2961 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2962 Patch by Serge Demonchaux.
2964 NM/35 Bugzilla 39: Base64 decode bug fixes.
2965 Patch by Jakob Hirsch.
2967 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2969 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2971 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2973 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2979 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2980 ALT-N's libdkim that I have put here:
2981 http://duncanthrax.net/exim-experimental/
2983 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2984 were called 'true' and it seems that C99 defines that as a reserved
2985 keyword to be used with 'bool' variable types. That means you could
2986 not include C99-style headers which use bools without triggering
2987 build errors in sieve.c.
2989 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2990 as mailq or other aliases. Changed the --help handling significantly
2991 to do whats expected. exim_usage() emits usage/help information.
2993 SC/01 Added the -bylocaldomain option to eximstats.
2995 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2997 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2999 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3005 PH/01 Another patch from the Sieve maintainer.
3007 PH/02 When an IPv6 address is converted to a string for single-key lookup
3008 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3009 dots are used instead of colons so that keys in lsearch files need not
3010 contain colons. This was done some time before quoting was made available
3011 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3012 (notated using the quote facility) so as to distinguish them from IPv4
3013 keys. This meant that lookups for IP addresses in host lists did not work
3014 for iplsearch lookups.
3016 This has been fixed by arranging for IPv6 addresses to be expressed with
3017 colons if the lookup type is iplsearch. This is not incompatible, because
3018 previously such lookups could never work.
3020 The situation is now rather anomalous, since one *can* have colons in
3021 ordinary lsearch keys. However, making the change in all cases is
3022 incompatible and would probably break a number of configurations.
3024 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3027 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3028 conversion specification without a maximum field width, thereby enabling
3029 a rogue spamd server to cause a buffer overflow. While nobody in their
3030 right mind would setup Exim to query an untrusted spamd server, an
3031 attacker that gains access to a server running spamd could potentially
3032 exploit this vulnerability to run arbitrary code as the Exim user.
3034 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3035 $primary_hostname instead of what libspf2 thinks the hosts name is.
3037 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3038 a directory entry by the name of the lookup key. Previously, if a
3039 symlink pointed to a non-existing file or a file in a directory that
3040 Exim lacked permissions to read, a lookup for a key matching that
3041 symlink would fail. Now it is enough that a matching directory entry
3042 exists, symlink or not. (Bugzilla 503.)
3044 PH/03 The body_linecount and body_zerocount variables are now exported in the
3047 PH/04 Added the $dnslist_matched variable.
3049 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3050 This means they are set thereafter only if the connection becomes
3053 PH/06 Added the client_condition to authenticators so that some can be skipped
3054 by clients under certain conditions.
3056 PH/07 The error message for a badly-placed control=no_multiline_responses left
3057 "_responses" off the end of the name.
3059 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3061 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3062 (without spaces) instead of just copying the configuration text.
3064 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3066 PH/11 Added $max_received_linelength.
3068 PH/12 Added +ignore_defer and +include_defer to host lists.
3070 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3071 way in which PCRE > 7.0 is built.
3073 PH/14 Implemented queue_only_load_latch.
3075 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3076 MAIL command. The effect was to mangle the value on 64-bit systems.
3078 PH/16 Another patch from the Sieve maintainer.
3080 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3082 PH/18 If a system quota error occurred while trying to create the file for
3083 a maildir delivery, the message "Mailbox is full" was not appended to the
3084 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3085 only to a quota excession during the actual writing of the file.
3087 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3088 characters?) which causes problems in log lines. The DN values are now
3089 passed through string_printing() before being added to log lines.
3091 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3092 and InterBase are left for another time.)
3094 PH/21 Added message_body_newlines option.
3096 PH/22 Guard against possible overflow in moan_check_errorcopy().
3098 PH/23 POSIX allows open() to be a macro; guard against that.
3100 PH/24 If the recipient of an error message contained an @ in the local part
3101 (suitably quoted, of course), incorrect values were put in $domain and
3102 $local_part during the evaluation of errors_copy.
3108 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3109 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3112 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3113 issue a MAIL command.
3115 PH/02 In an ACL statement such as
3117 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3119 if a client was not listed at all, or was listed with a value other than
3120 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3121 the condition was not true (as it should be), so access was not denied.
3122 The bug was that the ! inversion was incorrectly passed on to the second
3123 item. This has been fixed.
3125 PH/03 Added additional dnslists conditions == and =& which are different from
3126 = and & when the dns lookup returns more than one IP address.
3128 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3129 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3131 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3132 FSYNC, which compiles an option called disable_fsync that allows for
3133 bypassing fsync(). The documentation is heavily laced with warnings.
3135 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3137 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3138 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3139 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3140 including adding "make clean"; (3) Added -fPIC when compiling the test
3141 dynamically loaded module, to get rid of a warning.
3143 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3144 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3145 = 0s. The bug is otherwise harmless.
3147 PH/07 There was a bug in the dovecot authenticator such that the value of
3148 $auth1 could be overwritten, and so not correctly preserved, after a
3149 successful authentication. This usually meant that the value preserved by
3150 the server_setid option was incorrect.
3152 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3154 PH/09 Installed PCRE release 7.0.
3156 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3157 run for batched SMTP input. It is now run at the start of every message
3158 in the batch. While fixing this I discovered that the process information
3159 (output by running exiwhat) was not always getting set for -bs and -bS
3160 input. This is fixed, and it now also says "batched" for BSMTP.
3162 PH/11 Added control=no_pipelining.
3164 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3165 patch, slightly modified), and move the expansion of helo_data till after
3166 the connection is made in the smtp transport (so it can use these
3169 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3171 PH/14 Added log_selector = +pid.
3173 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3175 PH/16 Add ${if forany and ${if forall.
3177 PH/17 Added dsn_from option to vary the From: line in DSNs.
3179 PH/18 Flush SMTP output before performing a callout, unless control =
3180 no_callout_flush is set.
3182 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3183 was true (the default) a successful delivery failed to delete the retry
3184 item, thus causing premature timeout of the address. The bug is now
3187 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3189 PH/21 Long custom messages for fakedefer and fakereject are now split up
3190 into multiline responses in the same way that messages for "deny" and
3191 other ACL rejections are.
3193 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3194 with slight modification.
3196 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3197 draft, changing the syntax and factoring some duplicate code".
3199 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3200 for deliveries of the second and subsequent messages over the same SMTP
3203 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3204 ${reduce, with only minor "tidies".
3206 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3208 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3209 expansion side effects.
3211 PH/27 When a message times out after an over-quota error from an Exim-imposed
3212 quota, the bounce message says "mailbox is full". This message was not
3213 being given when it was a system quota that was exceeded. It now should
3216 MH/03 Made $recipients available in local_scan(). local_scan() already has
3217 better access to the recipient list through recipients_list[], but
3218 $recipients can be useful in postmaster-provided expansion strings.
3220 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3221 in the case of a MAIL command with additional options following the
3222 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3223 were accidentally chopped off.
3225 PH/29 SMTP synchronization checks are implemented when a command is read -
3226 there is a check that no more input is waiting when there shouldn't be
3227 any. However, for some commands, a delay in an ACL can mean that it is
3228 some time before the response is written. In this time, more input might
3229 arrive, invalidly. So now there are extra checks after an ACL has run for
3230 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3231 pipelining has not been advertised.
3233 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3235 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3236 returned in the SMTP response when smtp_return_error_details was false.
3237 This has been fixed.
3239 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3240 instead of the C I/O that was originally supplied, because problems were
3241 reported on Solaris.
3243 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3244 Exim which did not show up earlier: it was assuming that a call to
3245 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3246 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3247 was a macro that became an assignment, so it seemed to work. This has
3248 changed to a proper function call with a void return, hence the compile
3249 error. Exim's code has been fixed.
3251 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3254 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3256 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3258 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3259 because it turns out to be too big or there is a timeout) there is no
3260 "Completed" line in the log. When some messages of this type were
3261 selected by exigrep, they were listed as "not completed". Others were
3262 picked up by some special patterns. I have improved the selection
3263 criteria to be more general.
3265 PH/38 The host_find_failed option in the manualroute router can now be set
3266 to "ignore", to completely ignore a host whose IP address cannot be
3267 found. If all hosts are ignored, the behaviour is controlled by the new
3268 host_all_ignored option.
3270 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3271 homing or because of multiple MX records with /mx) generated more than
3272 one IP address, and the following item turned out to be the local host,
3273 all the secondary addresses of the first item were incorrectly removed
3274 from the list, along with the local host and any following hosts (which
3275 is what is supposed to happen).
3277 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3278 whoever called Exim into the -H file. In the case of the daemon it was
3279 behaving confusingly. When first started, it used values for whoever
3280 started the daemon, but after a SIGHUP it used the Exim user (because it
3281 calls itself on a restart). I have changed the code so that it now always
3284 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3285 message are rejected with the same error (e.g. no authentication or bad
3286 sender address), and a DATA command is nevertheless sent (as can happen
3287 with PIPELINING or a stupid MUA), the error message that was given to the
3288 RCPT commands is included in the rejection of the DATA command. This is
3289 intended to be helpful for MUAs that show only the final error to their
3292 PH/42 Another patch from the Sieve maintainer.
3294 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3295 Eximstats - Fixed some broken HTML links and added missing column headers
3297 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3298 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3300 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3302 PH/43 Yet another patch from the Sieve maintainer.
3304 PH/44 I found a way to check for a TCP/IP connection going away before sending
3305 the response to the final '.' that terminates a message, but only in the
3306 case where the client has not sent further data following the '.'
3307 (unfortunately, this is allowed). However, in many cases there won't be
3308 any further data because there won't be any more messages to send. A call
3309 to select() can be used: if it shows that the input is "ready", there is
3310 either input waiting, or the socket has been closed. An attempt to read
3311 the next input character can distinguish the two cases. Previously, Exim
3312 would have sent an OK response which the client would never have see.
3313 This could lead to message repetition. This fix should cure that, at
3314 least in a lot of common cases.
3316 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3317 advertised in response to EHLO.
3323 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3324 fixed by 4.65/MH/01 (is this a record?) are fixed:
3326 (i) An empty string was always treated as zero by the numeric comparison
3327 operators. This behaviour has been restored.
3329 (ii) It is documented that the numeric comparison operators always treat
3330 their arguments as decimal numbers. This was broken in that numbers
3331 starting with 0 were being interpreted as octal.
3333 While fixing these problems I realized that there was another issue that
3334 hadn't been noticed. Values of message_size_limit (both the global option
3335 and the transport option) were treated as octal if they started with 0.
3336 The documentation was vague. These values are now always treated as
3337 decimal, and I will make that clear in the documentation.
3343 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3344 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3347 MH/01 Don't check that the operands of numeric comparison operators are
3348 integers when their expansion is in "skipping" mode (fixes bug
3349 introduced by 4.64-PH/07).
3351 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3352 child addresses, Exim now panics and dies. Previously, because the count
3353 is held in a short int, deliveries were likely to be lost. As such a
3354 large number of recipients for a single message is ridiculous
3355 (performance will be very, very poor), I have chosen to impose a limit
3356 rather than extend the field.
3362 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3363 leftover -K file (the existence of which was triggered by #402).
3364 While we were at it, introduced process PID as part of the -K
3365 filename. This should rule out race conditions when creating
3368 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3369 processing considerably. Previous code took too long for large mails,
3370 triggering a timeout which in turn triggers #401.
3372 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3373 in the DK code in transports.c. sendfile() is not really portable,
3374 hence the _LINUX specificness.
3376 TF/01 In the add_headers option to the mail command in an Exim filter,
3377 there was a bug that Exim would claim a syntax error in any
3378 header after the first one which had an odd number of characters
3381 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3382 callout verification, Exim cached a "reject" for the entire domain. This
3383 is correct for most verifications, but it is not correct for a recipient
3384 verification with use_sender or use_postmaster set, because in that case
3385 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3386 case of MAIL FROM:<> rejection from other early rejections (e.g.
3387 rejection of HELO). When verifying a recipient using a non-null MAIL
3388 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3389 Whatever the result of the callout, the value of the domain cache is
3390 left unchanged (for any other kind of callout, getting as far as trying
3391 RCPT means that the domain itself is ok).
3393 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3396 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3397 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3398 session, the daemon ignores ECONNECT errors and logs others; it now
3399 ignores EPIPE as well.
3401 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3402 (quoted-printable decoding).
3404 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3405 later the small subsequent patch to fix an introduced bug.
3407 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3409 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3411 PH/08 An error is now given if message_size_limit is specified negative.
3413 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3414 to be given (somewhat) arbitrary names.
3416 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3419 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3420 miscellaneous code fixes
3422 PH/10 Added the log_reject_target ACL modifier to specify where to log
3425 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3426 hostname. This is wrong, because it relates to the incoming message (and
3427 probably the interface on which it is arriving) and not to the outgoing
3428 callout (which could be using a different interface). This has been
3429 changed to use the value of the helo_data option from the smtp transport
3430 instead - this is what is used when a message is actually being sent. If
3431 there is no remote transport (possible with a router that sets up host
3432 addresses), $smtp_active_hostname is used.
3434 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3435 tweaks were necessary in order to get it to work (see also 21 below):
3436 (a) The code assumed that strncpy() returns a negative number on buffer
3437 overflow, which isn't the case. Replaced with Exim's string_format()
3439 (b) There were several signed/unsigned issues. I just did the minimum
3440 hacking in of casts. There is scope for a larger refactoring.
3441 (c) The code used strcasecmp() which is not a standard C function.
3442 Replaced with Exim's strcmpic() function.
3443 (d) The code set only $1; it now sets $auth1 as well.
3444 (e) A simple test gave the error "authentication client didn't specify
3445 service in request". It would seem that Dovecot has changed its
3446 interface. Fortunately there's a specification; I followed it and
3447 changed what the client sends and it appears to be working now.
3449 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3452 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3453 address A is aliased to B and C, where B exists and C does not. Without
3454 -v the output is "A verified" because verification stops after a
3455 successful redirection if more than one address is generated. However,
3456 with -v the child addresses are also verified. Exim was outputting "A
3457 failed to verify" and then showing the successful verification for C,
3458 with its parentage. It now outputs "B failed to verify", showing B's
3459 parentage before showing the successful verification of C.
3461 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3462 look up a TXT record in a specific list after matching in a combined
3465 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3466 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3467 they consult the DNS. I had assumed they would set it the way they
3468 wanted; and indeed my experiments on Linux seem to show that in some
3469 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3470 To be on the safe side, however, I have now made the interface to
3471 host_find_byname() similar to host_find_bydns(), with an argument
3472 containing the DNS resolver options. The host_find_byname() function now
3473 sets these options at its start, just as host_find_bydns() does. The smtp
3474 transport options dns_qualify_single and dns_search_parents are passed to
3475 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3476 of host_find_byname() use the default settings of RES_DEFNAMES
3477 (qualify_single) but not RES_DNSRCH (search_parents).
3479 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3480 spool_read_header() do less string testing, by means of a preliminary
3481 switch on the second character of optional "-foo" lines. (This is
3482 overdue, caused by the large number of possibilities that now exist.
3483 Originally there were few.) While I was there, I also converted the
3484 str(n)cmp tests so they don't re-test the leading "-" and the first
3485 character, in the hope this might squeeze out yet more improvement.
3487 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3488 flag allowing group syntax was set by the header_syntax check but not
3489 turned off, possible causing trouble later; (2) The flag was not being
3490 set at all for the header_verify test, causing "group"-style headers to
3491 be rejected. I have now set it in this case, and also caused header_
3492 verify to ignore an empty address taken from a group. While doing this, I
3493 came across some other cases where the code for allowing group syntax
3494 while scanning a header line wasn't quite right (mostly, not resetting
3495 the flag correctly in the right place). These bugs could have caused
3496 trouble for malformed header lines. I hope it is now all correct.
3498 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3499 with the "reply" argument non-NULL. The code, however (which originally
3500 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3501 but it didn't always do it. This confused somebody who was copying the
3502 code for some other use. I have removed all the tests.
3504 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3505 feature that was used to support insecure browsers during the U.S. crypto
3506 embargo. It requires special client support, and Exim is probably the
3507 only MTA that supported it -- and would never use it because real RSA is
3508 always available. This code has been removed, because it had the bad
3509 effect of slowing Exim down by computing (never used) parameters for the
3510 RSA_EXPORT functionality.
3512 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3513 authenticator to fail if there's a tab character in the incoming data
3514 (there should never be unless someone is messing about, as it's supposed
3515 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3516 if the connection is using TLS or if the remote IP is the same as the
3517 local IP, and the "valid-client-cert option" if a client certificate has
3520 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3521 authenticators. This can be used for authorization after authentication
3522 succeeds. (In the case of plaintext, it servers for both authentication
3525 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3526 if any retry times were supplied.
3528 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3529 connection, where there is no client IP address to check. In this
3530 situation, the verify now always succeeds.
3532 PH/25 Applied John Jetmore's -Mset patch.
3534 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3536 PH/27 In a string expansion for a processed (not raw) header when multiple
3537 headers of the same name were present, leading whitespace was being
3538 removed from all of them, but trailing whitespace was being removed only
3539 from the last one. Now trailing whitespace is removed from each header
3540 before concatenation. Completely empty headers in a concatenation (as
3541 before) are ignored.
3543 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3544 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3546 PH/29 [Removed. This was a change that I later backed out, and forgot to
3547 correct the ChangeLog entry (that I had efficiently created) before
3548 committing the later change.]
3550 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3551 address errors (4xx response to RCPT) over the same connection as other
3552 messages routed to the same hosts. Such deliveries are always "forced",
3553 so retry times are not inspected. This resulted in far too many retries
3554 for the affected addresses. The effect occurred only when there were more
3555 hosts than the hosts_max_try setting in the smtp transport when it had
3556 the 4xx errors. Those hosts that it had tried were not added to the list
3557 of hosts for which the message was waiting, so if all were tried, there
3558 was no problem. Two fixes have been applied:
3560 (i) If there are any address or message errors in an SMTP delivery, none
3561 of the hosts (tried or untried) are now added to the list of hosts
3562 for which the message is waiting, so the message should not be a
3563 candidate for sending over the same connection that was used for a
3564 successful delivery of some other message. This seems entirely
3565 reasonable: after all the message is NOT "waiting for some host".
3566 This is so "obvious" that I'm not sure why it wasn't done
3567 previously. Hope I haven't missed anything, but it can't do any
3568 harm, as the worst effect is to miss an optimization.
3570 (ii) If, despite (i), such a delivery is accidentally attempted, the
3571 routing retry time is respected, so at least it doesn't keep
3572 hammering the server.
3574 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3575 in ${readsocket because some servers need this prod.
3577 PH/32 Added some extra debug output when updating a wait-xxx database.
3579 PH/33 The hint "could be header name not terminated by colon", which has been
3580 given for certain expansion errors for a long time, was not being given
3581 for the ${if def:h_colon_omitted{... case.
3583 PH/34 The spec says: "With one important exception, whenever a domain list is
3584 being scanned, $domain contains the subject domain." There was at least
3585 one case where this was not true.
3587 PH/35 The error "getsockname() failed: connection reset by peer" was being
3588 written to the panic log as well as the main log, but it isn't really
3589 panic-worthy as it just means the connection died rather early on. I have
3590 removed the panic log writing for the ECONNRESET error when getsockname()
3593 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3594 runs only) independently of the message's sender address. This meant
3595 that, if the 4xx error was in fact related to the sender, a different
3596 message to the same recipient with a different sender could confuse
3597 things. In particular, this can happen when sending to a greylisting
3598 server, but other circumstances could also provoke similar problems.
3599 I have changed the default so that the retry time for these errors is now
3600 based a combination of the sender and recipient addresses. This change
3601 can be overridden by setting address_retry_include_sender=false in the
3604 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3605 remote server are returned as part of bounce messages. This was not
3606 happening for LMTP over a pipe (the lmtp transport), but now it is the
3607 same for both kinds of LMTP.
3609 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3610 in header lines that were in fact CNAMEs. This is no longer the case.
3612 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3613 and queue runs started by the daemon processed all messages. This has
3614 been fixed so that -R and -S can now usefully be given with -q<time>.
3616 PH/40 Import PCRE release 6.7 (fixes some bugs).
3618 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3620 PH/42 Give an error if -q is specified more than once.
3622 PH/43 Renamed the variables $interface_address and $interface_port as
3623 $received_ip_address and $received_port, to make it clear that these
3624 values apply to message reception, and not to the outgoing interface when
3625 a message is delivered. (The old names remain recognized, of course.)
3627 PH/44 There was no timeout on the connect() call when using a Unix domain
3628 socket in the ${readsocket expansion. There now is.
3630 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3631 be meaningful with "accept".
3633 SC/01 Eximstats V1.43
3634 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3636 SC/02 Eximstats V1.44
3637 Use a glob alias rather than an array ref in the generated
3638 parser. This improves both readability and performance.
3640 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3641 Collect SpamAssassin and rejection statistics.
3642 Don't display local sender or destination tables unless
3643 there is data to show.
3644 Added average volumes into the top table text output.
3646 SC/04 Eximstats V1.46
3647 Collect data on the number of addresses (recipients)
3648 as well as the number of messages.
3650 SC/05 Eximstats V1.47
3651 Added 'Message too big' to the list of mail rejection
3652 reasons (thanks to Marco Gaiarin).
3654 SC/06 Eximstats V1.48
3655 Mainlog lines which have GMT offsets and are too short to
3656 have a flag are now skipped.
3658 SC/07 Eximstats V1.49 (Alain Williams)
3659 Added the -emptyok flag.
3661 SC/08 Eximstats V1.50
3662 Fixes for obtaining the IP address from reject messages.
3664 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3665 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3666 whitespace changes from 4.64-PH/27
3668 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3671 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3672 are found, allow negative numbers in numeric criteria)
3674 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3676 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3677 to match changes made in 4.64-PH/43
3679 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3681 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3682 "domain" is actually an IP address, and force a failure. This locks out
3683 those revolvers/nameservers that support "A-for-A" lookups, in
3684 contravention of the specifications.
3686 PH/48 When a host name was looked up from an IP address, and the subsequent
3687 forward lookup of the name timed out, the host name was left in
3688 $sender_host_name, contrary to the specification.
3690 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3691 restricted to single-key lookups, Exim was not diagnosing an error if
3692 * or *@ was used with a query-style lookup.
3694 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3696 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3697 long ago, but noone interested enough thought of it. Let's just say that
3698 the "1.1" means that there are some new functions that weren't there at
3699 some point in the past.
3701 PH/51 Error processing for expansion failure of helo_data from an smtp
3702 transport during callout processing was broken.
3704 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3705 tested/used via the -bh/-bhc/-bs options.
3707 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3708 bug, fixed in subsequent PCRE releases).
3710 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3711 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3713 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3719 SC/01 Use a glob alias rather than an array ref in eximstats generated
3720 parser. This improves both readability and performance.
3722 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3723 Don't display local sender or destination tables in eximstats unless
3724 there is data to show.
3725 Added average volumes into the eximstats top table text output.
3727 SC/03 Collect data on the number of addresses (recipients) as well
3728 as the number of messages in eximstats.
3730 TF/01 Correct an error in the documentation for the redirect router. Exim
3731 does (usually) call initgroups() when daemonizing.
3733 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3734 with consistent privilege compared to when running as a daemon.
3736 TF/03 Note in the spec that $authenticated_id is not set for local
3737 submissions from trusted users.
3739 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3740 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3742 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3743 by adding some example configuration directives to the default
3744 configuration file. A little bit of work is required to uncomment the
3745 directives and define how usernames and passwords are checked, but
3746 there is now a framework to start from.
3748 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3749 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3750 without this. I don't know how relevant this is to other LDAP libraries.
3752 PH/02 Add the verb name to the "unknown ACL verb" error.
3754 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3756 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3758 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3759 directory not expanded when it should be if an expanded home directory
3760 was set for the address (which is overridden by the transport).
3762 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3765 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3766 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3767 because it is too late at that time, and has no effect.
3769 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3770 security issue with \' (bugzilla #107). I could not use the
3771 PQescapeStringConn() function, because it needs a PGconn value as one of
3774 PH/08 When testing addresses using -bt, indicate those final addresses that
3775 are duplicates that would not cause an additional delivery. At least one
3776 person was confused, thinking that -bt output corresponded to deliveries.
3777 (Suppressing duplicates isn't a good idea as you lose the information
3778 about possibly different redirections that led to the duplicates.)
3780 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3781 systems where poll() doesn't work, in particular OS X.
3783 PH/10 Added more information to debugging output for retry time not reached.
3785 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3786 operations in malware.c.
3788 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3791 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3792 syslog_duplication set false, log lines that would normally be written
3793 both the the main log and to the reject log were not written to syslog at
3796 PH/14 In the default configuration, change the use of "message" in ACL warn
3797 statements to "add_header".
3799 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3800 not followed by a command (e.g. "seen endif").
3802 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3803 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3806 PH/17 Added extra conditions to the default value of delay_warning_condition
3810 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3811 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3812 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3815 The Auto-Submitted: and various List- headers are standardised, whereas I
3816 don't think Precedence: ever was.
3818 PH/18 Refactored debugging code in route_finduser() to show more information,
3819 in particular, the error code if getpwnam() issues one.
3821 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3822 This is apparently needed in addition to the PH/07 change above to avoid
3823 any possible encoding problems.
3825 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3826 but not after initializing Perl.
3828 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3829 output them only if debugging. By default they are written stderr,
3830 apparently, which is not desirable.
3832 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3835 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3838 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3840 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3841 authenticated or an ident call has been made. Suppress the default
3842 values for $authenticated_id and $authenticated_sender (but permit -oMai
3843 and -oMas) when testing with -bh.
3845 PH/24 Re-jigged the order of the tests in the default configuration so that the
3846 tests for valid domains and recipients precede the DNS black list and CSA
3847 tests, on the grounds that those ones are more expensive.
3849 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3850 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3851 This bug exists in every version of Exim that I still have, right back to
3854 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3855 However, an attempt to turn on case-sensitivity in a regex key by
3856 including (?-i) didn't work because the subject string was already
3857 lowercased, and the effects were non-intuitive. It turns out that a
3858 one-line patch can be used to allow (?-i) to work as expected.
3864 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3865 other effects) broke the use of negated acl sub-conditions.
3867 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3870 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3871 "Deny" causes Exim to reject the incoming connection with a 554 error.
3872 Unfortunately, if there is a major crisis, such as a disk failure,
3873 tcp-wrappers gives "deny", whereas what one would like would be some
3874 kind of temporary error. A kludge has been added to help with this.
3875 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3876 554 error is used if errno is still zero or contains ENOENT (which occurs
3877 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3880 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3882 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3883 errors. Otherwise a message that provokes a temporary error (when other
3884 messages do not) can cause a whole host to time out.
3886 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3887 addresses were routed directly to files or pipes from a redirect router.
3888 File deliveries just didn't batch; pipe deliveries might have suffered
3891 PH/06 A failure to get a lock for a hints database would erroneously always say
3892 "Failed to get write lock", even when it was really a read lock.
3894 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3895 of 0600. This has been changed to use the value of the lockfile_mode
3896 option (which defaults to 0600).
3898 PH/08 Applied small patch from the Sieve maintainer.
3900 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3901 folder from quota calculations, a direct delivery into this folder messed
3902 up the contents of the maildirsize file. This was because the regex was
3903 used only to exclude .Trash (or whatever) when the size of the mailbox
3904 was calculated. There was no check that a delivery was happening into an
3905 excluded directory. This bug has been fixed by ignoring all quota
3906 processing for deliveries into excluded directories.
3908 PH/10 Added the maildirfolder_create_regex option to appendfile.
3914 PH/01 The code for finding all the local interface addresses on a FreeBSD
3915 system running IPv6 was broken. This may well have applied to all BSD
3916 systems, as well as to others that have similar system calls. The broken
3917 code found IPv4 interfaces correctly, but gave incorrect values for the
3918 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3919 that it would not match correctly against @[] and not recognize the IPv6
3922 PH/02 The ipliteral router was not recognizing addresses of the form user@
3923 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3925 PH/03 Added disable_ipv6.
3927 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3928 decoded form, because it is most often used to construct To: headers
3929 lines in autoreplies, and the decoded form may well be syntactically
3930 invalid. However, $reply_address has leading white space removed, and all
3931 newlines turned into spaces so that the autoreply transport does not
3934 PH/05 If group was specified without a user on a router, and no group or user
3935 was specified on a transport, the group from the router was ignored.
3937 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3938 for visible compile-time settings that can be used to change these
3939 numbers, for those that want even more. Backwards compatibility with old
3940 spool files has been maintained. However, going back to a previous Exim
3941 release will lost any variables that are in spool files.
3943 PH/07 Two small changes when running in the test harness: increase delay when
3944 passing a TCP/IP connection to a new process, in case the original
3945 process has to generate a bounce, and remove special handling of
3946 127.0.0.2 (sic), which is no longer necessary.
3948 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3949 be the same on different OS.
3951 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3954 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3955 whether --show-vars was specified or not
3957 JJ/02 exipick: Added support for new ACL variable spool format introduced
3960 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3961 syntactically invalid From: or Reply-to: line, and a filter used this to
3962 generate an autoreply, and therefore failed to obtain an address for the
3963 autoreply, Exim could try to deliver to a non-existent relative file
3964 name, causing unrelated and misleading errors. What now happens is that
3965 it logs this as a hard delivery error, but does not attempt to create a
3968 PH/11 The exinext utility has a -C option for testing purposes, but although
3969 the given file was scanned by exinext itself; it wasn't being passed on
3970 when Exim was called.
3972 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3973 an end-of-file indication when reading a command response.
3975 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3976 compiled. In many other places in Exim, IPv6 addresses are always
3977 recognized, so I have changed this. It also means that IPv4 domain
3978 literals of the form [IPV4:n.n.n.n] are now always recognized.
3980 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3981 used if the router is not running as root, for example, when verifying at
3982 ACL time, or when using -bh. The debugging output from this situation was
3983 non-existent - all you got was a failure to exec. I have made two
3986 (a) Failures to set uid/gid, the current directory, or a process leader
3987 in a subprocess such as that created by queryprogram now generate
3988 suitable debugging output when -d is set.
3990 (b) The queryprogram router detects when it is not running as root,
3991 outputs suitable debugging information if -d is set, and then runs
3992 the subprocess without attempting to change uid/gid.
3994 PH/15 Minor change to Makefile for building test_host (undocumented testing
3997 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3998 additional section of a DNS packet that returns MX or SRV records.
3999 Instead, it always explicitly searches for A/AAAA records. This avoids
4000 major problems that occur when a DNS server includes only records of one
4001 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4002 fixed another bug: if SRV records were looked up and the corresponding
4003 address records were *not* found in the additional section, the port
4004 values from the SRV records were lost.
4006 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4007 using the correct key (the original address) when searching the retry
4008 rules in order to find which one to use for generating the retry hint.
4010 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4011 adding the default one. Similarly, if it contains a Reply-To: header, the
4012 errors_reply_to option, if set, is not used.
4014 PH/19 When calculating a retry time, Exim used to measure the "time since
4015 failure" by looking at the "first failed" field in the retry record. Now
4016 it does not use this if it is later than than the arrival time of the
4017 message. Instead it uses the arrival time. This makes for better
4018 behaviour in cases where some deliveries succeed, thus re-setting the
4019 "first failed" field. An example is a quota failure for a huge message
4020 when small messages continue to be delivered. Without this change, the
4021 "time since failure" will always be short, possible causing more frequent
4022 delivery attempts for the huge message than are intended.
4023 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4025 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4026 $1, $2, $3) because the numerical variables can be reset during some
4027 expansion items (e.g. "match"), thereby losing the authentication data.
4029 PH/21 Make -bV show the size of off_t variables so that the test suite can
4030 decide whether to run tests for quotas > 2G.
4032 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4033 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4034 filecount value is greater than 2G or if a quota value is greater than 2G
4035 on a system where the size of off_t is not greater than 4, a panic error
4038 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4039 never match. The debug and -bh output now contains an explicit error
4040 message indicating a malformed IPv4 address or mask.
4042 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4043 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4044 PH/23 above applies.
4046 PH/25 Do not write to syslog when running in the test harness. The only
4047 occasion when this arises is a failure to open the main or panic logs
4048 (for which there is an explicit test).
4050 PH/26 Added the /no_tell option to "control=freeze".
4052 PH/27 If a host name lookup failed very early in a connection, for example, if
4053 the IP address matched host_lookup and the reverse lookup yielded a name
4054 that did not have a forward lookup, an error message of the form "no IP
4055 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4056 could be logged. Now it outputs the IP address instead of "NULL".
4058 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4059 allows the sender and the authenticated sender to be set when
4060 submitting a message from within Exim. Since child_open_exim() is
4061 documented for local_scan(), the new function should be too.
4063 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4064 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4065 results in an empty string is now treated as unset.
4067 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4069 PH/31 Added sender_verify_fail logging option.
4071 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4072 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4073 tidied the source and removed it altogether.
4075 PH/33 When a queue run was abandoned because the load average was too high, a
4076 log line was always written; now it is written only if the queue_run log
4077 selector is set. In addition, the log line for abandonment now contains
4078 information about the queue run such as the pid. This is always present
4079 in "start" and "stop" lines but was omitted from the "abandon" line.
4081 PH/34 Omit spaces between a header name and the colon in the error message that
4082 is given when verify = headers_syntax fails (if there are lots of them,
4083 the message gets confusing).
4085 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4086 names, as there are now some PTR records that contain slashes. This check
4087 is only to protect against broken name servers that fall over on strange
4088 characters, so the fact that it applies to all lookups doesn't matter.
4090 PH/36 Now that the new test suite is complete, we can remove some of the
4091 special code in Exim that was needed for the old test suite. For example,
4092 sorting DNS records because real resolvers return them in an arbitrary
4093 order. The new test suite's fake resolver always returns records in the
4096 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4097 bounces) except when queue_only is set, to avoid logging races between
4098 the different processes.
4100 PH/38 Panic-die if .include specifies a non-absolute path.
4102 PH/39 A tweak to the "H" retry rule from its user.
4104 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4105 a label. They prevented compilation on older perls.
4107 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4108 a warning to be raised on newish perls.
4110 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4111 on queue. Changes to match documented behaviour of showing count of
4112 messages matching specified criteria.
4114 PH/40 Changed the default ident timeout from 30s to 5s.
4116 PH/41 Added support for the use of login_cap features, on those BSD systems
4117 that have them, for controlling the resources used by pipe deliveries.
4119 PH/42 The content-scanning code uses fopen() to create files in which to put
4120 message data. Previously it was not paying any attention to the mode of
4121 the files. Exim runs with umask(0) because the rest of the code creates
4122 files with open(), and sets the required mode explicitly. Thus, these
4123 files were ending up world-writeable. This was not a big issue, because,
4124 being within the spool directory, they were not world-accessible. I have
4125 created a function called modefopen, which takes an additional mode
4126 argument. It sets umask(777), creates the file, chmods it to the required
4127 mode, then resets the umask. All the relevant calls to fopen() in the
4128 content scanning code have been changed to use this function.
4130 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4131 to 24 hours. This avoids potential overflow problems when processing G
4132 and H retry rules. I suspect nobody ever tinkers with this value.
4134 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4136 PH/45 When the plaintext authenticator is running as a client, the server's
4137 challenges are checked to ensure they are valid base64 strings. By
4138 default, the authentication attempt is cancelled if an invalid string is
4139 received. Setting client_ignore_invalid_base64 true ignores these errors.
4140 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4141 they are received. Thus, the responses can be made to depend on the
4142 challenges. If an invalid string is ignored, an empty string is placed in
4145 PH/46 Messages that are created by the autoreply transport now contains a
4146 References: header, in accordance with RFCs 2822 and 3834.
4148 PH/47 Added authenticated_sender_force to the smtp transport.
4150 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4152 PH/49 Installed latest patch from the Sieve maintainer.
4154 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4155 was also set, the appendfile transport was unnecessarily scanning a
4156 directory of message files (e.g. for maildir delivery) to find the count
4157 of files (along with the size), even though it did not need this
4158 information. It now does the scan only if it needs to find either the
4159 size of the count of files.
4161 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4163 PH/52 Two bugs concerned with error handling when the smtp transport is
4166 (i) Exim was not creating retry information for temporary errors given
4167 for individual recipients after the DATA command when the smtp transport
4168 was used in LMTP mode. This meant that they could be retried too
4169 frequently, and not timed out correctly.
4171 (ii) Exim was setting the flag that allows error details to be returned
4172 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4173 recipients that were returned after the DATA command.
4175 PH/53 This is related to PH/52, but is more general: for any failing address,
4176 when detailed error information was permitted to be returned to the
4177 sender, but the error was temporary, then after the final timeout, only
4178 "retry timeout exceeded" was returned. Now it returns the full error as
4179 well as "retry timeout exceeded".
4181 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4182 do this, and (what is worse) MTAs that accept it.
4184 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4185 will now be deprecated.
4187 PH/56 New os.c-cygwin from the Cygwin maintainer.
4189 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4190 formats (previously only available in exim formats via -bpr, -bpru,
4191 and -bpra. Now also available in native and exiqgrep formats)
4193 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4194 with very large, slow to parse queues
4196 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4198 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4200 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4201 responses to authentication challenges, though it was showing the
4202 challenges; (ii) I've removed the CR characters from the debug output for
4205 PH/58 Allow for the insertion of a newline as well as a space when a string
4206 is turned into more than one encoded-word during RFC 2047 encoding. The
4207 Sieve code now uses this.
4209 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4210 data_4xx, lost_connection, tls_required.
4212 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4213 message was being sent as an SMTP response.
4215 PH/61 Add -l and -k options to exicyclog.
4217 PH/62 When verifying, if an address was redirected to one new address, so that
4218 verification continued, and the new address failed or deferred after
4219 having set something in $address_data, the value of $address_data was not
4220 passed back to the ACL. This was different to the case when no
4221 redirection occurred. The value is now passed back in both cases.
4223 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4224 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4225 use login_cap.h, so on its own it isn't the distinguishing feature. The
4226 new name refers directly to the setclassresources() function.
4228 PH/65 Added configuration files for NetBSD3.
4230 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4232 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4233 is preferred over IPv4.
4235 PH/68 The bounce_return_message and bounce_return_body options were not being
4236 honoured for bounces generated during the reception of non-SMTP messages.
4237 In particular, this applied to messages rejected by the ACL. This bug has
4238 been fixed. However, if bounce_return_message is true and bounce_return_
4239 body is false, the headers that are returned for a non-SMTP message
4240 include only those that have been read before the error was detected.
4241 (In the case of an ACL rejection, they have all been read.)
4243 PH/69 The HTML version of the specification is now built in a directory called
4244 spec_html instead of spec.html, because the latter looks like a path with
4245 a MIME-type, and this confuses some software.
4247 PH/70 Catch two compiler warnings in sieve.c.
4249 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4250 function verify_get_ident() calls ip_connect() to connect a socket, but
4251 if the "connect()" function timed out, ip_connect() used to close the
4252 socket. However, verify_get_ident() also closes the socket later, and in
4253 between Exim writes to the log, which may get opened at this point. When
4254 the socket was closed in ip_connect(), the log could get the same file
4255 descriptor number as the socket. This naturally causes chaos. The fix is
4256 not to close the socket in ip_connect(); the socket should be closed by
4257 the function that creates it. There was only one place in the code where
4258 this was missing, in the iplookup router, which I don't think anybody now
4259 uses, but I've fixed it anyway.
4261 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4262 well as to direct DNS lookups. Otherwise the handling of names in host
4263 lists is inconsistent and therefore confusing.
4269 PH/01 Two changes to the default runtime configuration:
4271 (1) Move the checks for relay_from_hosts and authenticated clients from
4272 after to before the (commented out) DNS black list checks.
4274 (2) Add control=submission to the relay_from_hosts and authenticated
4275 clients checks, on the grounds that messages accepted by these
4276 statements are most likely to be submissions.
4278 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4280 (1) Generate an error if the third argument for the ${prvs expansion is
4283 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4286 (3) Reset the variables that are obtained from the first argument of
4287 ${prvscheck and used in the second argument before leaving the code,
4288 because their memory is reclaimed, so using them afterwards may do
4291 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4292 one (it's much easier than Tom thought :-).
4294 (5) Because of (4), we can now allow for the use of $prvscheck_result
4295 inside the third argument.
4297 PH/03 For some reason, the default setting of PATH when running a command from
4298 a pipe transport was just "/usr/bin". I have changed it to
4301 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4302 anything to be listed in the output from -bV.
4304 PH/05 When a filter generated an autoreply, the entire To: header line was
4305 quoted in the delivery log line, like this:
4307 => >A.N.Other <ano@some.domain> <original@ddress> ...
4309 This has been changed so that it extracts the operative address. There
4310 may be more than one such address. If so, they are comma-separated, like
4313 => >ano@some.domain,ona@other.domain <original@ddress> ...
4315 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4316 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4317 not being looked up in the rDNS to get a host name, Exim was showing the
4318 IP address twice in Received: lines, even though the IP addresses were
4319 identical. For example:
4321 Received: from [1.2.3.4] (helo=[1.2.3.4])
4323 However, if the real host name was known, it was omitting the HELO data
4324 if it matched the actual IP address. This has been tidied up so that it
4325 doesn't show the same IP address twice.
4327 PH/07 When both +timestamp and +memory debugging was on, the value given by
4328 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4329 called by the debug printing, thereby overwriting the timestamp buffer.
4330 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4332 PH/08 When the original message was included in an autoreply transport, it
4333 always said "this is a copy of the message, including all the headers",
4334 even if body_only or headers_only was set. It now gives an appropriate
4337 PH/09 Applied a patch from the Sieve maintainer which:
4339 o fixes some comments
4340 o adds the (disabled) notify extension core
4341 o adds some debug output for the result of if/elsif tests
4342 o points to the current vacation draft in the documentation
4343 and documents the missing references header update
4347 o fixes a bug in processing the envelope test (when testing
4348 multiple envelope elements, the last element determined the
4351 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4352 Electronic Mail") by including:
4354 Auto-submitted: auto-generated
4356 in the messages that it generates (bounce messages and others, such as
4357 warnings). In the case of bounce messages for non-SMTP messages, there was
4358 also a typo: it was using "Auto_submitted" (underscore instead of
4359 hyphen). Since every message generated by Exim is necessarily in response
4360 to another message, thes have all been changed to:
4362 Auto-Submitted: auto-replied
4364 in accordance with these statements in the RFC:
4366 The auto-replied keyword:
4368 - SHOULD be used on messages sent in direct response to another
4369 message by an automatic process,
4371 - MUST NOT be used on manually-generated messages,
4373 - MAY be used on Delivery Status Notifications (DSNs) and Message
4374 Disposition Notifications (MDNs),
4376 - MUST NOT be used on messages generated by automatic or periodic
4377 processes, except for messages which are automatic responses to
4380 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4381 to the default Received: header definition.
4383 PH/12 Added log selector acl_warn_skipped (default on).
4385 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4386 variables because (a) they are in the wrong storage pool and (b) even if
4387 they were copied, it wouldn't work properly because of the caching.
4389 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4390 checking when decoding. Apparently there are clients that generate
4391 overlong encoded strings. Why am I not surprised?
4393 PH/15 If the first argument of "${if match_address" was not empty, but did not
4394 contain an "@" character, Exim crashed. Now it writes a panic log message
4395 and treats the condition as false.
4397 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4399 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4400 extension "envelope-auth". The code is finished and in agreement with
4401 other implementations, but there is no documentation so far and in fact,
4402 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4403 not changing the active code.
4405 Print executed "if" and "elsif" statements when debugging is used. This
4406 helps a great deal to understand what a filter does.
4408 Document more things not specified clearly in RFC3028. I had all this
4409 sorted out, when out of a sudden new issues came to my mind. Oops."
4411 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4414 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4415 canonical form (as documented). However, after a host name lookup from
4416 the IP address, check_host() was doing a simple string comparison with
4417 addresses acquired from the DNS when checking that the found name did
4418 have the original IP as one of its addresses. Since any found IPv6
4419 addresses are likely to be in abbreviated form, the comparison could
4420 fail. Luckily, there already exists a function for doing the comparison
4421 by converting both addresses to binary, so now that is used instead of
4422 the text comparison.
4424 PH/20 There was another similar case to PH/19, when a complete host name was
4425 given in a host list; looking up its IP address could give an abbreviated
4426 form, whereas the current host's name might or might not be abbreviated.
4427 The same fix has been applied.
4433 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4434 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4437 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4438 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4440 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4442 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4443 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4444 Also, for "Auto-Submitted", treat anything other than "no" as
4445 non-personal, in accordance with RFC 3834. (Previously it treated
4446 anything starting "auto-" as non-personal.)
4448 TF/01 The control=submission/name=... option had a problem with syntax
4449 errors if the name included a slash character. The /name= option
4450 now slurps the rest of the string, so it can include any characters
4451 but it must come last in the list of options (after /sender_retain
4454 PH/05 Some modifications to the interface to the fake nameserver for the new
4462 TK/01 Added the "success_on_redirect" address verification option. See
4463 NewStuff for rationale and an example.
4465 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4467 PH/02 Patch to exigrep to allow it to work on syslog lines.
4469 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4470 fread() to skip over the body file's header line, because in Cygwin the
4471 header line is locked and is inaccessible.
4473 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4474 co-exist for some time) to make it clear that it is the Exim ID that is
4475 referenced, not the Message-ID: header line.
4477 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4478 string_format() function, because snprintf() does not exist on all
4481 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4482 ${stat: expansion item.
4484 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4485 protocol synchronization error", to keep the pedants happy.
4487 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4488 well as for IRIX systems, when gcc is being used. See the host.c source
4491 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4493 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4496 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4497 transport and to the smtp transport in LMTP mode.
4499 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4501 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4502 run a filter in a subprocess. This could lead to confusion in subsequent
4503 lookups in the parent process. There should also be a search_tidyup() at
4504 the end of the subprocess.
4506 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4507 only if the host matched helo_try_verify_hosts, which caused the
4508 verification to occur when the EHLO/HELO command was issued. The ACL just
4509 tested the remembered result. Now, if a previous verification attempt has
4510 not happened, "verify = helo" does it there and then.
4512 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4514 TK/03 Fix log output including CR from clamd.
4516 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4517 exist provoked a memory error which could cause a segfault.
4519 PH/15 Installed PCRE 6.2
4521 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4523 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4524 of the problem. Specifically, suggested +O2 rather than +O1 for the
4527 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4529 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4530 record, the port was not correctly logged when the outgoing_port log
4531 selector was set (it logged the transort's default port).
4533 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4534 fallback_hosts, and "hosts" in the smtp transport.
4536 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4537 host errors such as "Connection refused".
4539 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4540 authentication with radiusclient 0.4.9:
4542 - Error returned from rc_read_config was caught wrongly
4543 - Username/password not passed on to radius server due to wrong length.
4545 The presumption is that some radiusclient API changes for 4.51/PH/17
4546 were not taken care of correctly. The code is still untested by me (my
4547 Linux distribution still has 0.3.2 of radiusclient), but it was
4548 contributed by a Radius user.
4550 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4551 expanding the "port" option of the smtp transport.
4553 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4554 while reading a MIME header. Thanks to Tom Hughes for a patch.
4556 PH/24 Include config.h inside local_scan.h so that configuration settings are
4559 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4560 that in an ACL for RCPT (for example), you can examine exactly what was
4563 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4564 commands, but it was not correctly comparing the address with the actual
4565 client host address. Thus, it would show the EHLO address in Received:
4566 header lines when this was not necessary.
4568 PH/27 Added the % operator to ${eval:}.
4570 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4571 it should be ignoring failures (because with -C, for example, it has lost
4572 privilege). It wasn't ignoring creation failures other than "already
4575 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4578 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4579 because an input line was too long, either on its own, or by virtue of
4580 too many continuations, the temporary file was not being removed, and the
4581 return code was incorrect.
4583 PH/31 Missing "BOOL" in function definition in filtertest.c.
4585 PH/32 Applied Sieve patches from the maintainer.
4587 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4589 PH/33 Added "verify = not_blind".
4591 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4592 Local/Makefile (with some defaults set). These are used in built scripts
4593 such as exicyclog, but they have never been used in the exim_install
4594 script (though there are many overriding facilities there). I have
4595 arranged that the exim_install script now takes note of these two
4598 PH/35 Installed configuration files for Dragonfly.
4600 PH/36 When a locally submitted message by a trusted user did not contain a
4601 From: header, and the sender address was obtained from -f or from an SMTP
4602 MAIL command, and the trusted user did not use -F to supply a sender
4603 name, $originator_name was incorrectly used when constructing a From:
4604 header. Furthermore, $originator_name was used for submission mode
4605 messages from external hosts without From: headers in a similar way,
4606 which is clearly wrong.
4608 PH/37 Added control=suppress_local_fixups.
4610 PH/38 When log_selector = +received_sender was set, and the addition of the
4611 sender made the log line's construction buffer exactly full, or one byte
4612 less than full, an overflow happened when the terminating "\n" was
4615 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4616 when the result of a list match is failure because a DNS lookup failed.
4618 PH/40 RM_COMMAND is now used in the building process.
4620 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4621 the "build-* directories that it finds.
4623 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4624 address in a domain literal was a prefix of an interface address.
4626 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4627 when verifying a sender address, unless rewrite_headers is false.
4629 PH/44 Wrote a long comment about why errors_to addresses are verified as
4630 recipients, not senders.
4632 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4633 the ratelimit ACL was added.
4635 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4637 PH/46 Added extra information about PostgreSQL errors to the error string.
4639 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4640 avoiding the need to install special zones in a real server. This is
4641 backwards compatible; if it can't find the fake resolver, it drops back.
4642 Thus, both old and new test suites can be run.
4644 TF/02 Added util/ratelimit.pl
4646 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4647 clock is set back in time.
4649 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4650 Candler <B.Candler@pobox.com>.
4652 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4653 OK for addresses that are the result of redirections.
4655 PH/48 A number of further additions for the benefit of the new test suite,
4656 including a fake gethostbyname() that interfaces to the fake DNS resolver
4659 TF/06 The fix for widen_domains has also been applied to qualify_single and
4660 search_parents which are the other dnslookup options that can cause
4663 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4666 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4668 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4669 service" if it encountered an MX record with an empty target hostname.
4670 The message is now "an MX or SRV record indicated no SMTP service".
4672 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4673 if the DNS of the sending site is misconfigured. This is quite a
4674 common situation. This change restores the behaviour of treating a
4675 helo verification defer as a failure.
4677 PH/51 If self=fail was set on a router, the bounce message did not include the
4678 actual error message.
4684 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4686 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4687 command itself ended in error, the underlying message about the transport
4688 filter timeout was being overwritten with the pipe command error. Now the
4689 underlying error message should be appended to the second error message.
4691 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4693 PH/02 Reduce the amount of output that "make" produces by default. Full output
4694 can still be requested.
4696 PH/03 The warning log line about a condition test deferring for a "warn" verb
4697 was being output only once per connection, rather than after each
4698 occurrence (because it was using the same function as for successful
4699 "warn" verbs). This seems wrong, so I have changed it.
4701 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4702 it should not have, which might have caused a crash in the right
4703 circumstances, but probably never did.
4705 PH/04 Installed a modified version of Tony Finch's patch to make submission
4706 mode fix the return path as well as the Sender: header line, and to
4707 add a /name= option so that you can make the user's friendly name appear
4710 TF/03 Added the control = fakedefer ACL modifier.
4712 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4713 Mark Lowes for thorough testing.
4715 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4717 TK/03 Merged latest SRS patch from Miles Wilton.
4719 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4720 with the definition in sysexits.h (which is #included earlier).
4721 Fortunately, Exim does not actually use EX_OK. The code used to try to
4722 preserve the sysexits.h value, by assuming that macro definitions were
4723 scanned for macro replacements. I have been disabused of this notion,
4724 so now the code just undefines EX_OK before #including unistd.h.
4726 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4727 in the smtp transport. When a block could not be written in a single
4728 write() function, the timeout was being re-applied to each part-write.
4729 This seems wrong - if the receiver was accepting one byte at a time it
4730 would take for ever. The timeout is now adjusted when this happens. It
4731 doesn't have to be particularly precise.
4733 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4734 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4736 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4737 without a domain if the check to <postmaster@domain> fails.
4739 SC/01 Eximstats: added -xls and the ability to specify output files
4740 (patch written by Frank Heydlauf).
4742 SC/02 Eximstats: use FileHandles for outputting results.
4744 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4746 SC/04 Eximstats: fixed display of large numbers with -nvr option
4748 SC/05 Eximstats: fixed merging of reports with empty tables.
4750 SC/06 Eximstats: added the -include_original_destination flag
4752 SC/07 Eximstats: removed tabs and trailing whitespace.
4754 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4756 TK/06 MBOX spool code: Add real "From " MBOX separator line
4757 so the .eml file is really in mbox format (even though
4758 most programs do not really care). Patch from Alex Miller.
4760 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4761 The latter is generated from $received_to and is only set if the
4762 message has one envelope recipient. SA can use these headers,
4763 obviously out-of-the-box. Patch from Alex Miller.
4765 PH/08 The ${def test on a variable was returning false if the variable's
4766 value was "0", contrary to what the specification has always said!
4767 The result should be true unless the variable is empty.
4769 PH/09 The syntax error of a character other than { following "${if
4770 def:variable_name" (after optional whitespace) was not being diagnosed.
4771 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4772 accidental colon was present, for example, could give incorrect results.
4774 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4775 result is used (not including appendfile, where other changes are about
4778 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4779 This involved changing a lot of size variables from int to off_t. It
4780 should work with maildirs and everything.
4782 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4783 spamd dying while we are connected to it.
4785 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4788 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4789 function for BDB 4.3.
4791 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4793 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4794 the sources for PCRE have been re-arranged and more files are now
4797 PH/15 The code I had for printing potentially long long variables in PH/11
4798 above was not the best (it lost precision). The length of off_t variables
4799 is now inspected at build time, and an appropriate printing format (%ld
4800 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4801 to be "long long int" or "long int". This is needed for the internal
4802 formatting function string_vformat().
4804 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4805 the configuration file to be ":syslog", then the script "guesses" where
4806 the logs files are, rather than using the compiled in default. In our
4807 case the guess is not the same as the compiled default, so the script
4808 suddenly stopped working when I started to use syslog. The patch checks
4809 to see if log_file_path is "". If so, it attempts to read it from exim
4810 with no configuration file to get the compiled in version, before it
4811 falls back to the previous guessing code."
4813 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4814 implementing BATV in an Exim configuration. See NewStuff for the gory
4817 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4818 Makefile that are specific to HP-UX.
4820 PH/18 If the "use_postmaster" option was set for a recipient callout together
4821 with the "random" option, the postmaster address was used as the MAIL
4822 FROM address for the random test, but not for the subsequent recipient
4823 test. It is now used for both.
4825 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4826 patch removes a few documentation additions to RFC 3028, because the
4827 latest draft now contains them. It adds the new en;ascii-case comparator
4828 and a new error check for 8bit text in MIME parts. Comparator and
4829 require names are now matched exactly. I enabled the subaddress
4830 extension, but it is not well tested yet (read: it works for me)."
4832 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4833 rework some of the code of TK/09 above to avoid the hardwired use of
4834 "%lld" and "long long". Replaced the call to snprintf() with a call to
4837 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4838 records point to non-existent hosts", "retry timeout exceeded", and
4839 "retry time not reached for any host after a long failure period".
4841 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4842 experimental DomainKeys support:
4844 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4845 (2) On an error such as an illegally used "control", the wrong name for
4846 the control was given.
4848 These problems did NOT occur unless DomainKeys support was compiled.
4850 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4852 PH/24 Added ${if match_ip condition.
4854 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4855 databases so that it will be absolutely obvious if a crash occurs in the
4856 DB library. This is a regular occurrence (often caused by mis-matched
4859 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4860 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4861 picked up on a user's system that detects such things. There doesn't seem
4862 to be a gcc warning option for this - only an attribute that has to be
4863 put on the function's prototype. It seems that in Fedora Core 4 they have
4864 set this on a number of new functions. No doubt there will be more in due
4867 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4868 specify a transport. However, if an address that was verified by such a
4869 router was the subject of a callout, Exim crashed because it tried to
4870 read the rcpt_include_affixes from the non-existent transport. Now it
4871 just assumes that the setting of that option is false. This bug was
4872 introduced by 4.51/PH/31.
4874 PH/28 Changed -d+all to exclude +memory, because that information is very
4875 rarely of interest, but it makes the output a lot bigger. People tend to
4876 do -d+all out of habit.
4878 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4879 code in os-type was giving problems when libc.so lives in lib64, like on
4882 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4883 aren't the modern standard, and it seems that some systems' include files
4884 don't always have them. Exim was already checking for some of the newer
4885 ones like T_AAAA, and defining it itself. I've added checks for all the
4886 record types that Exim uses.
4888 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4889 not automatically generating a new one, as it is supposed to. This
4890 prevented TLS from working. If the file did exist, but contained invalid
4891 data, a new version was generated, as expected. It was only the case of a
4892 non-existent file that was broken.
4894 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4895 with a change in libdomainkeys > 0.64.
4897 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4898 from DNS. If the selector record carries the flag, it now has
4899 precedence over the domain-wide flag.
4901 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4903 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4904 the use of a port name that isn't defined in /etc/services) occurred, the
4905 message was deferred as in a normal delivery, and thus remained on the
4906 spool, instead of being failed because of the mua_wrapper setting. This
4907 is now fixed, and I tidied up some of the mua_wrapper messages at the
4910 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4911 the messages in a hash of arrays rather than using individual hashes.
4912 This is a bit cleaner and results in dramatic memory savings, albeit
4913 at a slight CPU cost.
4915 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4916 as requested by Marc Sherman.
4918 SC/10 Eximstats: added histograms for user specified patterns as requested
4921 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4923 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4924 fopen() in the content-scanning modules that did not already have it.
4930 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4931 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4933 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4935 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4937 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4938 to test Sieve filters that use "vacation".
4940 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4941 that changes the way the GnuTLS parameters are stored in the cache file.
4942 The new format can be generated externally. For backward compatibility,
4943 if the data in the cache doesn't make sense, Exim assumes it has read an
4944 old-format file, and it generates new data and writes a new file. This
4945 means that you can't go back to an older release without removing the
4948 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4949 work if there are any delivery delays because "one_time" forces the
4950 parent to be marked "delivered", so its unseen clone is never tried
4951 again. For this reason, Exim now forbids the simultaneous setting of
4954 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4955 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4956 entry if you want to know the details. The fix, however, neglected to
4957 consider the case when local delivery batching is involved. The test for
4958 "previously delivered" was not happening when checking to see if an
4959 address could be batched with a previous (undelivered) one; under
4960 certain circumstances this could lead to multiple deliveries to the same
4963 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4964 in its include files, and this causes problems building Exim.
4966 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4967 header_syntax) but Exim was just ignoring anything given after a slash.
4968 In particular, this caused confusion with an attempt to use "verify =
4969 reverse_host_lookup/defer_ok". An error is now given when options are
4970 supplied for verify items that do not have them. (Maybe reverse_host_
4971 lookup should have a defer_ok option, but that's a different point.)
4973 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4974 defined by RFC 821) to 2048, because there were problems with some AUTH
4975 commands, and RFC 1869 says the size should be increased for extended
4976 SMTP commands that take arguments.
4978 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4981 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4982 "unknown" error; now it says that the functionality isn't in the binary.
4984 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4985 an address' error message when a string expansion fails (syntax or
4986 whatever). Otherwise the password may appear in the log. Following change
4987 PH/42 below, there is no longer a chance of it appearing in a bounce
4990 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4992 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4993 of its addresses was ever tried. (Bugzilla bug #2.)
4995 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4996 the result incorrectly in the debug output. (It correctly added a newline
4997 to what was transported.)
4999 TF/01 Added $received_time.
5001 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5002 commented out examples of how to interface to a virus scanner and to
5003 SpamAssassin. Also added commented examples of av_scanner and
5004 spamd_address settings.
5006 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5007 and controls are allowed in which ACLs. There were a couple of minor
5008 errors. Some of the entries in the conditions table (which is a table of
5009 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5010 negation of where the condition IS allowed.
5012 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5014 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5015 header file does not have a version number, so I've had to invent a new
5016 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5017 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5018 radiusclient), but it was contributed by a Radius user.
5020 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5021 files or directories, for OpenSSL.
5023 PH/19 When an Exim process that is running as root has to create an Exim log
5024 file, it does so in a subprocess that runs as exim:exim so as to get the
5025 ownership right at creation (otherwise, other Exim processes might see
5026 the file with the wrong ownership). There was no test for failure of this
5027 fork() call, which would lead to the process getting stuck as it waited
5028 for a non-existent subprocess. Forks do occasionally fail when resources
5029 run out. I reviewed all the other calls to fork(); they all seem to check
5032 PH/20 When checking for unexpected SMTP input at connect time (before writing
5033 the banner), Exim was not dealing correctly with a non-positive return
5034 from the read() function. If the client had disconnected by this time,
5035 the result was a log entry for a synchronization error with an empty
5036 string after "input=" when read() returned zero. If read() returned -1
5037 (an event I could not check), uninitialized data bytes were printed.
5038 There were reports of junk text (parts of files, etc) appearing after
5041 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5043 PH/22 Added support for macro redefinition, and (re)definition in between
5044 driver and ACL definitions.
5046 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5047 forgetting to use the resulting value; it was using the unexpanded value.
5049 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5050 hadn't been configured. The fix is from Juergen Kreileder, who
5051 understands it better than I do:
5053 "Here's what I see happening with three configured cyrus_sasl
5054 authenticators configured (plain, login, cram-md5):
5056 On startup auth_cyrus_sasl_init() gets called for each of these.
5057 This means three calls to sasl_listmech() without a specified mech_list.
5058 => SASL tests which mechs of all available mechs actually work
5059 => three warnings about OTP not working
5060 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5062 With the patch, sasl_listmech() also gets called three times. But now
5063 SASL's mech_list option is set to the server_mech specified in the the
5064 authenticator. Or in other words, the answer from sasl_listmech()
5065 gets limited to just the mech you're testing for (which is different
5067 => the return list contains just 'plain' or 'login', 'cram-md5' or
5068 nothing depending on the value of ob->server_mech.
5070 I've just tested the patch: Authentication still works fine,
5071 unavailable mechs specified in the exim configuration are still
5072 caught, and the auth.log warnings about OTP are gone."
5074 PH/25 When debugging is enabled, the contents of the command line are added
5075 to the debugging output, even when log_selector=+arguments is not
5078 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5079 answer is "GNU", and only if the return is "GNU/something" is the answer
5082 PH/27 $acl_verify_message is now set immediately after the failure of a
5083 verification in an ACL, and so is available in subsequent modifiers. In
5084 particular, the message can be preserved by coding like this:
5086 warn !verify = sender
5087 set acl_m0 = $acl_verify_message
5089 Previously, $acl_verify_message was set only while expanding "message"
5090 and "log_message" when a very denied access.
5092 PH/28 Modified OS/os.c-Linux with
5094 -#ifndef OS_LOAD_AVERAGE
5095 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5097 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5098 nomenclature these days.)
5100 PH/29 Installed patch from the Sieve maintainer that adds the options
5101 sieve_useraddress and sieve_subaddress to the redirect router.
5103 PH/30 In these circumstances:
5104 . Two addresses routed to the same list of hosts;
5105 . First host does not offer TLS;
5106 . First host accepts first address;
5107 . First host gives temporary error to second address;
5108 . Second host offers TLS and a TLS session is established;
5109 . Second host accepts second address.
5110 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5111 and peerdn, if requested) that were in fact used only for the second
5114 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5115 attention to any local part prefix or suffix that was matched by the
5116 router that accepted the address. It now behaves in the same way as it
5117 does for delivery: the affixes are removed from the local part unless
5118 rcpt_include_affixes is set on the transport.
5120 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5121 timeout during the DATA phase of an incoming message.
5123 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5124 applied a patch sanctioned by the Sieve maintainer.
5126 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5127 the uid or gid is negative. A case of a negative gid caused this to be
5128 noticed. The fix allows for either to be negative.
5130 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5131 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5134 PH/36 The change PH/12 above was broken. Fixed it.
5136 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5137 the grounds that routing the same address twice would always produce the
5138 same answer. This might have been true once, but it is certainly no
5139 longer true now. Routing a child address may depend on the previous
5140 routing that produced that child. Some complicated redirection strategies
5141 went wrong when messages had multiple recipients, and made Exim's
5142 behaviour dependent on the order in which the addresses were given.
5144 I have moved the duplicate checking until after the routing is complete.
5145 Exim scans the addresses that are assigned to local and remote
5146 transports, and removes any duplicates. This means that more work will be
5147 done, as duplicates will always all be routed, but duplicates are
5148 presumably rare, so I don't expect this is of any significance.
5150 For deliveries to pipes, files, and autoreplies, the duplicate checking
5151 still happens during the routing process, since they are not going to be
5154 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5155 It corrects a timeout issue with spamd. This is Ian's comment: "The
5156 background is that sometimes spamd either never reads data from a
5157 connection it has accepted, or it never writes response data. The exiscan
5158 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5159 blindly assumes that writes won't block so it may never time out."
5161 PH/39 Allow G after quota size as well as K and M.
5163 PH/40 The value set for $authenticated_id in an authenticator may not contain
5164 binary zeroes or newlines because the value is written to log lines and
5165 to spool files. There was no check on this. Now the value is run through
5166 the string_printing() function so that such characters are converted to
5167 printable escape sequences.
5169 PH/41 $message_linecount is a new variable that contains the total number of
5170 lines in the message. Compare $body_linecount, which is the count for the
5173 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5174 bounce and delay warning messages, except in certain special cases, which
5177 (a) An SMTP error message from a remote host;
5178 (b) A message specified in a :fail: redirection;
5179 (c) A message specified in a "fail" command in a system filter;
5180 (d) A message specified in a FAIL return from the queryprogram router;
5181 (e) A message specified by the cannot_route_message router option.
5183 In these cases only, Exim does include the error details in bounce and
5184 warning messages. There are also a few cases where bland messages such
5185 as "unrouteable address" or "local delivery error" are given.
5187 PH/43 $value is now also set for the "else" part of a ${run expansion.
5189 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5190 being worked on, but at least Exim now implements the latest version to
5193 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5194 process to complete was treated as a delivery failure, a timeout while
5195 writing the message to the pipe was logged, but erroneously treated as a
5196 successful delivery. Such timeouts include transport filter timeouts. For
5197 consistency with the overall process timeout, these timeouts are now
5198 treated as errors, giving rise to delivery failures by default. However,
5199 there is now a new Boolean option for the pipe transport called
5200 timeout_defer, which, if set TRUE, converts the failures into defers for
5201 both kinds of timeout. A transport filter timeout is now identified in
5204 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5205 systems where "make" and "gmake" are different, calling "gmake" at top
5206 level broke things. I've arranged for the value of $(MAKE) to be passed
5207 from the Makefile to this script so that it can call the same version of
5211 A note about Exim versions 4.44 and 4.50
5212 ----------------------------------------
5214 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5215 changes of various kinds. As a consequence, a big documentation update was
5216 needed. This delayed the release for rather longer than seemed good, especially
5217 in the light of a couple of (minor) security issues. Therefore, the changes
5218 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5219 release. So 4.44 and 4.50 are in effect two different branches that both start
5222 I have left the 4.50 change log unchanged; it contains all the changes since
5223 4.43. The change log for 4.44 is below; many of its items are identical to
5224 those for 4.50. This seems to be the most sensible way to preserve the
5225 historical information.
5231 1. Minor wording change to the doc/README.SIEVE file.
5233 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5234 computation of the current number of files was incorrect.
5236 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5237 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5240 4. Give more explanation in the error message when the command for a transport
5241 filter fails to execute.
5243 5. There are several places where Exim runs a non-Exim command in a
5244 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5245 was being done only for the command run by the queryprogram router. It is
5246 now done for all such subprocesses. The other cases are: ${run, transport
5247 filters, and the commands run by the lmtp and pipe transports.
5249 6. Added CONFIGURE_GROUP build-time option.
5251 7. Some older OS have a limit of 256 on the maximum number of file
5252 descriptors. Exim was using setrlimit() to set 1000 as a large value
5253 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5254 systems. I've change it so that if it can't get 1000, it tries for 256.
5256 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5257 was an oversight, and furthermore, ever since the addition of extra
5258 controls (e.g. 4.43/32), the checks on when to allow different forms of
5259 "control" were broken. There should now be diagnostics for all cases when a
5260 control that does not make sense is encountered.
5262 9. Added the /retain_sender option to "control=submission".
5264 10. $recipients is now available in the predata ACL (oversight).
5266 11. Tidy the search cache before the fork to do a delivery from a message
5267 received from the command line. Otherwise the child will trigger a lookup
5268 failure and thereby defer the delivery if it tries to use (for example) a
5269 cached ldap connection that the parent has called unbind on.
5271 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5272 of $address_data from the recipient verification was clobbered by the
5273 sender verification.
5275 13. The value of address_data from a sender verification is now available in
5276 $sender_address_data in subsequent conditions in the ACL statement.
5278 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5280 15. Added a new option "connect=<time>" to callout options, to set a different
5283 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5284 was its contents. (It was OK if the option was not defined at all.)
5286 17. A "Completed" log line is now written for messages that are removed from
5287 the spool by the -Mrm option.
5289 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5290 information about exactly what failed.
5292 19. Added -dd to debug only the daemon process.
5294 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5295 handles timeouts, both on the server side and network timeouts. Renamed the
5296 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5298 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5299 It is now set to "smtps".
5301 22. $host_address is now set to the target address during the checking of
5302 ignore_target_hosts.
5304 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5305 being passed; this would have caused $sender_host_name to have been used if
5306 matching the list had actually called for a host name (not very likely,
5307 since this list is usually IP addresses). A host name is now passed as
5310 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5311 code that specifies a non-restarting handler (typically sigaction() in
5312 modern systems) in an attempt to fix a rare and obscure crash bug.
5314 25. Narrowed the window for a race in the daemon that could cause it to ignore
5315 SIGCHLD signals. This is not a major problem, because they are used only to
5316 wake it up if nothing else does.
5318 26. A malformed maildirsize file could cause Exim to calculate negative values
5319 for the mailbox size or file count. Odd effects could occur as a result.
5320 The maildirsize information is now recalculated if the size or filecount
5323 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5324 support for a long time. Removed HAVE_SYS_VFS_H.
5326 28. Installed the latest version of exipick from John Jetmore.
5328 29. In an address list, if the pattern was not a regular expression, an empty
5329 subject address (from a bounce message) matched only if the pattern was an
5330 empty string. Non-empty patterns were not even tested. This was the wrong
5331 because it is perfectly reasonable to use an empty address as part of a
5332 database query. An empty address is now tested by patterns that are
5333 lookups. However, all the other forms of pattern expect the subject to
5334 contain a local part and a domain, and therefore, for them, an empty
5335 address still always fails if the pattern is not itself empty.
5337 30. Exim went into a mad DNS loop when attempting to do a callout where the
5338 host was specified on an smtp transport, and looking it up yielded more
5339 than one IP address.
5341 31. Re-factored the code for checking spool and log partition space into a
5342 function that finds that data and another that does the check. The former
5343 is then used to implement four new variables: $spool_space, $log_space,
5344 $spool_inodes, and $log_inodes.
5346 32. The RFC2047 encoding function was originally intended for short strings
5347 such as real names; it was not keeping to the 75-character limit for
5348 encoded words that the RFC imposes. It now respects the limit, and
5349 generates multiple encoded words if necessary. To be on the safe side, I
5350 have increased the buffer size for the ${rfc2047: expansion operator from
5353 33. It is now permitted to omit both strings after an "if" condition; if the
5354 condition is true, the result is "true". As before, when the second string
5355 is omitted, a false condition yields an empty string. This makes it less
5356 cumbersome to write custom ACL and router conditions.
5358 34. Failure to deliver a bounce message always caused it to be frozen, even if
5359 there was an errors_to setting on the router. The errors_to setting is now
5362 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5363 canonical form (fully expanded) before being placed in
5364 $sender_host_address.
5366 36. The table in the code that translates DNS record types into text (T_A to
5367 "A" for instance) was missing entries for NS and CNAME. It is just possible
5368 that this could have caused confusion if both these types were looked up
5369 for the same domain, because the text type is used as part of Exim's
5370 per-process caching. But the chance of anyone hitting this buglet seems
5373 37. The dnsdb lookup has been extended in a number of ways.
5375 (1) There is a new type, "zns", which walks up the domain tree until it
5376 finds some nameserver records. It should be used with care.
5378 (2) There is a new type, "mxh", which is like "mx" except that it returns
5379 just the host names, not the priorities.
5381 (3) It is now possible to give a list of domains (or IP addresses) to be
5382 looked up. The behaviour when one of the lookups defers can be
5383 controlled by a keyword.
5385 (4) It is now possible to specify the separator character for use when
5386 multiple records are returned.
5388 38. The dnslists ACL condition has been extended: it is now possible to supply
5389 a list of IP addresses and/or domains to be looked up in a particular DNS
5392 39. Added log_selector=+queue_time_overall.
5394 40. When running the queue in the test harness, wait just a tad after forking a
5395 delivery process, to get repeatability of debugging output.
5397 41. Include certificate and key file names in error message when GnuTLS fails
5398 to set them up, because the GnuTLS error message doesn't include the name
5399 of the failing file when there is a problem reading it.
5401 42. Allow both -bf and -bF in the same test run.
5403 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5405 44. The "Exiscan patch" is now merged into the mainline Exim source.
5407 45. Sometimes the final signoff response after QUIT could fail to get
5408 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5409 < 0 before doing a fflush(). This bug looks as though it goes back to the
5410 introduction of TLS in release 3.20, but "sometimes" must have been rare
5411 because the tests only now provoked it.
5413 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5414 (this can affect the format of dates).
5416 47. exim_tidydb, when checking for the continued existence of a message for
5417 which it has found a message-specific retry record, was not finding
5418 messages that were in split spool directories. Consequently, it was
5419 deleting retry records that should have stayed in existence.
5421 48. Steve fixed some bugs in eximstats.
5423 49. The SPA authentication driver was not abandoning authentication and moving
5424 on to the next authenticator when an expansion was forced to fail,
5425 contradicting the general specification for all authenticators. Instead it
5426 was generating a temporary error. It now behaves as specified.
5428 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5429 (the order specifies the preference for clients). The order is now AES256,
5430 AES128, 3DES, ARCFOUR128.
5432 51. Small patch to Sieve code - explicitly set From: when generating an
5435 52. Exim crashed if a remote delivery caused a very long error message to be
5436 recorded - for instance if somebody sent an entire SpamAssassin report back
5437 as a large number of 550 error lines. This bug was coincidentally fixed by
5438 increasing the size of one of Exim's internal buffers (big_buffer) that
5439 happened as part of the Exiscan merge. However, to be on the safe side, I
5440 have made the code more robust (and fixed the comments that describe what
5443 53. Now that there can be additional text after "Completed" in log lines (if
5444 the queue_time_overall log selector is set), a one-byte patch to exigrep
5445 was needed to allow it to recognize "Completed" as not the last thing in
5448 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5449 patch that reportedly fixes this has been added. I am not expert enough to
5450 create a test for it. This is what the patch creator wrote:
5452 "I found a little strange behaviour of ldap code when working with
5453 Windows 2003 AD Domain, where users was placed in more than one
5454 Organization Units. When I tried to give exim partial DN, the exit code
5455 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5456 But simultaneously result of request was absolutely normal ldap result,
5457 so I produce this patch..."
5459 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5460 REFERENCE, so I have modified the code to exclude the patch when that macro
5463 55. Some experimental protocols are using DNS PTR records for new purposes. The
5464 keys for these records are domain names, not reversed IP addresses. The
5465 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5466 leaves it alone. Component reversal etc. now happens only for IP addresses.
5469 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5471 57. Double the size of the debug message buffer (to 2048) so that more of very
5472 long debug lines gets shown.
5474 58. The exicyclog utility now does better if the number of log files to keep
5475 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5477 59. Two changes related to the smtp_active_hostname option:
5479 (1) $smtp_active_hostname is now available as a variable.
5480 (2) The default for smtp_banner uses $smtp_active_hostname instead
5481 of $primary_hostname.
5483 60. The host_aton() function is supposed to be passed a string that is known
5484 to be a valid IP address. However, in the case of IPv6 addresses, it was
5485 not checking this. This is a hostage to fortune. Exim now panics and dies
5486 if the condition is not met. A case was found where this could be provoked
5487 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5488 components; fortuitously, this particular loophole had already been fixed
5489 by change 4.50/55 above.
5491 If there are any other similar loopholes, the new check in host_aton()
5492 itself should stop them being exploited. The report I received stated that
5493 data on the command line could provoke the exploit when Exim was running as
5494 exim, but did not say which command line option was involved. All I could
5495 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5496 running as the user.
5499 61. There was a buffer overflow vulnerability in the SPA authentication code
5500 (which came originally from the Samba project). I have added a test to the
5501 spa_base64_to_bits() function which I hope fixes it.
5504 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5505 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5507 63. The daemon start-up calls getloadavg() while still root for those OS that
5508 need the first call to be done as root, but it missed one case: when
5509 deliver_queue_load_max is set with deliver_drop_privilege. This is
5510 necessary for the benefit of the queue runner, because there is no re-exec
5511 when deliver_drop_privilege is set.
5513 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5514 This has been fixed.
5516 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5517 was in use, was not putting the data itself into the right store pool;
5518 consequently, it could be overwritten for a subsequent message in the same
5519 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5522 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5524 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5525 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5526 the function were treating the return as a boolean value, which happened to
5527 work because 0=false and not-0=true, but is not correct code.
5529 68. The host_aton() function was not handling scoped IPv6 addresses (those
5530 with, for example, "%eth0" on the end) correctly.
5532 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5533 negated items (that is, ~something) in unsigned ints. Some compilers
5534 apparently mutter when there is no cast.
5536 70. If an address verification called from an ACL failed, and did not produce a
5537 user-specific message (i.e. there was only a "system" message), nothing was
5538 put in $acl_verify_message. In this situation, it now puts the system
5541 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5542 session; change 4.31/43 added the unwanted input to the log line - except
5543 that it did not do this in the start of session case. It now does.
5545 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5546 This is wrong and can cause the other end to generate a synchronization
5547 error if it is another Exim or anything else that does the synchronization
5548 check. A QUIT command is no longer sent after a timeout.
5550 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5551 during host lookups.
5553 74. The defer_ok option of callout verification was not working if it was used
5554 when verifying addresses in header lines, that is, for this case:
5556 verify = header_sender/callout=defer_ok
5558 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5559 those file descriptors could be used for SMTP connections. If anything
5560 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5561 could be sent to the SMTP client, causing chaos. The daemon now opens
5562 stdin, stdout, and stderr to /dev/null when it puts itself into the
5565 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5566 log by default. The user can override this with suitable Perl magic.
5568 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5569 the log message when discard triggers, was not working for the DATA ACL or
5570 for the non-SMTP ACL.
5572 78. Error message wording change in sieve.c.
5574 79. If smtp_accept_max_per_host was set, the number of connections could be
5575 restricted to fewer than expected, because the daemon was trying to set up
5576 a new connection before checking whether the processes handling previous
5577 connections had finished. The check for completed processes is now done
5578 earlier. On busy systems, this bug wouldn't be noticed because something
5579 else would have woken the daemon, and it would have reaped the completed
5582 80. If a message was submitted locally by a user whose login name contained one
5583 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5584 It caused a spool format error. I have fixed the spool reading code. A
5585 related problem was that the "from" clause in the Received: line became
5586 illegal because of the space(s). It is now covered by ${quote_local_part.
5588 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5591 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5593 83. Patch from Sieve maintainer to fix unterminated string problem in
5594 "vacation" handling.
5596 84. Some minor changes to the Linux configuration files to help with other
5597 OS variants using glibc.
5599 85. One more patch for Sieve to update vacation handling to latest spec.
5602 ----------------------------------------------------
5603 See the note above about the 4.44 and 4.50 releases.
5604 ----------------------------------------------------
5610 1. Change 4.43/35 introduced a bug that caused file counts to be
5611 incorrectly computed when quota_filecount was set in an appendfile
5614 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5615 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5618 3. Give more explanation in the error message when the command for a transport
5619 filter fails to execute.
5621 4. There are several places where Exim runs a non-Exim command in a
5622 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5623 was being done only for the command run by the queryprogram router. It is
5624 now done for all such subprocesses. The other cases are: ${run, transport
5625 filters, and the commands run by the lmtp and pipe transports.
5627 5. Some older OS have a limit of 256 on the maximum number of file
5628 descriptors. Exim was using setrlimit() to set 1000 as a large value
5629 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5630 systems. I've change it so that if it can't get 1000, it tries for 256.
5632 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5633 was an oversight, and furthermore, ever since the addition of extra
5634 controls (e.g. 4.43/32), the checks on when to allow different forms of
5635 "control" were broken. There should now be diagnostics for all cases when a
5636 control that does not make sense is encountered.
5638 7. $recipients is now available in the predata ACL (oversight).
5640 8. Tidy the search cache before the fork to do a delivery from a message
5641 received from the command line. Otherwise the child will trigger a lookup
5642 failure and thereby defer the delivery if it tries to use (for example) a
5643 cached ldap connection that the parent has called unbind on.
5645 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5646 of $address_data from the recipient verification was clobbered by the
5647 sender verification.
5649 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5650 was its contents. (It was OK if the option was not defined at all.)
5652 11. A "Completed" log line is now written for messages that are removed from
5653 the spool by the -Mrm option.
5655 12. $host_address is now set to the target address during the checking of
5656 ignore_target_hosts.
5658 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5659 being passed; this would have caused $sender_host_name to have been used if
5660 matching the list had actually called for a host name (not very likely,
5661 since this list is usually IP addresses). A host name is now passed as
5664 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5665 code that specifies a non-restarting handler (typically sigaction() in
5666 modern systems) in an attempt to fix a rare and obscure crash bug.
5668 15. Narrowed the window for a race in the daemon that could cause it to ignore
5669 SIGCHLD signals. This is not a major problem, because they are used only to
5670 wake it up if nothing else does.
5672 16. A malformed maildirsize file could cause Exim to calculate negative values
5673 for the mailbox size or file count. Odd effects could occur as a result.
5674 The maildirsize information is now recalculated if the size or filecount
5677 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5678 support for a long time. Removed HAVE_SYS_VFS_H.
5680 18. Updated exipick to current release from John Jetmore.
5682 19. Allow an empty sender to be matched against a lookup in an address list.
5683 Previously the only cases considered were a regular expression, or an
5686 20. Exim went into a mad DNS lookup loop when doing a callout where the
5687 host was specified on the transport, if the DNS lookup yielded more than
5690 21. The RFC2047 encoding function was originally intended for short strings
5691 such as real names; it was not keeping to the 75-character limit for
5692 encoded words that the RFC imposes. It now respects the limit, and
5693 generates multiple encoded words if necessary. To be on the safe side, I
5694 have increased the buffer size for the ${rfc2047: expansion operator from
5697 22. Failure to deliver a bounce message always caused it to be frozen, even if
5698 there was an errors_to setting on the router. The errors_to setting is now
5701 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5702 canonical form (fully expanded) before being placed in
5703 $sender_host_address.
5705 24. Updated eximstats to version 1.33
5707 25. Include certificate and key file names in error message when GnuTLS fails
5708 to set them up, because the GnuTLS error message doesn't include the name
5709 of the failing file when there is a problem reading it.
5711 26. Expand error message when OpenSSL has problems setting up cert/key files.
5714 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5715 (this can affect the format of dates).
5717 28. exim_tidydb, when checking for the continued existence of a message for
5718 which it has found a message-specific retry record, was not finding
5719 messages that were in split spool directories. Consequently, it was
5720 deleting retry records that should have stayed in existence.
5722 29. eximstats updated to version 1.35
5723 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5724 1.35 - bugfix such that pie charts by volume are generated correctly
5726 30. The SPA authentication driver was not abandoning authentication and moving
5727 on to the next authenticator when an expansion was forced to fail,
5728 contradicting the general specification for all authenticators. Instead it
5729 was generating a temporary error. It now behaves as specified.
5731 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5732 (the order specifies the preference for clients). The order is now AES256,
5733 AES128, 3DES, ARCFOUR128.
5735 31. Small patch to Sieve code - explicitly set From: when generating an
5738 32. Exim crashed if a remote delivery caused a very long error message to be
5739 recorded - for instance if somebody sent an entire SpamAssassin report back
5740 as a large number of 550 error lines. This bug was coincidentally fixed by
5741 increasing the size of one of Exim's internal buffers (big_buffer) that
5742 happened as part of the Exiscan merge. However, to be on the safe side, I
5743 have made the code more robust (and fixed the comments that describe what
5746 33. Some experimental protocols are using DNS PTR records for new purposes. The
5747 keys for these records are domain names, not reversed IP addresses. The
5748 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5749 leaves it alone. Component reversal etc. now happens only for IP addresses.
5752 34. The host_aton() function is supposed to be passed a string that is known
5753 to be a valid IP address. However, in the case of IPv6 addresses, it was
5754 not checking this. This is a hostage to fortune. Exim now panics and dies
5755 if the condition is not met. A case was found where this could be provoked
5756 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5757 components; fortuitously, this particular loophole had already been fixed
5758 by change 4.50/55 or 4.44/33 above.
5760 If there are any other similar loopholes, the new check in host_aton()
5761 itself should stop them being exploited. The report I received stated that
5762 data on the command line could provoke the exploit when Exim was running as
5763 exim, but did not say which command line option was involved. All I could
5764 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5765 running as the user.
5768 35. There was a buffer overflow vulnerability in the SPA authentication code
5769 (which came originally from the Samba project). I have added a test to the
5770 spa_base64_to_bits() function which I hope fixes it.
5773 36. The daemon start-up calls getloadavg() while still root for those OS that
5774 need the first call to be done as root, but it missed one case: when
5775 deliver_queue_load_max is set with deliver_drop_privilege. This is
5776 necessary for the benefit of the queue runner, because there is no re-exec
5777 when deliver_drop_privilege is set.
5779 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5780 was in use, was not putting the data itself into the right store pool;
5781 consequently, it could be overwritten for a subsequent message in the same
5782 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5785 38. Sometimes the final signoff response after QUIT could fail to get
5786 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5787 < 0 before doing a fflush(). This bug looks as though it goes back to the
5788 introduction of TLS in release 3.20, but "sometimes" must have been rare
5789 because the tests only now provoked it.
5795 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5796 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5797 Now it returns an integer. A number of calls were still expecting a T/F
5798 return. Fortuitously, in all cases, the tests worked in OK situations,
5799 which is the norm. However, things would have gone wrong on any write
5800 failures on the smtp file descriptor. This function is used when sending
5801 messages over SMTP and also when doing verify callouts.
5803 2. When Exim is called to do synchronous delivery of a locally submitted
5804 message (the -odf or -odi options), it no longer closes stderr before doing
5807 3. Implemented the mua_wrapper option.
5809 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5811 5. Implemented the functions header_remove(), header_testname(),
5812 header_add_at_position(), and receive_remove_recipient(), and exported them
5815 6. If an ACL "warn" statement specified the addition of headers, Exim already
5816 inserted X-ACL-Warn: at the start if there was no header name. However, it
5817 was not making this test for the second and subsequent header lines if
5818 there were newlines in the string. This meant that an invalid header could
5819 be inserted if Exim was badly configured.
5821 7. Allow an ACL "warn" statement to add header lines at the start or after all
5822 the Received: headers, as well as at the end.
5824 8. Added the rcpt_4xx retry error code.
5826 9. Added postmaster_mailfrom=xxx to callout verification option.
5828 10. Added mailfrom=xxxx to the callout verification option, for verify=
5831 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5832 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5834 12. Inserted some casts to stop certain compilers warning when using pointer
5835 differences as field lengths or precisions in printf-type calls (mostly
5836 affecting debugging statements).
5838 13. Added optional readline() support for -be (dynamically loaded).
5840 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5841 same clock tick as a message's arrival, so that its received time was the
5842 same as the "first fail" time on the retry record, and that message
5843 remained on the queue past the ultimate address timeout, every queue runner
5844 would try a delivery (because it was past the ultimate address timeout) but
5845 after another failure, the ultimate address timeout, which should have then
5846 bounced the address, did not kick in. This was a "< instead of <=" error;
5847 in most cases the first failure would have been in the next clock tick
5848 after the received time, and all would be well.
5850 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5851 being recognized when the domain list was tested by the match_domain
5852 condition in an expansion string.
5854 16. Added the ${str2b64: operator.
5856 17. Exim was always calling setrlimit() to set a large limit for the number of
5857 processes, without checking whether the existing limit was already
5858 adequate. (It did check for the limit on file descriptors.) Furthermore,
5859 errors from getrlimit() and setrlimit() were being ignored. Now they are
5860 logged to the main and panic logs, but Exim does carry on, to try to do its
5861 job under whatever limits there are.
5863 18. Imported PCRE 5.0.
5865 19. Trivial typo in log message " temporarily refused connection" (the leading
5868 20. If the log selector return_path_on_delivery was set and an address was
5869 redirected to /dev/null, the delivery process crashed because it assumed
5870 that a return path would always be set for a "successful" delivery. In this
5871 case, the whole delivery is bypassed as an optimization, and therefore no
5874 21. Internal re-arrangement: the function for sending a challenge and reading
5875 a response while authentication was assuming a zero-terminated challenge
5876 string. It's now changed to take a pointer and a length, to allow for
5877 binary data in such strings.
5879 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5881 23. Exim was not respecting finduser_retries when seeking the login of the
5882 uid under which it was called; it was always trying 10 times. (The default
5883 setting of finduser_retries is zero.) Also, it was sleeping after the final
5884 failure, which is pointless.
5886 24. Implemented tls_on_connect_ports.
5888 25. Implemented acl_smtp_predata.
5890 26. If the domain in control=submission is set empty, Exim assumes that the
5891 authenticated id is a complete email address when it generates From: or
5892 Sender: header lines.
5894 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5895 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5896 chown and chgrp in /bin and hostname in /usr/bin.
5898 28. Exim was keeping the "process log" file open after each use, just as it
5899 does for the main log. This opens the possibility of it remaining open for
5900 long periods when the USR1 signal hits a daemon. Occasional processlog
5901 errors were reported, that could have been caused by this. Anyway, it seems
5902 much more sensible not to leave this file open at all, so that is what now
5905 29. The long-running daemon process does not normally write to the log once it
5906 has entered its main loop, and it closes the log before doing so. This is
5907 so that log files can straightforwardly be renamed and moved. However,
5908 there are a couple of unusual error situations where the daemon does write
5909 log entries, and I had neglected to close the log afterwards.
5911 30. The text of an SMTP error response that was received during a remote
5912 delivery was being truncated at 512 bytes. This is too short for some of
5913 the long messages that one sometimes sees. I've increased the limit to
5916 31. It is now possible to make retry rules that apply only when a message has a
5917 specific sender, in particular, an empty sender.
5919 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5920 it possible to be selective about when SMTP synchronization is enforced.
5922 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5924 32. Implemented hosts_connection_nolog.
5926 33. Added an ACL for QUIT.
5928 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5931 35. Added mailbox_size and mailbox_filecount to appendfile.
5933 36. Added control = no_multiline_responses to ACLs.
5935 37. There was a bug in the logic of the code that waits for the clock to tick
5936 in the case where the clock went backwards by a substantial amount such
5937 that the microsecond fraction of "now" was more than the microsecond
5938 fraction of "then" (but the whole seconds number was less).
5940 38. Added support for the libradius Radius client library this is found on
5941 FreeBSD (previously only the radiusclient library was supported).
5947 1. When certain lookups returned multiple values in the form name=value, the
5948 quoting of the values was not always being done properly. Specifically:
5949 (a) If the value started with a double quote, but contained no whitespace,
5951 (b) If the value contained whitespace other than a space character (i.e.
5952 tabs or newlines or carriage returns) it was not quoted.
5953 This fix has been applied to the mysql and pgsql lookups by writing a
5954 separate quoting function and calling it from the lookup code. The fix
5955 should probably also be applied to nisplus, ibase and oracle lookups, but
5956 since I cannot test any of those, I have not disturbed their existing code.
5958 2. A hit in the callout cache for a specific address caused a log line with no
5959 reason for rejecting RCPT. Now it says "Previous (cached) callout
5960 verification failure".
5962 3. There was an off-by-one bug in the queryprogram router. An over-long
5963 return line was truncated at 256 instead of 255 characters, thereby
5964 overflowing its buffer with the terminating zero. As well as fixing this, I
5965 have increased the buffer size to 1024 (and made a note to document this).
5967 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5968 when Exim is waiting for an SMTP response from a remote server, Exim
5969 restarts its select() call on the socket, thereby resetting its timeout.
5970 This is not a problem when such interrupts are rare. Somebody set up a cron
5971 job to run exiwhat every 2 minutes, which is less than the normal select()
5972 timeout (5 or 10 minutes). This meant that the select() timeout never
5973 kicked in because it was always reset. I have fixed this by comparing the
5974 time when an interrupt arrives with the time at the start of the first call
5975 to select(). If more time than the timeout has elapsed, the interrupt is
5976 treated as a timeout.
5978 5. Some internal re-factoring in preparation for the addition of Sieve
5979 extensions (by MH). In particular, the "personal" test is moved to a
5980 separate function, and given an option for scanning Cc: and Bcc: (which is
5981 not set for Exim filters).
5983 6. When Exim created an email address using the login of the caller as the
5984 local part (e.g. when creating a From: or Sender: header line), it was not
5985 quoting the local part when it contained special characters such as @.
5987 7. Installed new OpenBSD configuration files.
5989 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5990 try to make them clearer.
5992 9. Callout options, other than the timeout value, were being ignored when
5993 verifying sender addresses in header lines. For example, when using
5995 verify = header_sender/callout=no_cache
5997 the cache was (incorrectly) being used.
5999 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6000 only the Cygwin environment.
6002 11. When return_path_on_delivery was set as a log selector, if different remote
6003 addresses in the same message used different return paths and parallel
6004 remote delivery occurred, the wrong values would sometimes be logged.
6005 (Whenever a remote delivery process finished, the return path value from
6006 the most recently started remote delivery process was logged.)
6008 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6009 lines when AUTH and/or TLS are in use. This is the "received protocol"
6010 field. Exim used to use "asmtp" for authenticated SMTP, without any
6011 indication (in the protocol name) for TLS use. Now it follows the RFC and
6012 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6013 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6014 names appear in log lines as well as in Received: header lines.
6016 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6017 extensions, and comparison tests, and to fix some bugs.
6019 14. Changes to the "personal" filter test:
6021 (1) The test was buggy in that it was just doing the equivalent of
6022 "contains" tests on header lines. For example, if a user's address was
6023 anne@some.where, the "personal" test would incorrectly be true for
6025 To: susanne@some.where
6027 This test is now done by extracting each address from the header in turn,
6028 and checking the entire address. Other tests that are part of "personal"
6029 are now done using regular expressions (for example, to check local parts
6030 of addresses in From: header lines).
6032 (2) The list of non-personal local parts in From: addresses has been
6033 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6034 taken from the Sieve specification recommendations.
6036 (3) If the message contains any header line starting with "List-" it is
6037 treated as non-personal.
6039 (4) The test for "circular" in the Subject: header line has been removed
6040 because it now seems ill-conceived.
6042 15. Minor typos in src/EDITME comments corrected.
6044 16. Installed latest exipick from John Jetmore.
6046 17. If headers_add on a router specified a text string that was too long for
6047 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6048 of string_sprintf() is now avoided.
6050 18. $message_body_size was not set (it was always zero) when running the DATA
6051 ACL and the local_scan() function.
6053 19. For the "mail" command in an Exim filter, no default was being set for
6054 the once_repeat time, causing a random time value to be used if "once" was
6055 specified. (If the value happened to be <= 0, no repeat happened.) The
6056 default is now 0s, meaning "never repeat". The "vacation" command was OK
6057 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6058 (I found it when inspecting the code).
6060 20. There is now an overall timeout for performing a callout verification. It
6061 defaults to 4 times the callout timeout, which applies to individual SMTP
6062 commands during the callout. The overall timeout applies when there is more
6063 than one host that can be tried. The timeout is checked before trying the
6064 next host. This prevents very long delays if there are a large number of
6065 hosts and all are timing out (e.g. when the network connections are timing
6066 out). The value of the overall timeout can be changed by specifying an
6067 additional sub-option for "callout", called "maxwait". For example:
6069 verify = sender/callout=5s,maxwait=20s
6071 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6072 to the end before writing, but this should make it even safer).
6074 22. Exim was forgetting that it had advertised PIPELINING for the second and
6075 subsequent messages on an SMTP connection. It was also not resetting its
6076 memory on STARTTLS and an internal HELO.
6078 23. When Exim logs an SMTP synchronization error within a session, it now
6079 records whether PIPELINING has been advertised or not.
6081 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6082 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6083 rather than long int.
6085 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6087 26. Added the never_mail option to autoreply.
6093 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6094 crash if the getsockname() call failed; this can happen if a connection is
6095 closed very soon after it is established. The problem was simply in the
6096 order in which certain operations were done, causing Exim to try to write
6097 to the SMTP stream before it had set up the file descriptor. The bug has
6098 been fixed by making things happen in the correct order.
6104 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6105 before the connection was closed, thus losing the rejection response.
6107 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6108 some early Solaris releases, but causes trouble in current releases where
6109 socklen_t is defined.
6111 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6114 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6117 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6118 that releasing the top of it at the end releases what was used for sub-
6119 expansions (unless the block got too big). However, discard this block if
6120 the first thing is a variable or header, so that we can use its block when
6121 it is dynamic (useful for very large $message_headers, for example).
6123 6. Lookups now cache *every* query, not just the most recent. A new, separate
6124 store pool is used for this. It can be recovered when all lookup caches are
6125 flushed. Lookups now release memory at the end of their result strings.
6126 This has involved some general refactoring of the lookup sources.
6128 7. Some code has been added to the store_xxx() functions to reduce the amount
6129 of flapping under certain conditions.
6131 8. log_incoming_interface used to affect only the <= reception log lines. Now
6132 it causes the local interface and port to be added to several more SMTP log
6133 lines, for example "SMTP connection from", and rejection lines.
6135 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6137 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6139 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6140 could be overwritten at the end of the current message (or the start of a
6141 new message if it was set in a HELO ACL). The value is now preserved for
6142 the duration of the SMTP connection.
6144 12. If a transport had a headers_rewrite setting, and a matching header line
6145 contained an unqualified address, that address was qualified, even if it
6146 did not match any rewriting rules. The underlying bug was that the values
6147 of the flags that permit the existence of unqualified sender and recipient
6148 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6149 non-local messages, and by -bnq for local messages) were not being
6150 preserved with the message after it was received.
6152 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6153 "next input=" as part of the text comprising the host identity instead of
6154 the correct text. The code was using the same buffer for two different
6155 strings. However, depending on which order the printing function evaluated
6156 its arguments, the bug did not always show up. Under Linux, for example, my
6157 test suite worked just fine.
6159 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6160 doesn't work with some older versions of Perl. It has been changed to "my",
6161 which in any case is probably the better facility to use.
6163 15. A really picky compiler found some instances of statements for creating
6164 error messages that either had too many or two few arguments for the format
6167 16. The size of the buffer for calls to the DNS resolver has been increased
6168 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6169 for addresses that have a lot of PTR records. This alleviates a problem; it
6170 does not fully solve it.
6172 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6173 buffer now truncates the list and logs the incident, which is the same
6174 action as happens when Exim is looking up a host name and its aliases.
6175 Previously in this situation something unpredictable would happen;
6176 sometimes it was "internal error: store_reset failed".
6178 18. If a server dropped the connection unexpectedly when an Exim client was
6179 using GnuTLS and trying to read a response, the client delivery process
6180 crashed while trying to generate an error log message.
6182 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6183 string, for example:
6185 warn message = H1: something\nH2: something
6187 the text was added as a single header line from Exim's point of view
6188 though it ended up OK in the delivered message. However, searching for the
6189 second and subsequent header lines using $h_h2: did not work. This has been
6190 fixed. Similarly, if a system filter added multiple headers in this way,
6191 the routers could not see them.
6193 20. Expanded the error message when iplsearch is called with an invalid key to
6194 suggest using net-iplsearch in a host list.
6196 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6197 ACLs are no longer actually imposed (and a message to that effect is
6200 22. If a "gecos" field in a passwd entry contained escaped characters, in
6201 particular, if it contained a \" sequence, Exim got it wrong when building
6202 a From: or a Sender: header from that name. A second bug also caused
6203 incorrect handling when an unquoted " was present following a character
6204 that needed quoting.
6206 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6207 was not being matched caselessly.
6209 24. Arranged for all hyphens in the exim.8 source to be escaped with
6212 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6213 properly. Recipient callout cache records were still being keyed to include
6214 the sender, even when use_sender was set false. This led to far more
6215 callouts that were necessary. The sender is no longer included in the key
6216 when use_sender is false.
6218 26. Added "control = submission" modifier to ACLs.
6220 27. Added the ${base62d: operator to decode base 62 numbers.
6222 28. dnsdb lookups can now access SRV records.
6224 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6225 the configuration file.
6227 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6228 (-v) mode. This makes the output for a verbose queue run more intelligible.
6230 31. Added a use_postmaster feature to recipient callouts.
6232 32. Added the $body_zerocount variable, containing the number of binary zero
6233 bytes in the message body.
6235 33. The time of last modification of the "new" subdirectory is now used as the
6236 "mailbox time last read" when there is a quota error for a maildir
6239 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6241 35. Added +ignore_unknown as a special item in host lists.
6243 36. Code for decoding IPv6 addresses in host lists is now included, even if
6244 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6245 address was recognized as an IP address, but was then not correctly decoded
6246 into binary, causing unexpected and incorrect effects when compared with
6253 1. Very minor rewording of debugging text in manualroute to say "list of
6254 hosts" instead of "hostlist".
6256 2. If verify=header_syntax was set, and a header line with an unqualified
6257 address (no domain) and a large number of spaces between the end of the
6258 name and the colon was received, the reception process suffered a buffer
6259 overflow, and (when I tested it) crashed. This was caused by some obsolete
6260 code that should have been removed. The fix is to remove it!
6262 3. When running in the test harness, delay a bit after writing a bounce
6263 message to get a bit more predictability in the log output.
6265 4. Added a call to search_tidyup() just before forking a reception process. In
6266 theory, someone could use a lookup in the expansion of smtp_accept_max_
6267 per_host which, without the tidyup, could leave open a database connection.
6269 5. Added the variables $recipient_data and $sender_data which get set from a
6270 lookup success in an ACL "recipients" or "senders" condition, or a router
6271 "senders" option, similar to $domain_data and $local_part_data.
6273 6. Moved the writing of debug_print from before to after the "senders" test
6276 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6277 problems for message scanning, either using a data ACL, or using
6278 local_scan() because the Received: header was not generated till after they
6279 were called (in order to set the time as the time of reception completion).
6280 I have revised the way this works. The header is now generated after the
6281 body is received, but before the ACL or local_scan() are called. After they
6282 are run, the timestamp in the header is updated.
6288 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6289 before starting a queue runner without re-exec. This happened only when
6290 deliver_drop_privilege was set or when the Exim user was set to root. The
6291 effect of the bug was that timeouts during subsequent deliveries caused
6292 crashes instead of being properly handled. The handler is now left at its
6293 default (and expected) setting.
6295 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6296 message, again when deliver_drop_privilege is set or Exim is run as root.
6297 The bug described in (1) was not present in this case, but the tidying up
6298 of the other signals was missing. I have made the two cases consistent.
6300 3. The ignore_target_hosts setting on a manualroute router was being ignored
6301 for hosts that were looked up using the /MX notation.
6303 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6306 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6307 operated on the sender address. After changing the $sender_address to <>
6308 for the sender address verify, Exim was re-instated it as the original
6309 (before rewriting) address, but remembering that it had rewritten it, so it
6310 wasn't rewriting it again. This bug also had the effect of breaking the
6311 sender address verification caching when the sender address was rewritten.
6313 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6314 This has been changed so that if the ip literal address matches
6315 ignore_target_hosts, the router declines.
6317 7. Added expansion conditions match_domain, match_address, and match_local_
6318 part (NOT match_host).
6320 8. The placeholder for the Received: header didn't have a length field set.
6322 9. Added code to Exim itself and to exim_lock to test for a specific race
6323 condition that could lead to file corruption when using MBX delivery. The
6324 issue is with the lockfile that is created in /tmp. If this file is removed
6325 after a process has opened it but before that process has acquired a lock,
6326 there is the potential for a second process to recreate the file and also
6327 acquire a lock. This could lead to two Exim processes writing to the file
6328 at the same time. The added code performs the same test as UW imapd; it
6329 checks after acquiring the lock that its file descriptor still refers to
6330 the same named file.
6332 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6333 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6336 11. Added the smtp_active_hostname option. If used, this will typically be made
6337 to depend on the incoming interface address. Because $interface_address is
6338 not set up until the daemon has forked a reception process, error responses
6339 that can happen earlier (such as "too many connections") no longer contain
6342 12. If an expansion in a condition on a "warn" statement fails because a lookup
6343 defers, the "warn" statement is abandoned, and the next ACL statement is
6344 processed. Previously this caused the whole ACL to be aborted.
6346 13. Added the iplsearch lookup type.
6348 14. Added ident_timeout as a log selector.
6350 15. Added tls_certificate_verified as a log selector.
6352 16. Added a global option tls_require_ciphers (compare the smtp transport
6353 option of the same name). This controls incoming TLS connections.
6355 17. I finally figured out how to make tls_require_ciphers do a similar thing
6356 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6357 before starting the TLS session.
6359 18. Tabs are now shown as \t in -bP output.
6361 19. If the log selector return_path_on_delivery was set, Exim crashed when
6362 bouncing a message because it had too many Received: header lines.
6364 20. If two routers both had headers_remove settings, and the first one included
6365 a superfluous trailing colon, the final name in the first list and the
6366 first name in the second list were incorrectly joined into one item (with a
6367 colon in the middle).
6373 1. Added -C and -D options to the exinext utility, mainly to make it easier
6374 to include in the automated testing, but these could be helpful when
6375 multiple configurations are in use.
6377 2. The exinext utility was not formatting the output nicely when there was
6378 an alternate port involved in the retry record key, nor when there was a
6379 message id as well (for retries that were specific to a specific message
6380 and a specific host). It was also confused by IPv6 addresses, because of
6381 the additional colons they contain. I have fixed the IPv4 problem, and
6382 patched it up to do a reasonable job for IPv6.
6384 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6385 delivery, the log line now contains "pipelined" if PIPELINING was used.
6387 4. An SMTP transport process used to panic and die if the bind() call to set
6388 an explicit outgoing interface failed. This has been changed; it is now
6389 treated in the same way as a connect() failure.
6391 5. A reference to $sender_host_name in the part of a conditional expansion
6392 that was being skipped was still causing a DNS lookup. This no longer
6395 6. The def: expansion condition was not recognizing references to header lines
6396 that used bh_ and bheader_.
6398 7. Added the _cache feature to named lists.
6400 8. The code for checking quota_filecount in the appendfile transport was
6401 allowing one more file than it should have been.
6403 9. For compatibility with Sendmail, the command line option
6411 and sets the incoming protocol and host name (for trusted callers). The
6412 host name and its colon can be omitted when only the protocol is to be set.
6413 Note the Exim already has two private options, -pd and -ps, that refer to
6414 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6415 "s", but I don't think that's a major issue.
6417 10. A number of refactoring changes to the code, none of which should affect
6420 (a) The number of logging options was getting close to filling up the
6421 32-bit word that was used as a bit map. I have split them into two classes:
6422 those that are passed in the argument to log_write(), and those that are
6423 only ever tested independently outside of that function. These are now in
6424 separate 32-bit words, so there is plenty of room for expansion again.
6425 There is no change in the user interface or the logging behaviour.
6427 (b) When building, for example, log lines, the code previously used a
6428 macro that called string_cat() twice, in order to add two strings. This is
6429 not really sufficiently general. Furthermore, there was one instance where
6430 it was actually wrong because one of the argument was used twice, and in
6431 one call a function was used. (As it happened, calling the function twice
6432 did not affect the overall behaviour.) The macro has been replaced by a
6433 function that can join an arbitrary number of extra strings onto a growing
6436 (c) The code for expansion conditions now uses a table and a binary chop
6437 instead of a serial search (which was left over from when there were very
6438 few conditions). Also, it now recognizes conditions like "pam" even when
6439 the relevant support is not compiled in: a suitably worded error message is
6440 given if an attempt is made to use such a condition.
6442 11. Added ${time_interval:xxxxx}.
6444 12. A bug was causing one of the ddress fields not to be passed back correctly
6445 from remote delivery subprocesses. The field in question was not being
6446 subsequently used, so this caused to problems in practice.
6448 13. Added new log selectors queue_time and deliver_time.
6450 14. Might have fixed a bug in maildirsizefile handling that threw up
6451 "unexpected character" debug warnings, and recalculated the data
6452 unnecessarily. In any case, I expanded the warning message to give more
6455 15. Added the message "Restricted characters in address" to the statements in
6456 the default ACL that block characters like @ and % in local parts.
6458 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6459 Three changes have been made:
6461 (a) There was a serious bug; a negative response to MAIL caused the whole
6462 recipient domain to be cached as invalid, thereby blocking all messages
6463 to all local parts at the same domain, from all senders. This bug has
6464 been fixed. The domain is no longer cached after a negative response to
6465 MAIL if the sender used is not empty.
6467 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6470 (c) A new callout option, "use_sender" has been added for people who want
6471 the modified behaviour.
6477 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6480 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6481 indeed breaks things for older releases.
6483 3. Added additional logging to the case where there is a problem reading data
6484 from a filter that is running in a subprocess using a pipe, in order to
6485 try to track down a specific problem.
6487 4. Testing facility fudge: when running in the test harness and attempting
6488 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6489 getting "No route to host". Convert this to a timeout.
6491 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6494 6. Some OS don't have socklen_t but use size_t instead. This affects the
6495 fifth argument of getsockopt() amongst other things. This is now
6496 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6497 can be set for individual OS. I have set it for SunOS5, OSF1, and
6498 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6499 some earlier ones do not.
6501 7. Change 4.30/15 was not doing the test caselessly.
6503 8. The standard form for an IPv6 address literal was being rejected by address
6504 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6505 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6506 this, as well as the form without the "IPv6" on the front (but only when
6507 address literals are enabled, of course).
6509 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6511 10. Exim crashed if a message with an empty sender address specified by -f
6512 encountered a router with an errors_to setting. This could be provoked only
6513 by a command such as
6517 where an empty string was supplied; "<>" did not hit this bug.
6519 11. Installed PCRE release 4.5.
6521 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6522 remained set. It is now erased.
6524 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6525 times from message ids (which are base 36 rather than the normal 62).
6527 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6528 were being counted as actual protocol errors, and logged if the log
6529 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6530 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6531 and DATA following a set of rejected RCPTs do not count as protocol errors.
6532 In other words, Exim assumes they were pipelined, though this may not
6533 actually be the case. Of course, in all cases the client gets an
6534 appropriate error code.
6536 15. If a lookup fails in an ACL condition, a message about the failure may
6537 be available; it is used if testing the ACL cannot continue, because most
6538 such messages specify what the cause of the deferral is. However, some
6539 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6540 that caused an old message to be retained and used if a later statement
6541 caused a defer, replacing the real cause of the deferral.
6543 16. If an IP address had so many PTR records that the DNS lookup buffer
6544 was not large enough to hold them, Exim could crash while trying to process
6545 the truncated data. It now detects and logs this case.
6547 17. Further to 4.21/58, another change has been made: if (and only if) the
6548 first line of a message (the first header line) ends with CRLF, a bare LF
6549 in a subsequent header line has a space inserted after it, so as not to
6550 terminate the header.
6552 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6553 unnecessarily, used atoi() instead of strtol(), and didn't check the
6554 termination when getting file sizes from file names by regex.
6556 19. Completely re-implemented the support for maildirsize files, in the light
6557 of a number of problems with the previous contributed implementation
6558 (4.30/29). In particular:
6560 . If the quota is zero, the maildirsize file is maintained, but no quota is
6563 . If the maildir directory does not exist, it is created before any attempt
6564 to write a maildirsize file.
6566 . The quota value in the file is just a cache; if the quota is changed in
6567 the transport, the new value overrides.
6569 . A regular expression is available for excluding directories from the
6572 20. The autoreply transport checks the characters in options that define the
6573 message's headers; it allows continued headers, but it was checking with
6574 isspace() after an embedded newline instead of explicitly looking for a
6577 21. If all the "regular" hosts to which an address was routed had passed their
6578 expiry times, and had not reached their retry times, the address was
6579 bounced, even if fallback hosts were defined. Now Exim should go on to try
6582 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6583 equivalent code in the SMTP transport. Some hosts send humungous responses
6584 to HELO/EHLO, more than 1024 it seems.
6586 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6587 gives compiler warnings in some environments. I've now done it "properly",
6590 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6591 (because of problems with the built-in one) was declared to return uschar *
6592 instead of char *, causing compiler failure.
6594 25. Fixed a file descriptor leak when processing alias/forward files.
6596 26. Fixed a minor format string issue in dbfn.c.
6598 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6600 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6601 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6604 29. When a sender address is verified, it is cached, to save repeating the test
6605 when there is more than one recipient in a message. However, when the
6606 verification involves a callout, it is possible for different callout
6607 options to be set for different recipients. It is too complicated to keep
6608 track of this in the cache, so now Exim always runs a verification when a
6609 callout is required, relying on the callout cache for the optimization.
6610 The overhead is duplication of the address routing, but this should not be
6613 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6614 to be verified with callout=postmaster, and the main callout worked but the
6615 postmaster check failed, the verification correctly failed. However, if a
6616 subsequent RCPT command asked for sender verification *without* the
6617 postmaster check, incorrect caching caused this verification also to fail,
6620 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6621 it was not caching the DNS options (qualify_single, search_parents) that
6622 were used when the lookup failed. A subsequent lookup with different
6623 options therefore always gave the same answer, though there were cases
6624 where it should not have. (Example: a "domains = !$mx_any" option on a
6625 dnslookup router: the "domains" option is always processed without any
6626 widening, but the router might have qualify_single set.) Now Exim uses the
6627 cached value only when the same options are set.
6629 32. Added John Jetmore's "exipick" utility to the distribution.
6631 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6632 than a timeout (e.g. a certificate is required, and is not provided), an
6633 Exim server now closes the connection immediately. Previously it waited for
6634 the client to close - but if the client is SSL, it seems that they each
6635 wait for each other, leading to a delay before one of them times out.
6637 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6638 maintained 0.8.x compatibility because I don't think many are using it, and
6639 it is clearly obsolete.
6641 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6644 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6645 client certificate was expired. A simple patch fixes this, though I don't
6646 understand the full logic of why the verify callback is called multiple
6649 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6650 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6651 which causes problems with some clients (such as the Certicom SSL Plus
6652 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6653 disables the coutermeasure allowing Eudora to connect."
6655 38. Exim was not checking that a write() to a log file succeeded. This could
6656 lead to Bad Things if a log got too big, in particular if it hit a file
6657 size limit. Exim now panics and dies if it cannot write to a log file, just
6658 as it does if it cannot open a log file.
6660 39. Modified OS/Makefile-Linux so that it now contains
6662 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6664 The two -D definitions ensure that Exim is compiled with large file
6665 support, which makes it possible to handle log files that are bigger than
6668 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6669 instance) a domain was checked against a named list that involved a lookup,
6670 causing $domain_data to be set, then another domain was checked against the
6671 same list, then the first domain was re-checked, the value of $domain_data
6672 after the final check could be wrong. In particular, if the second check
6673 failed, it could be set empty. This bug probably also applied to
6676 41. The strip_trailing_dot option was not being applied to the address given
6677 with the -f command-line option.
6679 42. The code for reading a message's header from the spool was incrementing
6680 $received_count, but never initializing it. This meant that the value was
6681 incorrect (doubled) while delivering a message in the same process in which
6682 it was received. In the most common configuration of Exim, this never
6683 happens - a fresh exec is done - but it can happen when
6684 deliver_drop_privilege is set.
6686 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6687 it now includes up to 150 characters of the unexpected data in the log
6690 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6691 and building data strings. The size of both of these buffers was 10 000
6692 bytes - far larger than anybody would *ever* want, thought I. Needless to
6693 say, somebody hit the limit. I have increased the maximum line length to
6694 20 000 and the maximum data length of concatenated lines to 100 000. I have
6695 also fixed two bugs, because there was no checking on these buffers. Tsk,
6696 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6697 buffer is too small.
6699 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6700 lsearch lookups. Now it does.
6702 46. When parsing a route_list item in a manualroute router, a fixed-length
6703 buffer was used for the list of hosts. I made this 1024 bytes long,
6704 thinking that nobody would ever have a list of hosts that long. Wrong.
6705 Somebody had a whole pile of complicated expansion conditions, and the
6706 string was silently truncated, leading to an expansion error. It turns out
6707 that it is easier to change to an unlimited length (owing to other changes
6708 that have happened since this code was originally written) than to build
6709 structure for giving a limitation error. The length of the item that
6710 expands into the list of hosts is now unlimited.
6712 47. The lsearch lookup could not handle data where the length of text line was
6713 more than 4095 characters. Such lines were truncated, leading to shortened
6714 data being returned. It should now handle lines of any length.
6716 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6717 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6720 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6721 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6722 no longer happens in comment lines. A list of replacements is now placed
6723 at the head of all of the source files, except those whose only change is
6724 to replace PERL_COMMAND in the very first #! line.
6726 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6727 messages on the queue, with a bottom-up merge sort, using code contributed
6728 by Michael Haardt. This should make operations like -bp somewhat faster on
6729 large queues. It won't affect queue runners, except when queue_run_in_order
6732 51. Installed eximstats 1.31 in the distribution.
6734 52. Added support for SRV lookups to the dnslookup router.
6736 53. If an ACL referred to $message_body or $message_body_end, the value was not
6737 reset for any messages that followed in the same SMTP session.
6739 54. The store-handling optimization for building very long strings was not
6740 differentiating between the different store pools. I don't think this
6741 actually made any difference in practice, but I've tidied it.
6743 55. While running the routers to verify a sender address, $sender_address
6744 was still set to the sender address. This is wrong, because when routing to
6745 send a bounce to the sender, it would be empty. Therefore, I have changed
6746 it so that, while verifying a sender address, $sender_address is set to <>.
6747 (There is no change to what happens when verifying a recipient address.)
6749 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6750 target A or AAAA records (if not already returned) without resetting the
6751 qualify_single or search_parents options of the DNS resolver. These are
6752 inappropriate in this case because the targets of MX and SRV records must
6753 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6754 target that, when qualified, matched something in the local domain. These
6755 two options are now turned off when doing these lookups.
6757 57. It seems that at least some releases of Reiserfs (which does not have the
6758 concept of a fixed number of inodes) returns zero and not -1 for the
6759 number of available inodes. This interacted badly with check_spool_inodes,
6760 which assumed that -1 was the "no such thing" setting. What I have done is
6761 to check that the total number of inodes is greater than zero before doing
6762 the test of how many are available.
6764 58. When a "warn" ACL statement has a log_message modifier, the message is
6765 remembered, and not repeated. This is to avoid a lot of repetition when a
6766 message has many recipients that cause the same warning to be written.
6767 However, Exim was preserving the list of already written lines for an
6768 entire SMTP session, which doesn't seem right. The memory is now reset if a
6769 new message is started.
6771 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6772 debugging output unless log_rewrite was also set.
6774 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6775 of (int)(handle) when we know that handle contains (void *)(-1).
6777 61. The Exim daemon panic-logs an error return when it closes the incoming
6778 connection. However "connection reset by peer" seems to be common, and
6779 isn't really an error worthy of noting specially, so that particular error
6782 62. When Exim is trying to find all the local interfaces, it used to panic and
6783 die if the ioctl to get the interface flags failed. However, it seems that
6784 on at least one OS (Solaris 9) it is possible to have an interface that is
6785 included in the list of interfaces, but for which you get a failure error
6786 for this call. This happens when the interface is not "plumbed" into a
6787 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6788 failure of the "get flags" call assumes that the interface is down.
6790 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6791 makes life easier for people who are doing arithmetic on fields extracted
6792 from dates, where you often get leading zeros that should not be
6793 interpreted as octal.
6795 64. Added qualify_domain to the redirect router, to override the global
6798 65. If a pathologically long header line contained very many addresses (the
6799 report of this problem mentioned 10 000) and each of them was rewritten,
6800 Exim could use up a very large amount of memory. (It kept on making new
6801 copies of the header line as it rewrote, and never released the old ones.)
6802 At the expense of a bit more processing, the header rewriting function has
6803 been changed so that it no longer eats memory in this way.
6805 66. The generation of the Received: header has been moved from the time that a
6806 message starts to be received, to the time that it finishes. The timestamp
6807 in the Received: header should now be very close to that of the <= log
6808 line. There are two side-effects of this change:
6810 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6811 logged header lines no longer include the local Received: line, because
6812 it has not yet been created. The same applies to a copy of the message
6813 that is returned to a non-SMTP sender when a message is rejected.
6815 (b) When a filter file is tested using -bf, no additional Received: header
6816 is added to the test message. After some thought, I decided that this
6819 This change does not affect the value of $received_for. It is still set
6820 after address rewriting, but before local_scan() is called.
6822 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6824 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6825 gave an unhelpful panic error message, and a defer error. I have managed to
6826 change this behaviour so that it now rejects any supplied certificate,
6827 which seems right, as the list of acceptable certificates is empty.
6829 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6830 gave an unhelpful defer error. I have not managed to make this reject any
6831 supplied certificates, but the error message it gives is "no certificate
6832 supplied", which is not helpful.
6834 70. exigrep's output now also includes lines that are not associated with any
6835 message, but which match the given pattern. Implemented by a patch from
6836 Martin Sluka, which also tidied up the Perl a bit.
6838 71. Recipient callout verification, like sender verification, was using <> in
6839 the MAIL FROM command. This isn't really the right thing, since the actual
6840 sender may affect whether the remote host accepts the recipient or not. I
6841 have changed it to use the actual sender in the callout; this means that
6842 the cache record is now keyed on a recipient/sender pair, not just the
6843 recipient address. There doesn't seem to be a real danger of callout loops,
6844 since a callout by the remote host to check the sender would use <>.
6845 [SEE ABOVE: changed after hitting problems.]
6847 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6848 temporary errors. However, in the case of such a code being given after
6849 the end of a data transmission (i.e. after ".") Exim was failing to write
6850 a retry record for the message. (Yes, there was some broken host that was
6851 actually sending 8xx at this point.)
6853 73. An unknown lookup type in a host list could cause Exim to panic-die when
6854 the list was checked. (An example that provoked this was putting <; in the
6855 middle of a list instead of at the start.) If this happened during a DATA
6856 ACL check, a -D file could be left lying around. This kind of configuration
6857 error no longer causes Exim to die; instead it causes a defer error. The
6858 incident is still logged to the main and panic logs.
6860 74. Buglet left over from Exim 3 conversion. The message "too many messages
6861 in one connection" was written to the rejectlog but not the mainlog, except
6862 when address rewriting (yes!) was being logged.
6864 75. Added write_rejectlog option.
6866 76. When a system filter was run not as root (that is, when system_filter_user
6867 was set), the values of the $n variables were not being returned to the
6868 main process; thus, they were not subsequently available in the $sn
6871 77. Added +return_path_on_delivery log selector.
6873 78. A connection timeout was being treated differently from recipients deferred
6874 when testing hosts_max_try with a message that was older than the host's
6875 retry timeout. (The host should not be counted, thus allowing all hosts to
6876 be tried at least once before bouncing.) This may have been the cause of an
6877 occasionally reported bug whereby a message would remain on the queue
6878 longer than the retry timeout, but would be bounced if a delivery was
6879 forced. I say "may" because I never totally pinned down the problem;
6880 setting up timeout/retry tests is difficult. See also the next item.
6882 79. The ultimate address timeout was not being applied to errors that involved
6883 a combination of host plus message (for example, a timeout on a MAIL
6884 command). When an address resolved to a number of possible hosts, and they
6885 were not all tried for each delivery (e.g. because of hosts_max_try), a
6886 message could remain on the queue longer than the retry timeout.
6888 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6891 81. Fixed an obscure SMTP outgoing bug which required at least the following
6892 conditions: (a) there was another message waiting for the same server;
6893 (b) the server returned 5xx to all RCPT commands in the first message so
6894 that the message was not completed; (c) the server dropped the connection
6895 or gave a negative response to the RSET that Exim sends to abort the
6896 transaction. The observed case was a dropped connection after DATA that had
6897 been sent in pipelining mode. That is, the server had advertised PIPELINING
6898 but was not implementing it correctly. The effect of the bug was incorrect
6899 behaviour, such as trying another host, and this could lead to a crash.
6905 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6906 and daemon.c were passed as pointers to ints; they should have been
6907 pointers to socklen_t variables (which are typically unsigned ints).
6909 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6912 3. Fixed a really odd bug that affected only the testing scheme; patching a
6913 certain fixed string in the binary changed the value of another string that
6914 happened to be identical to the end of the original first string.
6916 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6917 name", it returns that address as the IP address. On some operating
6918 systems (e.g. Solaris), it also passes back the IP address string as the
6919 "host name". However, on others (e.g. Linux), it passes back an empty
6920 string. Exim wasn't checking for this, and was changing the host name to an
6921 empty string, assuming it had been canonicalized.
6923 5. Although rare, it is permitted to have more than one PTR record for a given
6924 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6925 all the names associated with an address, because they do in Solaris.
6926 However, it seems that they do not in Linux for data that comes from the
6927 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6928 I found this out when I moved to a new Linux workstation and tried to run
6929 the Exim test suite.
6931 To get round this problem I have changed the code so that it now does its
6932 own call to the DNS to look up PTR records when searching for a host name.
6933 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6934 addresses that are only in /etc/hosts are still found.
6936 This behaviour is, however, controlled by an option called host_lookup_
6937 order, which defaults to "bydns:byaddr". If people want to use the other
6938 order, or indeed, just use one or the other means of lookup, they can
6939 specify it in this variable.
6941 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6942 some operating systems, this comes back from gethostbyaddr() as an empty
6943 string, and this is what Exim used to test for. However, it seems that in
6944 other systems, "." is yielded. Exim now tests for this case too.
6946 7. The values of check_spool_space and check_log_space are now held internally
6947 as a number of kilobytes instead of an absolute number of bytes. If a
6948 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6949 kilobyte. This means that much larger values can be stored.
6951 8. Exim monitor: an attempt to get the action menu when not actually pointing
6952 at a message produces an empty menu entitled "No message selected". This
6953 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6954 no entries in it ("Shell widget menu has zero width and/or height"). So I
6955 have added a single, blank menu entry in this case.
6957 9. Added ${quote_local_part.
6959 10. MIME decoding is now applied to the contents of Subject: header lines when
6962 11. Now that a reference to $sender_host_address automatically causes a reverse
6963 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6964 host lookup before query-style lookups in lists that might use this
6965 variable. This has therefore been abolished, and the "net-" prefix is no
6966 longer necessary for query-style lookups.
6968 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6969 have been a typo for LFLAGS, so it has been changed.
6971 13. The install script calls Exim with "-C /dev/null" in order to find the
6972 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6973 to be output. However, since Exim outputs its version number before the
6974 error, it didn't break the script. It just looked ugly. I fixed this by
6975 always allowing "-C /dev/null" if the caller is root.
6977 14. Ignore overlarge ACL variable number when reading spool file - insurance
6978 against a later release with more variables having written the file.
6980 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6981 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6982 this, as well as the form without the "IPv6" on the front.
6984 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6985 OS/Makefile-Darwin file.
6987 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6988 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6990 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6991 variable. However, some OS use ints for inodes. I've added cast to long int
6992 to get rid of the compiler warning.
6994 19. I had forgotten to lock out "/../" in configuration file names when
6995 ALT_CONFIG_PREFIX was set.
6997 20. Routers used for verification do not need to specify transports. However,
6998 if such a router generated a host list, and callout was configured, Exim
6999 crashed, because it could not find a port number from the (non-existent)
7000 transport. It now assumes port 25 in this circumstance.
7002 21. Added the -t option to exigrep.
7004 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7005 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7006 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7007 them is set, LOOKUP_LSEARCH is forced.
7009 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7011 24. Added sender and host information to the "rejected by local_scan()" log
7012 line; previously there was no indication of these.
7014 25. Added .include_if_exists.
7016 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7017 Linux. It turns out that not all file systems support this. Apparently some
7018 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7019 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7020 not supported on the file descriptor, is now ignored when Exim is trying to
7021 sync a directory. This applies only to Linux.
7023 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7025 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7026 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7029 29. Added support for maildirsize files from supplied patch (modified a bit).
7031 30. The use of :fail: followed by an empty string could lead Exim to respond to
7032 sender verification failures with (e.g.):
7034 550 Verification failed for <xxx>
7035 550 Sender verify failed
7037 where the first response line was missing the '-' that indicates it is not
7038 the final line of the response.
7040 31. The loop for finding the name of the user that called Exim had a hardwired
7041 limit of 10; it now uses the value of finduser_retries, which is used for
7042 all other user lookups.
7044 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7047 33. Exim was neglecting to zero errno before one call of strtol() when
7048 expanding a string and expecting an integer value. On some systems this
7049 resulted in spurious "integer overflow" errors. Also, it was casting the
7050 result into an int without checking.
7052 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7053 did not work. The code looks as if it has *never* worked, though it appears
7054 to have been documented since at least release 1.62. I have made it work.
7056 35. The "timeout_DNS" error in retry rules, also documented since at least
7057 1.62, also never worked. As it isn't clear exactly what this means, and
7058 clearly it isn't a major issue, I have abolished the feature by treating it
7059 as "timeout", and writing a warning to the main and panic logs.
7061 36. The display of retry rules for -brt wasn't always showing the error code
7064 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7065 timeout_connect_A, timeout_connect_MX.
7067 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7068 to the empty sender.
7070 39. The daemon was not analysing the content of -oX till after it had closed
7071 stderr and disconnected from the controlling terminal. This meant that any
7072 syntax errors were only noted on the panic log, and the return code from
7073 the command was 0. By re-arranging the code a little, I've made the
7074 decoding happen first, so such errors now appear on stderr, and the return
7075 code is 1. However, the actual setting up of the sockets still happens in
7076 the disconnected process, so errors there are still only recorded on the
7079 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7080 connections (as happens on some IP stacks) was logged at start up time as
7081 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7082 it from "IPv6 and IPv4", which means that two separate sockets are being
7085 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7086 says whether AF_INET or AF_INET6 was passed as an argument.
7088 42. Exiwhat output was messed up when time zones were included in log
7091 43. Exiwhat now gives more information about the daemon's listening ports,
7092 and whether -tls-on-connect was used.
7094 44. The "port" option of the smtp transport is now expanded.
7096 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7097 silently ignored. Now an error message is written to the main and panic
7100 46. There's a new ACL modifier called "logwrite" which writes to a log file
7101 as soon as it is encountered.
7103 47. Added $local_user_uid and $local_user_gid at routing time.
7105 48. Exim crashed when trying to verify a sender address that was being
7108 49. Exim was recognizing only a space character after ".include". It now also
7109 recognizes a tab character.
7111 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7112 extracting the relevant information from the specification. The man page no
7113 longer contains scrambled data for the -d option, and I've added a section
7114 at the front about calling Exim under different names.
7116 51. Added "extra_headers" argument to the "mail" command in filter files.
7118 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7121 53. Installed eximstats 1.29.
7123 54. Added transport_filter_timeout as a generic transport option.
7125 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7126 Cc: header lines. This was required by RFC 822, but it not required by RFC
7129 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7130 incoming messages that did not have them. Now it does so only if the
7131 message originates locally, that is, if there is no associated remote host
7132 address. When Resent- header lines are present, this applies to the Resent-
7133 lines rather than the non-Resent- lines.
7135 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7136 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7138 58. Messages for configuration errors now include the name of the main
7139 configuration file - useful now that there may be more than one file in a
7140 list (.included file names were always shown).
7142 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7143 for those rare installations that do not start the daemon as root or run it
7144 setuid root. I've cut out the call to initgroups() if the daemon is not
7147 60. The Exim user and group can now be bound into the binary as text strings
7148 that are looked up at the start of Exim's processing.
7150 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7152 62. Added $mailstore_basename variable.
7154 63. Installed patch to sieve.c from Michael Haardt.
7156 64. When Exim failed to open the panic log after failing to open the main log,
7157 the original message it was trying to log was written to stderr and debug
7158 output, but if they were not available (the usual case in production), it
7159 was lost. Now it is written to syslog before the two lines that record the
7160 failures to open the logs.
7162 65. Users' Exim filters run in subprocesses under the user's uid. It is
7163 possible for a "deliver" command or an alias in a "personal" command to
7164 provoke an address rewrite. If logging of address rewriting is configured,
7165 this fails because the process is not running as root or exim. There may be
7166 a better way of dealing with this, but for the moment (because 4.30 needs
7167 to be released), I have disabled address rewrite logging when running a
7168 filter in a non-root, non-exim process.
7174 1. The buildconfig auxiliary program wasn't quoting the value set for
7175 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7176 not defined. This bug was masked in 4.22 by the effect that was fixed in
7179 2. Some messages that were rejected after a message id was allocated were
7180 shown as "incomplete" by exigrep. It no longer does this for messages that
7181 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7183 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7184 have allow_domain_literals set, the ID did not get logged in the <= line.
7185 Domain literals are now always recognized in Message-ID: header lines.
7187 4. The first argument for a ${extract expansion item is the key name or field
7188 number. Leading and trailing spaces in this item were not being ignored,
7189 causing some misleading effects.
7191 5. When deliver_drop_privilege was set, single queue runner processes started
7192 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7193 same command in the process it spins off) were not dropping privilege.
7195 6. When the daemon running as "exim" started a queue runner, it always
7196 re-executed Exim in the spun-off process. This is a waste of effort when
7197 deliver_drop_privilege is set. The new process now just calls the
7198 queue-runner function directly.
7204 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7207 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7208 ignored. Though the use of -oP was forcing the writing of a pid file, it
7209 was always written to the default place.
7211 3. If the message "no IP address found for host xxxx" is generated during
7212 incoming verification, it is now followed by identification of the incoming
7213 connection (so you can more easily find what provoked it).
7215 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7217 5. Added some features to "harden" Exim a bit more against certain attacks:
7219 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7220 be put in Local/Makefile. This is like the never_users runtime option,
7221 but it cannot be overridden. The default setting is "root".
7223 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7224 prefix string with which any file named in a -C command line option
7227 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7228 is retained for -C and -D only if the caller of Exim is root. Without
7229 it, the exim user may also use -C and -D and retain privilege.
7231 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7232 command line option is disabled.
7234 6. Macro names set by the -D option must start with an upper case letter, just
7235 like macro names defined in the configuration file.
7237 7. Added "dereference=" facility to LDAP.
7239 8. Two instances of the typo "uknown" in the source files are fixed.
7241 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7242 the Configure-Makefile script screwed up while processing it.
7244 10. Incorporated PCRE 4.4.
7246 11. The SMTP synchronization check was not operating right at the start of an
7247 SMTP session. For example, it could not catch a HELO sent before the client
7248 waited for the greeting. There is now a check for outstanding input at the
7249 point when the greeting is written. Because of the duplex, asynchronous
7250 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7251 way, but not yet received, when the check is performed.
7253 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7254 on TCP/IP sockets, because this apparently causes some broken clients to
7257 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7258 unchanged) from the Cygwin maintainer.
7260 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7261 was supported instead of testing for mbx. Effectively a typo.
7263 15. The spa authenticator server code was not checking that the input it
7264 received was valid base64.
7266 16. The debug output line for the "set" modifier in ACLs was not showing the
7267 name of the variable that was being set.
7269 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7271 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7272 lookup on its own. Something else was needed to trigger the lookup. For
7273 example, a match in host_lookup or the need for a host name in a host list.
7274 Now, if $sender_host_name is referenced and the host name has not yet been
7275 looked up, a lookup is performed. If the lookup fails, the variable remains
7276 empty, and $host_lookup_failed is set to "1".
7278 19. Added "eqi" as a case-independent comparison operator.
7280 20. The saslauthd authentication condition could segfault if neither service
7281 nor realm was specified.
7283 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7284 error message that was logged was misleading, and incoming SMTP
7285 connections were dropped. The message is now more accurate, and temporary
7286 errors are given to SMTP connections.
7288 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7289 (except RSET) with a 421 error, until QUIT is received. However, it was
7290 failing to send a response to QUIT.
7292 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7293 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7294 ACL is now run after the helo_try_verify_hosts code.
7296 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7297 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7298 case-independent, so other case variants are also recognized). Apparently
7299 some systems use these upper case variants.
7301 25. If more than two messages were waiting for the same host, and a transport
7302 filter was specified for the transport, Exim sent two messages over the
7303 same TCP/IP connection, and then failed with "socket operation on non-
7304 socket" when it tried to send the third.
7306 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7308 27. The extern definition of crypt16() in expand.c was not being excluded when
7309 the OS had its own crypt16() function.
7311 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7312 as a preferred synonym for return_size_limit, both as an option and as an
7315 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7317 30. Changed the default configuration ACL to relax the local part checking rule
7318 for addresses that are not in any local domains. For these addresses,
7319 slashes and pipe symbols are allowed within local parts, but the sequence
7320 /../ is explicitly forbidden.
7322 31. SPA server authentication was not clearing the challenge buffer before
7325 32. log_message in a "warn" ACL statement was writing to the reject log as
7326 well as to the main log, which contradicts the documentation and doesn't
7327 seem right (because no rejection is happening). So I have stopped it.
7329 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7330 However, I am unable to do any testing of this.
7332 34. Fixed an infelicity in the appendfile transport. When checking directories
7333 for a mailbox, to see if any needed to be created, it was accidentally
7334 using path names with one or more superfluous leading slashes; tracing
7335 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7337 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7338 contents are added to the log line that is written for every discarded
7339 recipient. (Previously a log_message setting was ignored.)
7341 36. The ${quote: operator now quotes the string if it is empty.
7343 37. The install script runs exim in order to find its version number. If for
7344 some reason other than non-existence or emptiness, which it checks, it
7345 could not run './exim', it was installing it with an empty version number,
7346 i.e. as "exim-". This error state is now caught, and the installation is
7349 38. An argument was missing from the function that creates an error message
7350 when Exim fails to connect to the socket for saslauthd authentication.
7351 This could cause Exim to crash, or give a corrupted message.
7353 39. Added isip, isip4, and isip6 to ${if conditions.
7355 40. The ACL variables $acl_xx are now saved with the message, and can be
7356 accessed later in routers, transports, and filters.
7358 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7359 strings in the file are not string-expanded.
7361 42. If a MAIL command specified a SIZE value that was too large to fit into an
7362 int variable, the check against message_size_limit failed. Such values are
7363 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7364 day this will have to be increased, but I don't think I want to be around
7365 when emails are that large.
7372 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7373 iconv() is not standard in FreeBSD.
7375 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7376 IPv6 enabled. The observed symptom was a segmentation fault on return from
7377 the function os_common_find_running_interfaces() in src/os.c.
7379 3. In the check_special_case() function in daemon.c I had used "errno" as an
7380 argument name, which causes warnings on some systems. This was basically a
7381 typo, since it was named "eno" in the comments!
7383 4. The code that waits for the clock to tick (at a resolution of some fraction
7384 of a second) so as to ensure message-id uniqueness was always waiting for
7385 at least one whole tick, when it could have waited for less. [This is
7386 almost certainly not relevant at current processor speeds, where it is
7387 unlikely to ever wait at all. But we try to future-proof.]
7389 5. The function that sleeps for a time interval that includes fractions of a
7390 second contained a race. It did not block SIGALRM between setting the
7391 timer, and suspending (a couple of lines later). If the interval was short
7392 and the sigsuspend() was delayed until after it had expired, the suspension
7393 never ended. On busy systems this could lead to processes getting stuck for
7396 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7397 process, before it forks any delivery processes. The open lookup caching
7398 mechanism meant that the open file or database connection was passed into
7399 the delivery process. The problem was that delivery processes always tidy
7400 up cached lookup data. This could cause a problem for the next delivery
7401 process started by the queue runner, because the external queue runner
7402 process does not know about the closure. So the next delivery process
7403 still has data in the lookup cache. In the case of a file lookup, there was
7404 no problem because closing a file descriptor in a subprocess doesn't affect
7405 the parent. However, if the lookup was caching a connection to a database,
7406 the connection was closed, and the second delivery process was likely to
7407 see errors such as "PGSQL: query failed: server closed the connection
7408 unexpectedly". The problem has been fixed by closing all cached lookups
7409 in a queue runner before running a delivery process.
7411 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7412 seem to have the "const" qualifier which it has on other OS. I've
7415 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7416 *of the same type* (client or server) with the same public name that an
7417 error should be diagnosed.
7419 9. When Exim looked up a host name for an IP address, but failed to find the
7420 original IP address when looking up the host name (a safety check), it
7421 output the message "<ip address> does not match any IP for NULL", which was
7422 confusing, to say the least. The bug was that the host name should have
7423 appeared instead of "NULL".
7425 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7426 user that is built into the binary, and the -C or -D options is used, root
7427 privilege is dropped before the configuration file is read. In addition,
7428 logging is switched to stderr instead of the normal log files. If the
7429 configuration then re-defines the Exim user, the unprivileged environment
7430 is probably not what is expected, so Exim logs a panic warning message (but
7433 However, if deliver_drop_privilege is set, the unprivileged state may well
7434 be exactly what is intended, so the warning has been cut out in that case,
7435 and Exim is allowed to try to write to its normal log files.
7441 1. smtp_return_error_details was not giving details for temporary sender
7442 or receiver verification errors.
7444 2. Diagnose a configuration error if two authenticators have the same public
7447 3. Exim used not to create the message log file for a message until the first
7448 delivery attempt. This could be confusing when incoming messages were held
7449 for policy or load reasons. The message log file is now created at the time
7450 the message is received, and an initial "Received" line is written to it.
7452 4. The automatically generated man page for command line options had a minor
7453 bug that caused no ill effects; however, a more serious problem was that
7454 the procedure for building the man page automatically didn't always
7455 operate. Consequently, release 4.20 contains an out-of-date version. This
7456 shouldn't happen again.
7458 5. When building Exim with embedded Perl support, the script that builds the
7459 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7460 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7462 6. The freeze_tell option was not being used for messages that were frozen on
7463 arrival, either by an ACL or by local_scan().
7465 7. Added the smtp_incomplete_transaction log selector.
7467 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7468 was accepting AUTH without a new EHLO.
7470 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7471 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7474 10. Logging of TCP/IP connections (when configured) now happens in the main
7475 daemon process instead of the child process, so that the TCP/IP connection
7476 count is more accurate (but it can never be perfect).
7478 11. The use of "drop" in a nested ACL was not being handled correctly in the
7479 outer ACL. Now, if condition failure induced by the nested "drop" causes
7480 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7481 or "require"), the connection is dropped.
7483 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7484 that yield "discard" can now be used with an "accept" or a "discard" verb,
7485 but an error is generated for any others (because I can't see a useful way
7486 to define what should happen).
7488 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7489 are now processed in the same way as lines in the Exim configuration file.
7490 In particular, continuation lines are supported.
7492 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7494 15. Added -ti meaning -t -i.
7496 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7497 domains, and warn by logging if others are found.
7499 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7500 returned from GIFCONF when Exim is trying to find the list of interfaces on
7501 a host. The code in os.c has been modified to copy each ifreq to an aligned
7502 structure in all cases.
7504 Also, in some cases, the returned ifreq's were being copied to a 'struct
7505 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7506 means the last couple of bytes of an IPv6 address could be chopped if the
7507 ifreq contained only a normal sockaddr (14 bytes storage).
7509 18. Named domain lists were not supported in the hosts_treat_as_local option.
7510 An entry such as +xxxx was not recognized, and was treated as a literal
7513 19. Ensure that header lines added by a DATA ACL are included in the reject log
7514 if the ACL subsequently rejects the message.
7516 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7517 MD5 (which is deprecated).
7519 21. When testing a filter file using -bf, Exim was writing a message when it
7520 took the sender from a "From " line in the message, but it was not doing so
7521 when it took $return_path from a Return-Path: header line. It now does.
7523 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7524 with a valid header line field name (a series of printing characters
7525 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7527 23. Changed "disc" in the source to "disk" to conform to the documentation and
7528 the book and for uniformity.
7530 24. Ignore Sendmail's -Ooption=value command line item.
7532 25. When execve() failed while trying to run a command in a pipe transport,
7533 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7534 could be confused with a return value of 69 from the command itself. This
7535 has been changed to 127, the value the shell returns if it is asked to run
7536 a non-existent command. The wording for the related log line suggests a
7537 non-existent command as the problem.
7539 26. If received_header_text expands to an empty string, do not add a Received:
7540 header line to the message. (Well, it adds a token one on the spool, but
7541 marks it "old" so that it doesn't get used or transmitted.)
7543 27. Installed eximstats 1.28 (addition of -nt option).
7545 28. There was no check for failure on the call to getsockname() in the daemon
7546 code. This can fail if there is a shortage of resources on the system, with
7547 ENOMEM, for example. A temporary error is now given on failure.
7549 29. Contrary to the C standard, it seems that in some environments, the
7550 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7551 program. Exim now does this explicitly; it affects the formatting of
7552 timestamps using strftime().
7554 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7555 complaints. I've now initialized all the variables, to avoid this.
7557 32. Header lines added by a system filter were not being "seen" during
7558 transport-time rewrites.
7560 33. The info_callback() function passed to OpenSSL is set up with type void
7561 (*)(SSL *, int, int), as described somewhere. However, when calling the
7562 function (actually a macro) that sets it up, the type void(*)() is
7563 expected. I've put in a cast to prevent warnings from picky compilers.
7565 34. If a DNS black list lookup found a CNAME record, but there were no A
7566 records associated with the domain it pointed at, Exim crashed.
7568 35. If a DNS black list lookup returned more than one A record, Exim ignored
7569 all but the first. It now scans all returned addresses if a particular IP
7570 value is being sought. In this situation, the contents of the
7571 $dnslist_value variable are a list of all the addresses, separated by a
7574 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7575 to accept a host name and all its aliases if the forward lookup for any of
7576 them yielded the IP address of the incoming connection. Now it accepts only
7577 those names whose forward lookup yields the correct IP address. Any other
7578 names are discarded. This closes a loophole whereby a rogue DNS
7579 administrator could create reverse DNS records to break through a
7580 wildcarded host restriction in an ACL.
7582 37. If a user filter or a system filter that ran in a subprocess used any of
7583 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7584 the wrong values were passed to the pipe command ($thisaddress had the
7585 value of $0, $0 had the value of $1, etc). This bug was introduced by
7586 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7588 38. Improved the line breaking for long SMTP error messages from ACLs.
7589 Previously, if there was no break point between 40 and 75 characters, Exim
7590 left the rest of the message alone. Two changes have been made: (a) I've
7591 reduced the minimum length to 35 characters; (b) if it can't find a break
7592 point between 35 and 75 characters, it looks ahead and uses the first one
7593 that it finds. This may give the occasional overlong line, but at least the
7594 remaining text gets split now.
7596 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7597 file descriptors might be low, and that setting 1000 would always raise it.
7598 It turns out that in some environments, the limit is already over 1000 and
7599 that lowering it causes trouble. So now Exim takes care not to decrease it.
7601 40. When delivering a message, the value of $return_path is set to $sender_
7602 address at the start of routing (routers may change the value). By an
7603 oversight, this default was not being set up when an address was tested by
7604 -bt or -bv, which affected the outcome if any router or filter referred to
7607 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7608 log only when "message" or "log_message" are set. However, if one of the
7609 conditions was an address verification, or a call to a nested ACL, the
7610 messages generated by the underlying test were being passed through. This
7611 no longer happens. The underlying message is available in $acl_verify_
7612 message for both "message" and "log_message" expansions, so it can be
7613 passed through if needed.
7615 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7616 new expansion $bh_ to give the encoded byte string without charset
7617 translation. Translation happens only if iconv() is available; HAVE_ICONV
7618 indicates this at build time. HEADERS_CHARSET gives the charset to
7619 translate to; headers_charset can change it in the configuration, and
7620 "headers charset" can change it in an individual filter file.
7622 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7623 that creates RFC 2047 encoded "words" labels them as that charset instead
7624 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7625 expansion operator; (ii) when Exim creates a From: line for a local
7626 message; (iii) when a header line is rewritten to include a "phrase" part.
7628 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7629 buggy, causing it to skip the first lines of messages whose message ID
7630 ended in 'D'. This would not have bitten before Exim release 4.14, because
7631 message IDs were unlikely to end in 'D' before then. The effect was to have
7632 incorrect size information for certain domains.
7634 45. #include "config.h" was missing at the start of the crypt16.c module. This
7635 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7638 46. If there was a timeout during a "random" callout check, Exim treated it as
7639 a failure of the random address, and carried on sending RSET and the real
7640 address. If the delay was just some slowness somewhere, the response to the
7641 original RCPT would be taken as a response to RSET and so on, causing
7642 mayhem of various kinds.
7644 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7645 when I implemented it. It didn't allow for the fact that some option values
7646 may legitimately be negative (e.g. size_addition), and it didn't even do
7647 the right test for positive values.
7649 48. Domain names in DNS records are case-independent. Exim always looks them up
7650 in lower case. Some resolvers return domain names in exactly the case they
7651 appear in the zone file, that is, they may contain uppercase letters. Not
7652 all resolvers do this - some return always lower case. Exim was treating a
7653 change of case by a resolver as a change of domain, similar to a widening
7654 of a domain abbreviation. This triggered its re-routing code and so it was
7655 trying to route what was effectively the same domain again. This normally
7656 caused routing to fail (because the router wouldn't handle the domain
7657 twice). Now Exim checks for this case specially, and just changes the
7658 casing of the domain that it ultimately uses when it transmits the message
7661 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7664 50. If a filter generated a file delivery with a non-absolute name (possible if
7665 no home directory exists for the router), the forbid_file option was not
7668 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7669 the existing equality matching.
7671 52. Exim was using ints instead of ino_t variables in some places where it was
7672 dealing with inode numbers.
7674 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7675 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7676 called TMPDIR, and if it finds it is different, it changes its value.
7678 54. The smtp_printf() function is now made available to local_scan() so
7679 additional output lines can be written before returning. There is also an
7680 smtp_fflush() function to enable the detection of a dropped connection.
7681 The variables smtp_input and smtp_batched_input are exported to
7684 55. Changed the default runtime configuration: the message "Unknown user"
7685 has been removed from the ACL, and instead placed on the localuser router,
7686 using the cannot_route_message feature. This means that any verification
7687 failures that generate their own messages won't get overridden. Similarly,
7688 the "Unrouteable address" message that was in the ACL for unverifiable
7689 relay addresses has also been removed.
7691 56. Added hosts_avoid_esmtp to the smtp transport.
7693 57. The exicyclog script was not checking for the esoteric option
7694 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7695 will work only if exicyclog is run under the appropriate euid.
7697 58. Following a discussion on the list, the rules by which Exim recognises line
7698 endings on incoming messages have been changed. The -dropcr and drop_cr
7699 options are now no-ops, retained only for backwards compatibility. The
7700 following line terminators are recognized: LF CRLF CR. However, special
7701 processing applies to CR:
7703 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7704 nor a local message in the state where . is a terminator.
7706 (ii) If a bare CR is encountered in a header line, an extra space is added
7707 after the line terminator so as not to end the header. The reasoning
7708 behind this is that bare CRs in header lines are most likely either
7709 to be mistakes, or people trying to play silly games.
7711 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7712 was being incorrectly given as 18 bytes larger than it should have been.
7713 This is a VOB (very old bug).
7715 60. This may never have affected anything current, but just in case it has:
7716 When the local host is found other than at the start of a list of hosts,
7717 the local host, those with the same MX, and any that follow, are discarded.
7718 When the list in question was part of a longer list of hosts, the following
7719 hosts (not currently being processed) were also being discarded. This no
7720 longer happens. I'm not sure if this situation could ever has previously
7723 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7726 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7727 header. This is something that is recommended in a new Internet Draft, and
7728 is something that is documented as being done by Sendmail. There are two
7729 possible values. For messages generated by the autoreply transport, Exim
7732 Auto-Submitted: auto-replied
7734 whereas for all other generated messages (e.g. bounces) it adds
7736 Auto-Submitted: auto-generated
7738 63. The "personal" condition in filters now includes a test for the
7739 Auto-Submitted: header. If it contains the string "auto-" the message it
7740 not considered personal.
7742 64. Added rcpt_include_affixes as a generic transport option.
7744 65. Added queue_only_override (default true).
7746 66. Added the syslog_duplication option.
7748 67. If what should have been the first header line of a message consisted of
7749 a space followed by a colon, Exim was mis-interpreting it as a header line.
7750 It isn't of course - it is syntactically invalid and should therefore be
7751 treated as the start of the message body. The misbehaviour could have
7752 caused a number of strange effects, including loss of data in subsequent
7753 header lines, and spool format errors.
7755 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7756 client host had authenticated. This control can now be exercised by an ACL
7757 for more flexibility.
7759 69. By default, callouts do not happen when testing with -bh. There is now a
7760 variant, -bhc, which does actually run the callout code, including
7761 consulting and updating the callout cache.
7763 70. Added support for saslauthd authentication, courtesy of Alexander
7766 71. If statvfs() failed on the spool or log directories while checking their
7767 size for availability, Exim confusingly gave the error "space shortage".
7768 Furthermore, in debugging mode it crashed with a floating point exception.
7769 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7770 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7771 serious problem, Exim now writes to the main and panic logs when this
7772 happens, with details of the failure. It then refuses to accept the
7773 incoming message, giving the message "spool directory problem" or "log
7774 directory problem" with a 421 code for SMTP messages.
7776 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7777 0, 1, and 2 exist, because some OS complain for execs without them (see
7778 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7779 descriptors. However, the code omitted to check that the open succeeded,
7780 causing mysterious errors if for some reason the permissions on /dev/null
7781 got screwed. Now Exim writes a message to the main and panic logs, and
7782 bombs out if it can't open /dev/null.
7784 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7785 interact so that it is all more flexible. It is supposed to remain
7786 backwards compatible. Also added extra_local_interfaces.
7788 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7789 to bomb out with an assertion failure - to the client this appears as a
7790 connection drop. This problem occurs in the part of the code that was taken
7791 from the Samba project. Fortunately, the assertion is in a very simple
7792 function, so I have fixed this by reproducing the function inline in the
7793 one place where it is called, and arranging for authentication to fail
7794 instead of killing the process with assert().
7796 75. The SPA client code was not working when the server requested OEM rather
7797 than Unicode encoding.
7799 76. Added code to make require_files with a specific uid setting more usable in
7800 the case where statting the file as root fails - usually a non-root-mounted
7801 NFS file system. When this happens and the failure is EACCES, Exim now
7802 forks a subprocess and does the per-uid checking as the relevant uid.
7804 77. Added process_log_path.
7806 78. If log_file_path was not explicitly set, a setting of check_log_space or
7807 check_log_inodes was ignored.
7809 79. If a space check for the spool or log partitions fails, the incident is now
7810 logged. Of course, in the latter case the data may get lost...
7812 80. Added the %p formatting code to string_format() so that it can be used to
7813 print addresses in debug_print(). Adjusted all the address printing in the
7814 debugging in store.c to use %p rather than %d.
7816 81. There was a concern that a line of code in smtp_in.c could overflow a
7817 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7818 initially expressed, the concern was not well-founded, because trailing
7819 spaces are removed early. However, if the trailing spaces were followed by
7820 a NULL, they did not get removed, so the overflow was possible. Two fixes
7823 (a) I re-wrote the offending code in a cleaner fashion.
7824 (b) If an incoming SMTP command contains a NULL character, it is rejected
7827 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7828 runs initgroups(), so that if the Exim user is in any additional groups,
7829 they will be used during message reception.
7835 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob