1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
57 JH/01 Avoid costly startup code when not strictly needed. This reduces time
58 for some exim process initialisations. It does mean that the logging
59 of TLS configuration problems is only done for the daemon startup.
61 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
63 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
64 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
67 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
68 DKIM signing is being done. Previously a transport_filter would always
69 disable CHUNKING, falling back to traditional DATA.
71 JH/05 Regard command-line receipients as tainted.
73 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
75 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
76 PAM library frees one of the arguments given to it, despite the
77 documentation. Therefore a plain malloc must be used.
79 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
80 on-stack buffers were used, resulting in a taint trap when DSN information
81 copied from a received message was written into the buffer.
83 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
84 the ordering of its ARC headers. This caused a crash.
86 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
87 a new record was being constructed with information from the peer, a trap
90 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
91 installation would get error messages from DMARC verify, when it hit the
92 nonexistent file indicated by the default. Distros wanting DMARC enabled
93 should both provide the file and set the option.
94 Also enforce no DMARC verification for command-line sourced messages.
96 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
97 could, depending on the platform, hang at the STARTTLS response.
99 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
100 message on a connection. Previously if one message had ARC headers and
101 the following one did not, a crash could result when adding an
102 Authentication-Results: header.
104 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
105 Exim main code and Exim-related utities. The introduction of taint
106 tracking also did many adjustments to string handling. Since then, eximon
107 frequently terminated with an assert failure.
109 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
110 check for 452 responses. This slightly helps the inefficieny of doing
111 a large alias-expansion into a recipient-limited target. The max_rcpt
112 transport option still applies (and at the current default, will override
113 the new feature). The check is done for either cause of synch, and forces
114 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
115 connection. The new facility is not tunable at this time.
117 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
118 library live data was being used, so the results became garbage. Make
119 copies while it is still usable.
121 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
122 on delivery deferred (==) and failed (**) lines (if a delivery was
123 attemtped). Previously it was only on completion (=>) lines.
125 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
126 for the expansion of the server_scram_iter and server_scram_salt options.
128 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
129 are now specifically given a NO_DATA response without hitting the system
130 resolver. The library goes on to do the now-standard TXT lookup.
131 Use of dnsdb lookups is not affected.
133 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
134 only retrieve the errormessage once. Previously two calls to dlerror()
135 were used, and the second one (for mainlog/paniclog) retrieved null
138 JH/20 Taint checking: disallow use of tainted data for
139 - the appendfile transport file and directory options
140 - the pipe transport command
141 - the autoreply transport file, log and once options
142 - file names used by the redirect router (including filter files)
144 - paths used by single-key lookups
145 Previously this was permitted.
147 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
148 adjusted the size of a major service buffer; this failed because the
149 buffer was in use at the time. Change to a compile-time increase in the
150 buffer size, when this authenticator is compiled into exim.
152 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
153 previous fast-mode was untenable in the face of glibs using mmap to
154 support larger malloc requests.
156 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
157 New values supported, if defined on system where compiled:
158 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
159 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
161 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
162 running a limited number of proceses in parallel, a benefit is gained. The
163 amount varies with the platform hardware and load. The use of the option
164 queue_run_in_order means we cannot do this, as ordering becomes
167 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
168 had introduced a string-copy (for ensuring NUL-termination) which was not
169 appropriate for that case, which can include embedded NUL bytes in the
170 block of data. Investigation showed the copy to actually be needless, the
171 data being length-specified.
173 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
174 done during a receiving connection, and both used TLS, global info was
175 used rather than per-connection info for tracking the state of data
176 queued for transmission. This could result in a connection hang.
178 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
179 Previously, when delivering serveral messages down a single connection
180 only the first would provide a SIZE. This was due to the size information
181 not being properly tracked.
183 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
184 TAI (at 37 seconds currently), pretend to be in UTC for time-related
185 expansion and logging. Previously, spurious values such as a future
186 minute could be seen.
188 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
189 it could crash from a null-deref. This could also affect the
190 ${addresses: } operator and ${readsock } item.
192 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
193 message following a mime one, the variable was not reset.
195 JH/30 When an pipelined-connect fails at the first response, assume incorrect
196 cached capability (perhaps the peer reneged?) and immediately retry in
199 JH/31 Fix spurious detection of timeout while writing to transport filter.
201 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
202 an attempt to copy the string was made before checking it.
204 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
205 taint of the lookup key was maintained; we now regard the presence in the
206 filesystem as sufficient validation.
208 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
209 argument is supplied.
211 JH/35 The dsearch lookup now requires that the directory is an absolute path.
212 Previously this was not checked, and nonempty relative paths made an
213 access under Exim's current working directory.
215 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
216 Previously no event was raised.
218 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
219 parameter supplied by the sender MAIL FROM command. Previously it was
220 ignored, and only the check_spool_space option value for the required
223 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
224 the size of the signing public-key. Previously it was instead giving
225 the size of the signature hash.
227 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
228 the default. See the (new) dkim_verify_min_keysizes option.
230 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
231 and an ACL use a lookup for checking either the local_part or domain,
232 stale data could be accessed. Ensure that variable references are
233 dropped between messages.
235 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
236 by the client was not checked as pointing within response data before
237 being used. A malicious client could thus cause an out-of-bounds read and
238 possibly gain authentication. Fix by adding the check.
240 JH/42 Internationalisation: change the default for downconversion in the smtp
241 transport to be "if needed". Previously it was "as previously set" for
242 the message, which usually meant "if needed" for message-submission but
243 "no" for everything else. However, MTAs have been seen using SMTPUTF8
244 even when the envelope addresses did not need it, resulting in forwarding
245 failures to non-supporting MTAs. A downconvert in such cases will be
246 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
247 transport. The change does mean that addresses needing conversion will
248 be converted when previously a delivery failure would occur.
250 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
251 response was received it would be used unchecked in a fail-DSN, violating
252 standards on line-length limits. Truncate if needed.
254 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
255 doesn't work. (Additionally add a new main config option to configure the
262 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
263 side, to decode a TLS 1.3 packet capture.
265 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
266 Previously the default library behaviour applied, sending two, each in
269 JH/03 Debug output for ACL now gives the config file name and line number for
272 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
274 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
276 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
277 buffer overrun for (non-chunking) other transports.
279 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
280 TLS1.3, means that a server rejecting a client certificate is not visible
281 to the client until the first read of encrypted data (typically the
282 response to EHLO). Add detection for that case and treat it as a failed
283 TLS connection attempt, so that the normal retry-in-clear can work (if
284 suitably configured).
286 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
287 and/or domain. Found and fixed by Jason Betts.
289 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
290 configuration). If a CNAME target was not a wellformed name pattern, a
293 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
294 the OS reports them interleaved with other addresses.
296 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
297 used both for input and for a verify callout, both encrypted, SMTP
298 responses being sent by the server could be lost. This resulted in
299 dropped connections and sometimes bounces generated by a peer sending
302 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
303 string. Previously it was possible to cause undefined behaviour in a
304 library routine (usually a crash). Found by "zerons".
306 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
309 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
310 API was removed, so update to use the newer ones.
312 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
313 any timeout set, is taking a long time. Previously we would hang on to a
314 rotated logfile "forever" if the input was arriving with long gaps
315 (a previous attempt to fix addressed lack, for a long time, of initial
318 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
319 shared (NFS) environment. The length of the tempfile name is now
320 4 + 16 ("hdr.$message_exim_id") which might break on file
321 systems which restrict the file name length to lower values.
324 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
325 shared (NFS) environment.
327 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
328 did for all versions <4.90). Notably -M, -m, --invert, -I may be
331 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
332 on some platforms for bit 31.
334 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
335 to changes apparently associated with TLS1.3 handling some of the APIs
336 previously used were either nonfunctional or inappropriate. Strings
337 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
338 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
339 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
340 This affects log line X= elements, the $tls_{in,out}_cipher variables,
341 and the use of specific cipher names in the encrypted= ACL condition.
343 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
345 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
346 verification result was not updated unless hosts_require_ocsp applied.
348 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
349 queue_list_requires_admin set to false, non-admin users were denied the
352 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
353 directory-of-certs mode. Previously they were advertised despite the
356 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
357 A single TCP connection by a client will now hold a TLS connection open
358 for multiple message deliveries, by default. Previously the default was to
361 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
362 default. If built with the facility, DANE will be used. The facility
363 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
365 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
366 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
367 must be defined and you must still, unless you define DISABLE_TLS, manage
368 the the include-dir and library-file requirements that go with that
369 choice. Non-TLS builds are still supported.
371 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
374 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
375 default. If the platform supports and has the facility enabled, it will
376 be requested on all coneections.
378 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
379 controlled by the build-time option SUPPORT_PIPE_CONNECT.
381 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
383 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
384 success-DSN messages. Previously the From: header was always the default
385 one for these; the option was ignored.
387 JH/28 Fix the timeout on smtp response to apply to the whole response.
388 Previously it was reset for every read, so a teergrubing peer sending
389 single bytes within the time limit could extend the connection for a
390 long time. Credit to Qualsys Security Advisory Team for the discovery.
392 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
393 delivery address, which leaked information of the results of local
394 forwarding. Change to the original envelope recipient address, per
397 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
398 requested. Previously not bounce was generated and a log entry of
399 error ignored was made.
401 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
403 JH/32 Introduce a general tainting mechanism for values read from the input
404 channel, and values derived from them. Refuse to expand any tainted
405 values, to catch one form of exploit.
407 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
408 was unused and the unexpanded text used for the test. Found and
409 fixed by Ruben Jenster.
411 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
412 an attempt to use a TLS library read routine dereffed a nul pointer,
415 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
416 them in our smtp response.
418 JH/36 Have the general router option retry_use_local_part default to true when
419 any of the restrictive preconditions are set (to anything). Previously it
420 was only for check_local user. The change removes one item of manual
421 configuration which is required for proper retries when a remote router
422 handles a subset of addresses for a domain.
424 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
425 link count into consideration.
427 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
428 caused the extension of big_buffer, the following lines were ignored.
430 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
431 accordance with RFC 2308. Previously there was no expiry, so a longlived
432 receive process (eg. due to ACL delays) versus a short SOA value could
435 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
437 JH/39 Promote DMARC support to mainline.
439 JH/40 Bug 2452: Add a References: header to DSNs.
441 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
442 parameters. The relevant library call is documented as "Deprecated: This
443 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
444 3.6.0, DH parameters are negotiated following RFC7919."
446 HS/06 Change the default of dnssec_request_domains to "*"
448 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
449 carried on and emitted a BDAT command, even when PIPELINING was not
452 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
453 buffer was used for the filename, resulting in a trap when tainted
454 arguments (eg. $domain) were used.
456 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
457 recommended to avoid a possible server-load attack. The feature can be
458 re-enabled via the openssl_options main cofiguration option.
460 JH/45 local_scan API: documented the current smtp_printf() call. This changed
461 for version 4.90 - adding a "more data" boolean to the arguments.
462 Bumped the ABI version number also, this having been missed previously;
463 release versions 4.90 to 4.92.3 inclusive were effectively broken in
464 respect of usage of smtp_printf() by either local_scan code or libraries
465 accessed via the ${dlfunc } expansion item. Both will need coding
466 adjustment for any calls to smtp_printf() to match the new function
467 signature; a FALSE value for the new argument is always safe.
469 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
470 the file-offset (which the Linux syscall does, and exim expects); this
471 resulted in an indefinite loop.
473 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
474 to do ARC verification. The Authentication-Results: header line added
475 by the configuration then had no ARC item.
481 JH/01 Remove code calling the customisable local_scan function, unless a new
482 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
484 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
485 non-signal-safe functions being used.
487 JH/03 Bug 2269: When presented with a received message having a stupidly large
488 number of DKIM-Signature headers, disable DKIM verification to avoid
489 a resource-consumption attack. The limit is set at twenty.
491 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
492 report of oldest_pass in ${authres } in consequence, and separate out
493 some descriptions of reasons for verification fail.
495 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
496 files in the spool were present and unlocked. A queue-runner could spot
497 them, resulting in a duplicate delivery. Fix that by doing the unlock
498 after the unlink. Investigation by Tim Stewart. Take the opportunity to
499 add more error-checking on spoolfile handling while that code is being
502 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
503 No known attacks, no CVE, this is defensive hardening.
505 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
506 a queue-runner could start a delivery while other operations were ongoing.
507 Cutthrough delivery was a common victim, resulting in duplicate delivery.
508 Found and investigated by Tim Stewart. Fix by using the open message data
509 file handle rather than opening another, and not locally closing it (which
510 releases a lock) for that case, while creating the temporary .eml format
511 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
513 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
514 $sender_verify_failure/$recipient_verify_failure to "random".
516 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
519 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
520 Previously this would segfault.
522 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
525 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
526 like zero, since the resolver should be doing this for us, But we need one
527 as a CNAME but no MX presence gets the CNAME returned; we need to check
528 that doesn't point to an MX to declare it "no MX returned" rather than
529 "error, loop". A new main option is added so the older capability of
530 following some limited number of chain links is maintained.
532 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
534 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
535 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
536 now-deprecated earlier definitions used only specified the range up to TLS
537 1.2 (in the older-version library docs).
539 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
541 JH/15 Rework TLS client-side context management. Stop using a global, and
542 explicitly pass a context around. This enables future use of TLS for
543 connections to service-daemons (eg. malware scanning) while a client smtp
544 connection is using TLS; with cutthrough connections this is quite likely.
546 JH/16 Fix ARC verification to do AS checks in reverse order.
548 JH/17 Support a "tls" option on the ${readsocket } expansion item.
550 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
551 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
552 Previously the "utf8" would be re-prepended for every additional message.
554 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
555 Previously thery were accepted, resulting in issues when attempting to
556 forward messages to a non-supporting MTA.
558 PP/02 Let -n work with printing macros too, not just options.
560 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
561 one parent address was copied, and bogus data was used at delivery-logging
562 time. Either a crash (after delivery) or bogus log data could result.
563 Discovery and analysis by Tim Stewart.
565 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
566 Previously if the string ended mid-character, we did not insert the
567 promised '?' replacement.
569 PP/04 Documentation: current string operators work on bytes, not codepoints.
571 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
572 should pack well giving a smaller memory footprint so better caching and
573 therefore performance. Group the declarations where this can't be done so
574 that the byte-sized flag variables are not interspersed among pointer
575 variables, giving a better chance of good packing by the compiler.
577 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
578 non-null, to avoid issues with sites running BATV. Previously reports were
579 sent with an empty envelope sender so looked like bounces.
581 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
582 The ignore_error flag wasn't being returned from the filter subprocess so
583 was not set for later routers. Investigation and fix by Matthias Kurz.
585 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
586 and a msg:complete for the whole, when a message is manually removed using
587 -Mrm. Developement by Matthias Kurz, hacked on by JH.
589 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
590 a "Gnu special" function, asprintf() in the DB utility binary builds; I
591 hope that is portable enough.
593 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
594 requiring a known-CA anchor certificate; make it now rely entirely on the
595 TLSA as an anchor. Checking the name on the leaf cert against the name
596 on the A-record for the host is still done for TA (but not for EE mode).
598 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
599 included in delivery lines for non-proxied connections, when compiled with
600 SUPPORT_SOCKS and running with proxy logging enabled.
602 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
603 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
604 move the existing event to fire before the normal logging of message
605 failure so that custom logging is bracketed by normal logging.
607 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
608 msg:fail:internal event. Developement by Matthias Kurz.
610 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
611 far too small for todays use of crypto signatures stored there. Go all
612 the way to the max DNS message size of 64kB, even though this might be
613 overmuch for IOT constrained device use.
615 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
616 copy a string over itself. The library routine is documented as not
617 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
619 JH/32 For main options check_spool_space and check_inode_space, where the
620 platform supports 64b integers, support more than the previous 2^31 kB
621 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
622 the previous G, M, k.
624 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
625 $authenticated_fail_id variable on authentication failure. Previously
628 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
629 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
630 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
631 GNUTLS_SEC_PARAM_MEDIUM.
633 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
634 side. Previously we would continue as if no SNI had been received.
636 JH/36 Harden the handling of string-lists. When a list consisted of a sole
637 "<" character, which should be a list-separator specification, we walked
638 off past the nul-terimation.
640 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
641 causes) even when the retry time is not yet met. Previously they were
642 not, meaning that when (say) an account was over-quota and temp-rejecting,
643 and multiple senders' messages were queued, only one sender would get
644 notified on each configured delay_warning cycle.
646 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
648 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
649 cipher-suites, an error can be left on the stack even for a succeeding
650 accept; this results in impossible error messages when a later operation
653 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
654 return error codes indicating retry. Under TLS1.3 this becomes required.
656 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
657 it only wrote the new authenticators, resulting in a lack of tracking of
658 peer changes of ESMTP extensions until the next cache flush.
660 JH/41 Fix the loop reading a message header line to check for integer overflow,
661 and more-often against header_maxsize. Previously a crafted message could
662 induce a crash of the recive process; now the message is cleanly rejected.
664 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
665 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
671 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
672 When redis_servers is set to a list of > 1 element, and the Redis servers
673 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
674 case of MOVED into a DEFER case instead, thus moving the query onto the
675 next server in the list. For a cluster of N elements, all N servers must
676 be defined in redis_servers.
678 GF/02 Catch and remove uninitialized value warning in exiqsumm
679 Check for existence of @ARGV before looking at $ARGV[0]
681 JH/01 Replace the store_release() internal interface with store_newblock(),
682 which internalises the check required to safely use the old one, plus
683 the allocate and data copy operations duplicated in both (!) of the
684 extant use locations.
686 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
687 modifier. This matches the restriction on the commandline.
689 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
690 Previously only the last row was returned.
692 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
693 we assumed that tags in the header were well-formed, and parsed the
694 element content after inspecting only the first char of the tag.
695 Assumptions at that stage could crash the receive process on malformed
698 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
699 While running the DKIM ACL we operate on the Permanent memory pool so that
700 variables created with "set" persist to the DATA ACL. Also (at any time)
701 DNS lookups that fail create cache records using the Permanent pool. But
702 expansions release any allocations made on the current pool - so a dnsdb
703 lookup expansion done in the DKIM ACL releases the memory used for the
704 DNS negative-cache, and bad things result. Solution is to switch to the
705 Main pool for expansions.
706 While we're in that code, add checks on the DNS cache during store_reset,
707 active in the testsuite.
708 Problem spotted, and debugging aided, by Wolfgang Breyha.
710 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
711 When none of the hosts presented to a transport match an already-open
712 connection, close it and proceed with the list. Previously we would
713 queue the message. Spotted by Lena with Yahoo, probably involving
716 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
717 Previously a spurious "250 OK id=" response was appended to the proper
720 JH/08 The "support for" informational output now, which built with Content
721 Scanning support, has a line for the malware scanner interfaces compiled
722 in. Interface can be individually included or not at build time.
724 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
725 by the template makefile "src/EDITME". The "STREAM" support for an older
726 ClamAV interface method is removed.
728 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
729 rows affected is given instead).
731 JH/11 The runtime Berkeley DB library version is now additionally output by
732 "exim -d -bV". Previously only the compile-time version was shown.
734 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
735 SMTP connection. Previously, when one had more receipients than the
736 first, an abortive onward connection was made. Move to full support for
737 multiple onward connections in sequence, handling cutthrough connection
738 for all multi-message initiating connections.
740 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
741 routers. Previously, a multi-recipient message would fail to match the
742 onward-connection opened for the first recipient, and cause its closure.
744 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
745 a timeout on read on a GnuTLS initiating connection, resulting in the
746 initiating connection being dropped. This mattered most when the callout
747 was marked defer_ok. Fix to keep the two timeout-detection methods
750 JH/15 Relax results from ACL control request to enable cutthrough, in
751 unsupported situations, from error to silently (except under debug)
752 ignoring. This covers use with PRDR, frozen messages, queue-only and
755 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
757 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
758 metadata, resulting in a crash in free().
760 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
761 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
762 Broken also in d185889f4, with init system revamp.
764 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
765 Previously we abruptly closed the connection after reading a malware-
766 found indication; now we go on to read the "scan ok" response line,
769 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
770 ACL. Previously, a crash would result.
772 JH/19 Speed up macro lookups during configuration file read, by skipping non-
773 macro text after a replacement (previously it was only once per line) and
774 by skipping builtin macros when searching for an uppercase lead character.
776 JH/20 DANE support moved from Experimental to mainline. The Makefile control
777 for the build is renamed.
779 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
780 was allocated for every new TLS startup, meaning one per message. Fix
781 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
783 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
784 reported the original. Fix to report (as far as possible) the ACL
785 result replacing the original.
787 JH/23 Fix memory leak during multi-message connections using STARTTLS under
788 OpenSSL. Certificate information is loaded for every new TLS startup,
789 and the resources needed to be freed.
791 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
793 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
796 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
797 DATA response info to the (existing) per-recipient response info for
798 the "C=" log element. It can have useful tracking info from the
799 destination system. Patch from Simon Arlott.
801 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
802 length value. Previously this would segfault.
804 HS/02 Support Avast multiline protoocol, this allows passing flags to
805 newer versions of the scanner.
807 JH/28 Ensure that variables possibly set during message acceptance are marked
808 dead before release of memory in the daemon loop. This stops complaints
809 about them when the debug_store option is enabled. Discovered specifically
810 for sender_rate_period, but applies to a whole set of variables.
811 Do the same for the queue-runner and queue-list loops, for variables set
812 from spool message files. Do the same for the SMTP per-message loop, for
813 certain variables indirectly set in ACL operations.
815 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
816 as a multi-recipient message from a mailinglist manager). The coding had
817 an arbitrary cutoff number of characters while checking for more input;
818 enforced by writing a NUL into the buffer. This corrupted long / fast
819 input. The problem was exposed more widely when more pipelineing of SMTP
820 responses was introduced, and one Exim system was feeding another.
821 The symptom is log complaints of SMTP syntax error (NUL chars) on the
822 receiving system, and refused recipients seen by the sending system
823 (propating to people being dropped from mailing lists).
824 Discovered and pinpointed by David Carter.
826 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
827 replaced by the ${authresults } expansion.
829 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
831 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
832 allows proper process termination in container environments.
834 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
835 Previously the "final dot" had a newline after it; ensure it is CR,LF.
837 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
838 and "err_perm", deprecated since 4.83 when the RFC-defined words
839 "temperror" and "permerror" were introduced.
841 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
842 transport-filters or DKIM-signing. The restriction was lost in the
843 consolidation of verify-callout and delivery SMTP handling.
844 Extend the restriction to also cover ARC-signing.
846 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
847 in defer=pass mode supply a 450 to the initiator. Previously the message
850 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
851 tls_require_ciphers is used as before.
853 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
854 "pass_unscanned". Only tmpfails from the scanner are written to
855 the paniclog, as they may require admin intervention (permission
856 denied, license issues). Other scanner errors (like decompression
857 bombs) do not cause a paniclog entry.
859 JH/36 Fix reinitialisation of DKIM logging variable between messages.
860 Previously it was possible to log spurious information in receive log
863 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
864 triggered odd behaviour from Outlook Express clients.
866 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
869 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
870 since the IETF WG has not yet settled on that versus the original
871 "bare" representation.
873 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
874 Previously the millisecond value corrupted the output.
875 Fix also for syslog_pid=no and log_selector +pid, for which the pid
876 corrupted the output.
882 JH/01 Rework error string handling in TLS interface so that the caller in
883 more cases is responsible for logging. This permits library-sourced
884 string to be attached to addresses during delivery, and collapses
885 pairs of long lines into single ones.
887 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
888 during configuration. Wildcards are allowed and expanded.
890 JH/02 Rework error string handling in DKIM to pass more info back to callers.
891 This permits better logging.
893 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
894 do not close it down and have the child transport start it up again on
895 the passed-on TCP connection. Instead, proxy the child (and any
896 subsequent ones) for TLS via a unix-domain socket channel. Logging is
897 affected: the continued delivery log lines do not have any DNSSEC, TLS
898 Certificate or OCSP information. TLS cipher information is still logged.
900 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
901 identical IP addresses on different listening ports. Will also affect
904 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
905 add noisy ifdef guards to special-case this sillyness.
906 Patch from Bernd Kuhls.
908 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
909 than 255 are no longer allowed.
911 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
912 Disable the session-cache too, which might reduce our load. Since we
913 currrectly use a new context for every connection, both as server and
914 client, there is no benefit for these.
915 GnuTLS appears to not support tickets server-side by default (we don't
916 call gnutls_session_ticket_enable_server()) but client side is enabled
917 by default on recent versions (3.1.3 +) unless the PFS priority string
920 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
921 <https://reproducible-builds.org/specs/source-date-epoch/>.
923 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
924 the check for any unsuccessful recipients did not notice the limit, and
925 erroneously found still-pending ones.
927 JH/08 Pipeline CHUNKING command and data together, on kernels that support
928 MSG_MORE. Only in-clear (not on TLS connections).
930 JH/09 Avoid using a temporary file during transport using dkim. Unless a
931 transport-filter is involved we can buffer the headers in memory for
932 creating the signature, and read the spool data file once for the
933 signature and again for transmission.
935 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
936 4.77 as the kernel support then wasn't solid, having issues in 64bit
937 mode. Now, it's been long enough. Add support for FreeBSD also.
939 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
940 case where the routing stage had gathered several addresses to send to
941 a host before calling the transport for the first, we previously failed
942 to close down TLS in the old transport process before passing the TCP
943 connection to the new process. The new one sent a STARTTLS command
944 which naturally failed, giving a failed delivery and bloating the retry
945 database. Investigation and fix prototype from Wolfgang Breyha.
947 JH/12 Fix check on SMTP command input synchronisation. Previously there were
948 false-negatives in the check that the sender had not preempted a response
949 or prompt from Exim (running as a server), due to that code's lack of
950 awareness of the SMTP input buffering.
952 PP/04 Add commandline_checks_require_admin option.
953 Exim drops privileges sanely, various checks such as -be aren't a
954 security problem, as long as you trust local users with access to their
955 own account. When invoked by services which pass untrusted data to
956 Exim, this might be an issue. Set this option in main configuration
957 AND make fixes to the calling application, such as using `--` to stop
960 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
961 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
962 responses to those, into a single TLS record each way (this usually means
963 a single packet). As a side issue, smtp_enforce_sync now works on TLS
966 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
967 affects you only if you're dancing at the edge of the param size limits.
968 If you are, and this message makes sense to you, then: raise the
969 configured limit or use OpenSSL 1.1. Nothing we can do for older
972 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
973 cmdline element to get the documented default one. Previously it was
976 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
977 are unsuable for non-message-specific reasons - eg. connection timeout,
978 banner-time rejection.
980 JH/16 Fix logging of delivery remote port, when specified by router, under
983 PP/06 Repair manualroute's ability to take options in any order, even if one
984 is the name of a transport.
987 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
989 JH/17 Change the list-building routines interface to use the expanding-string
990 triplet model, for better allocation and copying behaviour.
992 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
993 Previously it was constructed the first time a possibly-matching string
994 was met in the configuration file input during startup; now it is done
997 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
998 compatible one, to avoid the (poorly documented) possibility of a config
999 file in the working directory redirecting the DB files, possibly correpting
1000 some existing file. CVE-2017-10140 assigned for BDB.
1002 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1003 cache-hot. Previously, although the result was properly cached, the
1004 initial verify call returned a defer.
1006 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1007 the main verify for receipient in uncached-mode.
1009 JH/22 Retire historical build files to an "unsupported" subdir. These are
1010 defined as "ones for which we have no current evidence of testing".
1012 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1013 if present. Previously it was ignored.
1015 JH/24 Start using specified-initialisers in C structure init coding. This is
1016 a C99 feature (it's 2017, so now considered safe).
1018 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1019 if was a fixed-sized field and bitmask ops via macros; it is now more
1022 PP/07 GitHub PR 56: Apply MariaDB build fix.
1023 Patch provided by Jaroslav Å karvada.
1025 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1026 during Coverity cleanups [4.87 JH/47]
1027 Diagnosis and fix provided by Michael Fischer v. Mollard.
1029 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1030 the right size to place the terminating semicolon on its own folded
1031 line, the header hash was calculated to an incorrect value thanks to
1032 the (relaxed) space the fold became.
1034 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1035 and confused the parent.
1037 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1038 which could crash as a result. This could lead to undeliverable messages.
1040 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1043 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1044 matters on fast-turnover and PID-randomising systems, which were getting
1045 out-of-order delivery.
1047 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1048 a possibly-overlapping copy. The symptom was that "Remote host closed
1049 connection in response to HELO" was logged instead of the actual 4xx
1052 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1053 Previously only that bufferd was discarded, resulting in SYMTP command
1056 JH/32 DKIM: when a message has multiple signatures matching an identity given
1057 in dkim_verify_signers, run the dkim acl once for each. Previously only
1058 one run was done. Bug 2189.
1060 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1061 "panic the current process" to "deliberately defer". The panic log is
1062 still written with the problem list name; the mail and reject logs now
1063 get a temp-reject line for the message that was being handled, saying
1064 something like "domains check lookup or other defer". The SMTP 451
1065 message is still "Temporary local problem".
1067 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1068 A crafted sequence of BDAT commands could result in in-use memory beeing
1069 freed. CVE-2017-16943.
1071 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1072 from SMTP input. Previously it was always done; now only done for DATA
1073 and not BDAT commands. CVE-2017-16944.
1075 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1076 to the message (such as an overlong header line). Previously this was
1077 not done and we did not exit BDAT mode. Followon from the previous item
1078 though a different problem.
1084 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1085 than -2003 did; needs libidn2 in addition to libidn.
1087 JH/02 The path option on a pipe transport is now expanded before use.
1089 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1090 Patch provided by "Björn", documentation fix added too.
1092 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1093 missing a wire-to-host endian conversion.
1095 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1096 close after a BDAT command line could be taken as a following command,
1097 giving a synch failure. Fix by only checking for synch immediately
1098 before acknowledging the chunk.
1100 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1101 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1102 macro. Patches provided by Josh Soref.
1104 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1105 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1106 by IANA as a service extension. However, John Klensin suggests that we
1109 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1110 the dkim code may be unix-mode line endings rather than smtp wire-format
1111 CRLF, so prepend a CR to any bare LF.
1113 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1114 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1116 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1117 the first were themselves being wrongly included in the feed into dkim
1118 processing; with most chunk sizes in use this resulted in an incorrect
1119 body hash calculated value.
1121 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1122 DKIM signature block, for verification. Although advised against by
1123 standards it is specifically not ruled illegal.
1125 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1127 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1128 missing a body hash (the bh= tag).
1130 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1131 It seems that HAProxy sends the Proxy Protocol information in clear and
1132 only then does a TLS startup, so do the same.
1134 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1135 TCP connections (such as for Spamd) unless the daemon successfully set
1136 Fast Open mode on its listening sockets. This fixes breakage seen on
1137 too-old kernels or those not configured for Fast Open, at the cost of
1138 requiring both directions being enabled for TFO, and TFO never being used
1139 by non-daemon-related Exim processes.
1141 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1142 endings, at least on the first header line. Try to canonify any that get
1143 past that check, despite the cost.
1145 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1146 now limited to an arbitrary five deep, while parsing addresses with the
1147 strip_excess_angle_brackets option enabled.
1149 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1150 instead leave the unprompted TLS handshake in socket buffer for the
1151 TLS library to consume.
1153 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1155 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1157 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1158 region in one operation, for speed, and this leaves assigned pointers
1159 dangling. Add checks run only under the testsuite which checks all
1160 variables at a store-reset and panics on a dangling pointer; add code
1161 explicitly nulling out all the variables discovered. Fixes one known
1162 bug: a transport crash, where a dangling pointer for $sending_ip_address
1163 originally assigned in a verify callout, is re-used.
1165 PP/06 Drop '.' from @INC in various Perl scripts.
1167 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1169 PP/08 Reduce a number of compilation warnings under clang; building with
1170 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1171 should be warning-free.
1173 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1175 HS/01 Fix portability problems introduced by PP/08 for platforms where
1176 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1178 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1179 chunk. This allows us to accept broken chunked messages. We need a more
1180 general solution here.
1182 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1183 already-broken messages in the queue.
1185 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1187 JH/19 Fix reference counting bug in routing-generated-address tracking.
1193 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1194 supports it and a size is available (ie. the sending peer gave us one).
1196 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1197 years of being deprecated). The replacements are the ACLs
1198 acl_smtp_mime and acl_not_smtp_mime.
1200 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1201 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1202 or even an in-clear connection were permitted. Now, if the host lookup
1203 was dnssec and dane was requested then the host is only used if the
1204 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1205 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1206 if one fails this test.
1207 This means that a poorly-configured remote DNS will make it incommunicado;
1208 but it protects against a DNS-interception attack on it.
1210 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1211 non-noisy when a race steals the message being considered.
1213 JH/05 If main configuration option tls_certificate is unset, generate a
1214 self-signed certificate for inbound TLS connections.
1216 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1217 in rewrites and routers.
1219 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1220 and logged a warning sing 4.83; now they are a configuration file error.
1222 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1223 (lacking @domain). Apply the same qualification processing as RCPT.
1225 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1227 JH/10 Support ${sha256:} applied to a string (as well as the previous
1230 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1231 a cutthrough deliver is pending, as we always want to make a connection.
1232 This also avoids re-routing the message when later placing the cutthrough
1233 connection after a verify cache hit.
1234 Do not update it with the verify result either.
1236 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1237 when routing results in more than one destination address.
1239 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1240 signing (which inhibits the cutthrough capability). Previously only
1241 the presence of an option was tested; now an expansion evaluating as
1242 empty is permissible (obviously it should depend only on data available
1243 when the cutthrough connection is made).
1245 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1246 the relevant preceding SMTP command did not note the pipelining mode.
1248 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1249 Previously they were not counted.
1251 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1252 as one having no matching records. Previously we deferred the message
1253 that needed the lookup.
1255 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1256 distinguished as "(=".
1258 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1259 for missing MX records. Previously it only worked for missing A records.
1261 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1263 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1264 after the data-go-ahead and data-ack. Patch from Jason Betts.
1266 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1267 even for a "none" policy. Patch from Tony Meyer.
1269 JH/22 Fix continued use of a connection for further deliveries. If a port was
1270 specified by a router, it must also match for the delivery to be
1273 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1274 When one of the recipients of a message was unsuitable for the connection
1275 (has no matching addresses), we lost track of needing to mark it
1276 deferred. As a result mail would be lost.
1278 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1280 JH/25 Decoding ACL controls is now done using a binary search; the source code
1281 takes up less space and should be simpler to maintain. Merge the ACL
1282 condition decode tables also, with similar effect.
1284 JH/26 Fix problem with one_time used on a redirect router which returned the
1285 parent address unchanged. A retry would see the parent address marked as
1286 delivered, so not attempt the (identical) child. As a result mail would
1289 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1290 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1291 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1294 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1295 Main config options check_{log,spool}_{inodes,space} are now
1296 100 inodes, 10MB unless set otherwise in the configuration.
1298 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1299 Previously it only applied to the main-section connection policy
1302 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1304 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1305 by me. Added RFC7919 DH primes as an alternative.
1307 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1308 are not in the system include path.
1310 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1311 GnuTLS, when a session startup failed (eg because the client disconnected)
1312 Exim did stdio operations after fclose. This was exposed by a recent
1313 change which nulled out the file handle after the fclose.
1315 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1316 signed directly by the cert-signing cert, rather than an intermediate
1317 OCSP-signing cert. This is the model used by LetsEncrypt.
1319 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1321 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1322 an incoming connection.
1324 HS/02 Bug 1802: Do not half-close the connection after sending a request
1327 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1328 fallback to "prime256v1".
1330 JH/34 SECURITY: Use proper copy of DATA command in error message.
1331 Could leak key material. Remotely exploitable. CVE-2016-9963.
1337 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1338 and 3.4.4 - once the server is enabled to respond to an OCSP request
1339 it does even when not requested, resulting in a stapling non-aware
1340 client dropping the TLS connection.
1342 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1343 support variable-length bit vectors. No functional change.
1345 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1346 The I= interface field on outgoing lines is now after the H= remote
1347 host field, same as incoming lines. There is a separate
1348 outgoing_interface log selector which allows you to disable the
1351 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1352 If not running log_selector +smtp_connection the mainlog would be held
1353 open indefinitely after a "too many connections" event, including to a
1354 deleted file after a log rotate. Leave the per net connection logging
1355 leaving it open for efficiency as that will be quickly detected by the
1356 check on the next write.
1358 HS/01 Bug 1671: Fix post transport crash.
1359 Processing the wait-<transport> messages could crash the delivery
1360 process if the message IDs didn't exist for some reason. When
1361 using 'split_spool_directory=yes' the construction of the spool
1362 file name failed already, exposing the same netto behaviour.
1364 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1365 mime_regex ACL conditions.
1367 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1368 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1369 to HELO, local diagnostic string.
1371 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1372 log line to debug. Even when configured with a tls authenticator many
1373 client connections are expected to not authenticate in this way, so
1374 an authenticate fail is not an error.
1376 HS/02 Add the Exim version string to the process info. This way exiwhat
1377 gives some more detail about the running daemon.
1379 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1380 matter for fast-change records such as DNSBLs.
1382 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1383 retry record, even if constant. There may be multiple transports with
1384 different interface settings and the retry behaviour needs to be kept
1387 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1389 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1391 JH/10 Bug 840: fix log_defer_output option of pipe transport
1393 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1394 hosts_randomize. This matters a lot when combined with mua_wrapper.
1396 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1397 ${quote_pgsql:<string>} operator.
1399 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1401 JH/14 Change header file name for freeradius-client. Relevant if compiling
1402 with Radius support; from the Gentoo tree and checked under Fedora.
1404 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1406 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1407 option transport_filter, meaning no filtering.
1409 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1411 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1412 defaults to "*" (all hosts). The variable is now available when not built
1413 with TLS, default unset, mainly to enable keeping the testsuite sane.
1414 If a server certificate is not supplied (via tls_certificate) an error is
1415 logged, and clients will find TLS connections fail on startup. Presumably
1416 they will retry in-clear.
1417 Packagers of Exim are strongly encouraged to create a server certificate
1418 at installation time.
1420 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1421 with the $config_file variable.
1423 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1424 in transport context, after the attempt, and per-recipient. The latter type
1425 is per host attempted. The event data is the error message, and the errno
1426 information encodes the lookup type (A vs. MX) used for the (first) host,
1427 and the trailing two digits of the smtp 4xx response.
1429 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1430 to write to mainlog (or rejectlog, paniclog) in the window between file
1431 creation and permissions/ownership being changed. Particularly affects
1432 installations where exicyclog is run as root, rather than exim user;
1433 result is that the running daemon panics and dies.
1435 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1437 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1438 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1439 "pri" and "weight". Note that the previous implicit priority given by the
1440 list order is no longer honoured.
1442 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1443 for DKIM processing.
1445 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1446 by defining SUPPORT_SOCKS.
1448 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1449 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1450 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1451 variables are renamed to proxy_{local,external}_{address,port}.
1453 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1454 for a build by defining SUPPORT_I18N
1456 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1457 of the query string, and make ${quote_redis:} do that quoting.
1459 JH/29 Move Events support from Experimental to mainline, enabled by default
1460 and removable for a build by defining DISABLE_EVENT.
1462 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1464 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1465 cached by the daemon.
1467 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1468 by defining LOOKUP_REDIS. The libhiredis library is required.
1470 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1471 keys are given for lookup.
1473 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1474 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1475 only supported when built with TLS support. The PolarSSL SHA routines
1476 are still used when the TLS library is too old for convenient support.
1478 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1479 openssl_options), for security. OpenSSL forces this from version 1.1.0
1480 server-side so match that on older versions.
1482 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1483 allocation for $value could be released as the expansion processing
1484 concluded, but leaving the global pointer active for it.
1486 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1487 and to use the domains and local_parts ACL conditions.
1489 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1490 incorrectly not doubled on cutthrough transmission, hence seen as a
1491 body-termination at the receiving system - resulting in truncated mails.
1492 Commonly the sender saw a TCP-level error, and retransmitted the message
1493 via the normal store-and-forward channel. This could result in duplicates
1494 received - but deduplicating mailstores were liable to retain only the
1495 initial truncated version.
1497 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1499 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1501 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1502 we're in there, support oversigning also; bug 1309.
1504 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1506 HS/04 Add support for keep_environment and add_environment options.
1508 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1509 either intentional arithmetic overflow during PRNG, or testing config-
1512 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1513 delivery resulted in actual delivery. Cancel cutthrough before DATA
1516 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1517 rejects a recipient: pass the reject to the originator.
1519 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1520 Many were false-positives and ignorable, but it's worth fixing the
1523 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1524 for the new environment-manipulation done at startup. Move the routines
1525 from being local to tls.c to being global via the os.c file.
1527 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1528 an extract embedded as result-arg for a map, the first arg for extract
1529 is unavailable so we cannot tell if this is a numbered or keyed
1530 extraction. Accept either.
1536 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1539 JH/02 The smtp transport option "multi_domain" is now expanded.
1541 JH/03 The smtp transport now requests PRDR by default, if the server offers
1544 JH/04 Certificate name checking on server certificates, when exim is a client,
1545 is now done by default. The transport option tls_verify_cert_hostnames
1546 can be used to disable this per-host. The build option
1547 EXPERIMENTAL_CERTNAMES is withdrawn.
1549 JH/05 The value of the tls_verify_certificates smtp transport and main options
1550 default to the word "system" to access the system default CA bundle.
1551 For GnuTLS, only version 3.0.20 or later.
1553 JH/06 Verification of the server certificate for a TLS connection is now tried
1554 (but not required) by default. The verification status is now logged by
1555 default, for both outbound TLS and client-certificate supplying inbound
1558 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1561 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1562 Status Notification (bounce) messages are now MIME format per RFC 3464.
1563 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1564 under the control of the dsn_advertise_hosts option, and routers may
1565 have a dsn_lasthop option.
1567 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1568 default, modifiable by a malware= option. The list separator for
1569 the options can now be changed in the usual way. Bug 68.
1571 JH/10 The smtp_receive_timeout main option is now expanded before use.
1573 JH/11 The incoming_interface log option now also enables logging of the
1574 local interface on delivery outgoing connections.
1576 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1577 if the interface and destination host and port all match.
1579 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1582 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1583 Patch from Andrew Lewis.
1585 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1586 now supports optional time-restrictions, weighting, and priority
1587 modifiers per server. Patch originally by <rommer@active.by>.
1589 JH/16 The spamd_address main option now supports a mixed list of local
1590 and remote servers. Remote servers can be IPv6 addresses, and
1591 specify a port-range.
1593 JH/17 Bug 68: The spamd_address main option now supports an optional
1594 timeout value per server.
1596 JH/18 Bug 1581: Router and transport options headers_add/remove can
1597 now have the list separator specified.
1599 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1602 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1605 JH/21 Support for the A6 type of dns record is withdrawn.
1607 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1608 rather than the verbs used.
1610 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1611 from 255 to 1024 chars.
1613 JH/24 Verification callouts now attempt to use TLS by default.
1615 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1616 are generic router options now. The defaults didn't change.
1618 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1619 Original patch from Alexander Shikoff, worked over by JH.
1621 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1622 files if ZCAT_COMMAND is not executable.
1624 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1626 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1628 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1629 Normally benign, it bites when the pair was led to by a CNAME;
1630 modern usage is to not canonicalize the domain to a CNAME target
1631 (and we were inconsistent anyway for A-only vs AAAA+A).
1633 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1635 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1636 when evaluating $sender_host_dnssec.
1638 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1639 $sender_helo_dnssec variable.
1641 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1643 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1645 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1647 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1648 documented as working, but never had. Support all but $spam_report.
1650 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1651 added for tls authenticator.
1653 HS/03 Add perl_taintmode main config option
1659 TL/01 When running the test suite, the README says that variables such as
1660 no_msglog_check are global and can be placed anywhere in a specific
1661 test's script, however it was observed that placement needed to be near
1662 the beginning for it to behave that way. Changed the runtest perl
1663 script to read through the entire script once to detect and set these
1664 variables, reset to the beginning of the script, and then run through
1665 the script parsing/test process like normal.
1667 TL/02 The BSD's have an arc4random API. One of the functions to induce
1668 adding randomness was arc4random_stir(), but it has been removed in
1669 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1670 function when detected.
1672 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1673 cause callback expansion.
1675 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1676 syntax errors in an expansion can be treated as a string instead of
1677 logging or causing an error, due to the internal use of bool_lax
1678 instead of bool when processing it.
1680 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1681 server certificates when making smtp deliveries.
1683 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1685 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1687 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1689 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1690 Merged patch from Sebastian Wiedenroth.
1692 JH/05 Fix results-pipe from transport process. Several recipients, combined
1693 with certificate use, exposed issues where response data items split
1694 over buffer boundaries were not parsed properly. This eventually
1695 resulted in duplicates being sent. This issue only became common enough
1696 to notice due to the introduction of connection certificate information,
1697 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1699 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1700 size buffer was used, resulting in syntax errors when an expansion
1703 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1704 version 3.3.6 or later.
1706 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1707 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1708 both become "event_action", the variables become $event_name, $event_data
1709 and $event_defer_errno. There is a new variable $verify_mode, usable in
1710 routers, transports and related events. The tls:cert event is now also
1711 raised for inbound connections, if the main configuration event_action
1714 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1715 early OCSP support, but no stapling (appears to be less than 1.0.0).
1717 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1718 server certificate names available under the smtp transport option
1719 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1722 JH/10 Time-related extraction expansions from certificates now use the main
1723 option "timezone" setting for output formatting, and are consistent
1724 between OpenSSL and GnuTLS compilations. Bug 1541.
1726 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1727 encoded parameter in the incoming message. Bug 1558.
1729 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1730 include certificate info, eximon was claiming there were spoolfile
1733 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1735 JH/14 Log delivery-related information more consistently, using the sequence
1736 "H=<name> [<ip>]" wherever possible.
1738 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1739 are problematic for Debian distribution, omit them from the release
1742 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1744 JH/16 Fix string representation of time values on 64bit time_t architectures.
1747 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1748 output list separator was used.
1753 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1754 checkers that were complaining about end of non-void function with no
1757 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1758 This was a regression introduced in 4.83 by another bugfix.
1760 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1762 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1763 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1769 TF/01 Correctly close the server side of TLS when forking for delivery.
1771 When a message was received over SMTP with TLS, Exim failed to clear up
1772 the incoming connection properly after forking off the child process to
1773 deliver the message. In some situations the subsequent outgoing
1774 delivery connection happened to have the same fd number as the incoming
1775 connection previously had. Exim would try to use TLS and fail, logging
1776 a "Bad file descriptor" error.
1778 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1779 utilities have not been installed.
1781 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1782 temporary space as the ACL may create new global variables.
1784 TL/01 LDAP support uses per connection or global context settings, depending
1785 upon the detected version of the libraries at build time.
1787 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1788 to extract and use the src ip:port in logging and expansions as if it
1789 were a direct connection from the outside internet. PPv2 support was
1790 updated based on HAProxy spec change in May 2014.
1792 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1794 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1795 Properly escape header and check for NULL return.
1797 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1800 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1802 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1803 characters in header names, implemented as a verify condition.
1804 Contributed by Michael Fischer v. Mollard.
1806 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1807 results permerror and temperror. Previous values are deprecated but
1808 still accepted. In a future release, err_perm and err_temp will be
1809 completely removed, which will be a backward incompatibility if the
1810 ACL tests for either of these two old results. Patch contributed by
1811 user bes-internal on the mailing list.
1813 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1815 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1816 selectors, in both main and reject logs.
1818 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1821 JH/07 Add malware type "sock" for talking to simple daemon.
1823 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1825 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1826 routers/transports under cutthrough routing.
1828 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1829 numbers. Touch up "bool" conditional to keep the same definition.
1831 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1833 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1835 JH/12 Expand items in router/transport headers_add or headers_remove lists
1836 individually rather than the list as a whole. Bug 1452.
1838 Required for reasonable handling of multiple headers_ options when
1839 they may be empty; requires that headers_remove items with embedded
1840 colons must have them doubled (or the list-separator changed).
1842 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1843 view the policy declared in the DMARC record. Currently, $dmarc_status
1844 is a combined value of both the record presence and the result of the
1847 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1849 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1850 dnslookup router and the smtp transport (applying to the forward
1853 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1854 of ldap servers used for a specific lookup. Patch provided by Heiko
1857 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1858 New variable $lookup_dnssec_authenticated for observability.
1860 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1861 Patch submitted by Lars Timman.
1863 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1865 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1866 Requires trusted mode and valid format message id, aborts otherwise.
1867 Patch contributed by Heiko Schlichting.
1869 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1870 certextract with support for various fields. Bug 1358.
1872 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1873 is requested by default, modifiable by smtp transport option
1876 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1877 operate on certificate variables to give certificate fingerprints
1878 Also new ${sha256:cert_variable}.
1880 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1882 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1883 Christian Aistleitner.
1885 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1887 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1888 file. Patch from Wolfgang Breyha.
1890 JH/25 Expand the coverage of the delivery $host and $host_address to
1891 client authenticators run in verify callout. Bug 1476.
1893 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1894 align with daemon_smtp_ports. Bug 72.
1896 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1897 support and error reporting did not work properly.
1899 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1900 and is readable. Patch from Andrew Colin Kissa.
1902 TL/14 Enhance documentation of ${run expansion and how it parses the
1903 commandline after expansion, particularly in the case when an
1904 unquoted variable expansion results in an empty value.
1906 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1908 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1909 by renaming away from it. Addresses GH issue 15, reported by
1912 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1914 TL/15 SECURITY: prevent double expansion in math comparison functions
1915 (can expand unsanitized data). Not remotely exploitable.
1922 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1924 PP/02 Make -n do something, by making it not do something.
1925 When combined with -bP, the name of an option is not output.
1927 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1930 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1931 $sender_host_name and config options to manage this, and basic check
1934 PP/05 DSCP support for outbound connections and control modifier for inbound.
1936 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1937 (Only plugin which currently uses this is kerberos4, which nobody should
1938 be using, but we should make it available and other future plugins might
1939 conceivably use it, even though it would break NAT; stuff *should* be
1940 using channel bindings instead).
1942 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1943 name; added for Sendmail compatibility; requires admin caller.
1944 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1945 just ignore it); requires trusted caller.
1946 Also parse but ignore: -Ac -Am -X<logfile>
1949 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1951 TL/02 Add +smtp_confirmation as a default logging option.
1953 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1954 Patch by Magnus Holmgren from 2007-02-20.
1956 TL/04 Bugzilla 1281 - Spec typo.
1957 Bugzilla 1283 - Spec typo.
1958 Bugzilla 1290 - Spec grammar fixes.
1960 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1962 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1964 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1965 mailing list by by Dmitry Isaikin.
1967 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1969 JH/02 Support "G" suffix to numbers in ${if comparisons.
1971 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1973 NM/01 Bugzilla 1197 - Spec typo
1974 Bugzilla 1196 - Spec examples corrections
1976 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1978 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1979 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1982 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1983 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1985 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1986 "acl {{name}{arg}...}", and optional args on acl condition
1989 JH/05 Permit multiple router/transport headers_add/remove lines.
1991 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1993 JH/07 Avoid using a waiting database for a single-message-only transport.
1994 Performance patch from Paul Fisher. Bugzilla 1262.
1996 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1999 JH/09 Add $headers_added variable, with content from use of ACL modifier
2000 add_header (but not yet added to the message). Bugzilla 199.
2002 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2003 Pulled from Bugzilla 817 by Wolfgang Breyha.
2005 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2007 (nb: this is the same fix as in Exim 4.80.1)
2009 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2012 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2015 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2017 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2018 advertise SMTP AUTH mechanism to us, instead of a generic
2019 protocol violation error. Also, make Exim more robust to bad
2020 data from the Dovecot auth socket.
2022 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2024 When a queue runner is handling a message, Exim first routes the
2025 recipient addresses, during which it prunes them based on the retry
2026 hints database. After that it attempts to deliver the message to
2027 any remaining recipients. It then updates the hints database using
2030 So if a recipient address works intermittently, it can get repeatedly
2031 deferred at routing time. The retry hints record remains fresh so the
2032 address never reaches the final cutoff time.
2034 This is a fairly common occurrence when a user is bumping up against
2035 their storage quota. Exim had some logic in its local delivery code
2036 to deal with this. However it did not apply to per-recipient defers
2037 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2039 This change adds a proper retry rule check during routing so that the
2040 final cutoff time is checked against the message's age. We only do
2041 this check if there is an address retry record and there is not a
2042 domain retry record; this implies that previous attempts to handle
2043 the address had the retry_use_local_parts option turned on. We use
2044 this as an approximation for the destination being like a local
2045 delivery, as in LMTP.
2047 I suspect this new check makes the old local delivery cutoff check
2048 redundant, but I have not verified this so I left the code in place.
2050 TF/02 Correct gecos expansion when From: is a prefix of the username.
2052 Test 0254 submits a message to Exim with the header
2056 When I ran the test suite under the user fanf2, Exim expanded
2057 the header to contain my full name, whereas it should have added
2058 a Resent-Sender: header. It erroneously treats any prefix of the
2059 username as equal to the username.
2061 This change corrects that bug.
2063 GF/01 DCC debug and logging tidyup
2064 Error conditions log to paniclog rather than rejectlog.
2065 Debug lines prefixed by "DCC: " to remove any ambiguity.
2067 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2069 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2070 Bug spotted by Jeremy Harris; was flawed since initial commit.
2071 Would have resulted in OCSP responses post-SNI triggering an Exim
2072 NULL dereference and crash.
2074 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2076 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2077 Bug detection, analysis and fix by Samuel Thibault.
2078 Bugzilla 1331, Debian bug #698092.
2080 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2082 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2083 Server implementation by Todd Lyons, client by JH.
2084 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2085 config variable "prdr_enable" controls whether the server
2086 advertises the facility. If the client requests PRDR a new
2087 acl_data_smtp_prdr ACL is called once for each recipient, after
2088 the body content is received and before the acl_smtp_data ACL.
2089 The client is controlled by both of: a hosts_try_prdr option
2090 on the smtp transport, and the server advertisement.
2091 Default client logging of deliveries and rejections involving
2092 PRDR are flagged with the string "PRDR".
2094 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2095 fclose(). Diagnosis by Todd Lyons.
2097 PP/17 Update configure.default to handle IPv6 localhost better.
2098 Patch by Alain Williams (plus minor tweaks).
2101 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2102 This is now consistent with GnuTLS, and is now documented: the
2103 previous undocumented portable approach to treating the option as
2104 unset was to force an expansion failure. That still works, and
2105 an empty string is now equivalent.
2107 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2108 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2109 not performing validation itself.
2111 PP/20 Added force_command boolean option to pipe transport.
2112 Patch from Nick Koston, of cPanel Inc.
2114 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2117 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2119 PP/21 Fix eximon continuous updating with timestamped log-files.
2120 Broken in a format-string cleanup in 4.80, missed when I repaired the
2121 other false fix of the same issue.
2122 Report and fix from Heiko Schlichting.
2125 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2126 Report from Prashanth Katuri.
2128 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2129 It's SecureTransport, so affects any MacOS clients which use the
2130 system-integrated TLS libraries, including email clients.
2132 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2133 using a MIME ACL for non-SMTP local injection.
2134 Report and assistance in diagnosis by Warren Baker.
2136 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2138 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2140 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2141 last id that failed so it may be referenced in subsequent ACL's.
2143 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2146 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2147 ldap library initialization, allowing self-signed CA's to be
2148 used. Also properly sets require_cert option later in code by
2149 using NULL (global ldap config) instead of ldap handle (per
2150 session). Bug diagnosis and testing by alxgomz.
2152 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2153 the src/util/ subdirectory.
2155 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2156 renamed to Transport Post Delivery Action by Jeremy Harris, as
2159 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2160 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2161 redis_servers = needs to be configured which will be used by the redis
2162 lookup. Patch from Warren Baker, of The Packet Hub.
2164 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2166 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2167 hostname or reverse DNS when processing a host list. Used suggestions
2168 from multiple comments on this bug.
2170 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2172 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2173 Missed a few lines, added it to make the runtest require no keyboard
2176 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2177 contains upper case chars. Make router use caseful_local_part.
2179 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2180 support when GnuTLS has been built with p11-kit.
2186 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2188 This, or similar/improved, will also be change PP/11 of 4.82.
2194 PP/01 Handle short writes when writing local log-files.
2195 In practice, only affects FreeBSD (8 onwards).
2196 Bugzilla 1053, with thanks to Dmitry Isaikin.
2198 NM/01 Bugzilla 949 - Documentation tweak
2200 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2203 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2205 PP/02 Implemented gsasl authenticator.
2207 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2209 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2210 `pkg-config foo` for cflags/libs.
2212 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2213 with rest of GSASL and with heimdal_gssapi.
2215 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2216 `pkg-config foo` for cflags/libs for the TLS implementation.
2218 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2219 properties get this fed in as external SSF. A number of robustness
2220 and debugging improvements to the cyrus_sasl authenticator.
2222 PP/08 cyrus_sasl server now expands the server_realm option.
2224 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2225 Patch by Jeremy Harris.
2227 PP/10 Added dbmjz lookup type.
2229 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2231 PP/12 MAIL args handles TAB as well as SP, for better interop with
2232 non-compliant senders.
2233 Analysis and variant patch by Todd Lyons.
2235 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2236 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2237 Patch from Dirk Mueller <dmueller@suse.com>
2239 PP/13 tls_peerdn now print-escaped for spool files.
2240 Observed some $tls_peerdn in wild which contained \n, which resulted
2241 in spool file corruption.
2243 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2244 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2245 or write after TLS renegotiation, which otherwise led to messages
2248 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2249 as a tracking header (ie: a signed header comes before the signature).
2250 Patch from Wolfgang Breyha.
2252 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2253 comma-sep list; embedded commas doubled.
2255 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2257 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2259 Report and patch from Dmitry Banschikov.
2261 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2262 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2263 failures. We appear to now support TLS1.1+ with Exim.
2265 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2266 lets Exim select keys and certificates based upon TLS SNI from client.
2267 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2268 before an outbound SMTP session. New log_selector, +tls_sni.
2270 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2271 NULL dereference. Report and patch from Alun Jones.
2273 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2274 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2275 Not seeing resolver debug output on NetBSD, but suspect this is a
2276 resolver implementation change.
2278 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2279 Left warnings. Added "eximon gdb" invocation mode.
2281 PP/21 Defaulting "accept_8bitmime" to true, not false.
2283 PP/22 Added -bw for inetd wait mode support.
2285 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2286 locate the relevant includes and libraries. Made this the default.
2288 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2289 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2291 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2292 This may cause build issues on older platforms.
2294 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2295 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2296 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2297 Added SNI support via GnuTLS too.
2298 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2300 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2302 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2303 Applied second patch from Janne, implementing suggestion to default
2304 multiple-strings-in-record handling to match SPF spec.
2306 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2308 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2309 read-only, out of scope).
2310 Patch from Wolfgang Breyha, report from Stuart Northfield.
2312 PP/29 Fix three issues highlighted by clang analyser static analysis.
2313 Only crash-plausible issue would require the Cambridge-specific
2314 iplookup router and a misconfiguration.
2315 Report from Marcin Mirosław.
2317 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2319 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2320 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2321 As part of this, removing so much warning spew let me fix some minor
2322 real issues in debug logging.
2324 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2325 assignment on my part. Fixed.
2327 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2328 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2329 Janne Snabb (who went above and beyond: thank you).
2331 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2332 string otherwise requires a connection and a bunch more work and it's
2333 relatively easy to get wrong. Should also expose TLS library linkage
2336 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2337 64-bit ${eval} (JH/03).
2339 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2340 GNU libc to support some of the 64-bit stuff, should not lead to
2341 conflicts. Defined before os.h is pulled in, so if a given platform
2342 needs to override this, it can.
2344 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2345 protection layer was required, which is not implemented.
2346 Bugzilla 1254, patch from Wolfgang Breyha.
2348 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2349 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2350 tls_dhparam take prime identifiers. Also unbreak combination of
2351 OpenSSL+DH_params+TLSSNI.
2353 PP/39 Disable SSLv2 by default in OpenSSL support.
2359 PP/01 Solaris build fix for Oracle's LDAP libraries.
2360 Bugzilla 1109, patch from Stephen Usher.
2362 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2364 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2367 TF/02 Fix a couple more cases where we did not log the error message
2368 when unlink() failed. See also change 4.74-TF/03.
2370 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2371 lock up or crash if it happened to be inside a call to libc when it
2372 got a SIGUSR1 from exiwhat.
2374 The SIGUSR1 handler appends the current process status to the process
2375 log which is later printed by exiwhat. It used to use the general
2376 purpose logging code to do this, but several functions it calls are
2377 not safe for signals.
2379 The new output code in the SIGUSR1 handler is specific to the process
2380 log, and simple enough that it's easy to inspect for signal safety.
2381 Removing some special cases also simplifies the general logging code.
2382 Removing the spurious timestamps from the process log simplifies
2385 TF/04 Improved ratelimit ACL condition.
2387 The /noupdate option has been deprecated in favour of /readonly which
2388 has clearer semantics. The /leaky, /strict, and /readonly update modes
2389 are mutually exclusive. The update mode is no longer included in the
2390 database key; it just determines when the database is updated. (This
2391 means that when you upgrade Exim will forget old rate measurements.)
2393 Exim now checks that the per_* options are used with an update mode that
2394 makes sense for the current ACL. For example, when Exim is processing a
2395 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2396 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2397 must specify per_mail/readonly. If you omit the update mode it defaults to
2398 /leaky where that makes sense (as before) or /readonly where required.
2400 The /noupdate option is now undocumented but still supported for
2401 backwards compatibility. It is equivalent to /readonly except that in
2402 ACLs where /readonly is required you may specify /leaky/noupdate or
2403 /strict/noupdate which are treated the same as /readonly.
2405 A useful new feature is the /count= option. This is a generalization
2406 of the per_byte option, so that you can measure the throughput of other
2407 aggregate values. For example, the per_byte option is now equivalent
2408 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2410 The per_rcpt option has been generalized using the /count= mechanism
2411 (though it's more complicated than the per_byte equivalence). When it is
2412 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2413 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2414 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2415 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2416 Note that using per_rcpt with a non-readonly update mode in more than
2417 one ACL will cause the recipients to be double-counted. (The per_mail
2418 and per_byte options don't have this problem.)
2420 The handling of very low rates has changed slightly. If the computed rate
2421 is less than the event's count (usually one) then this event is the first
2422 after a long gap. In this case the rate is set to the same as this event's
2423 count, so that the first message of a spam run is counted properly.
2425 The major new feature is a mechanism for counting the rate of unique
2426 events. The new per_addr option counts the number of different
2427 recipients that someone has sent messages to in the last time period. It
2428 behaves like per_rcpt if all the recipient addresses are different, but
2429 duplicate recipient addresses do not increase the measured rate. Like
2430 the /count= option this is a general mechanism, so the per_addr option
2431 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2432 example, measure the rate that a client uses different sender addresses
2433 with the options per_mail/unique=$sender_address. There are further
2434 details in the main documentation.
2436 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2438 TF/06 Removed a few PCRE remnants.
2440 TF/07 Automatically extract Exim's version number from tags in the git
2441 repository when doing development or release builds.
2443 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2444 Bugzilla 879. Patch from Paul Fisher.
2446 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2447 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2450 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2452 PP/05 Variable $av_failed, true if the AV scanner deferred.
2453 Bugzilla 1078. Patch from John Horne.
2455 PP/06 Stop make process more reliably on build failure.
2456 Bugzilla 1087. Patch from Heiko Schlittermann.
2458 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2459 Bugzilla 1089. Patch from Heiko Schlittermann.
2461 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2462 Bugzilla 1131. Patch from Holger Weiß.
2464 PP/09 Handle IPv6 addresses with SPF.
2465 Bugzilla 860. Patch from Wolfgang Breyha.
2467 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2469 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2472 PP/11 match_* no longer expand right-hand-side by default.
2473 New compile-time build option, EXPAND_LISTMATCH_RHS.
2474 New expansion conditions, "inlist", "inlisti".
2476 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2478 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2480 PP/14 fix log_write() format string regression from TF/03.
2481 Bugzilla 1152. Patch from Dmitry Isaikin.
2487 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2489 PP/02 Harmonised TLS library version reporting; only show if debugging.
2490 Layout now matches that introduced for other libraries in 4.74 PP/03.
2492 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2494 PP/04 New "dns_use_edns0" global option.
2496 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2499 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2500 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2502 TK/01 Updated PolarSSL code to 0.14.2.
2503 Bugzilla 1097. Patch from Andreas Metzler.
2505 PP/07 Catch divide-by-zero in ${eval:...}.
2506 Fixes bugzilla 1102.
2508 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2511 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2512 format-string attack -- SECURITY: remote arbitrary code execution.
2514 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2515 time unintentionally subject to list matching rules, letting the header
2516 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2517 arbitrary string expansion). This allowed for information disclosure.
2519 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2520 INT_MIN/-1 -- value coerced to INT_MAX.
2526 NM/01 Workaround for PCRE version dependency in version reporting
2529 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2530 This fixes portability to compilers other than gcc, notably
2531 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2533 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2534 makefiles for portability to HP-UX and POSIX correctness.
2536 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2537 Also via indented variable definition in the Makefile.
2538 (Debugging by Oliver Heesakkers).
2540 PP/02 Restore caching of spamd results with expanded spamd_address.
2541 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2543 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2544 Improves build reliability. Fix from: Frank Elsner
2546 NM/02 Fix wide character breakage in the rfc2047 coding
2547 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2549 NM/03 Allow underscore in dnslist lookups
2550 Fixes bug 1026. Patch from Graeme Fowler
2552 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2553 Code patches from Adam Ciarcinski of NetBSD.
2555 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2558 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2559 is logged, to avoid truncation. Patch from John Horne.
2561 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2562 Patch from Jakob Hirsch.
2564 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2565 SQL string expansion failure details.
2566 Patch from Andrey Oktyabrski.
2568 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2569 Patch from Simon Arlott.
2571 PP/09 New lookups functionality failed to compile on old gcc which rejects
2572 extern declarations in function scope.
2573 Patch from Oliver Fleischmann
2575 PP/10 Use sig_atomic_t for flags set from signal handlers.
2576 Check getgroups() return and improve debugging.
2577 Fixed developed for diagnosis in bug 927 (which turned out to be
2580 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2581 Patch from Mark Zealey.
2583 PP/12 Bugzilla 1056: Improved spamd server selection.
2584 Patch from Mark Zealey.
2586 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2587 Based on patch from Heiko Schlittermann.
2589 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2590 Patch from Uwe Doering, sign-off by Michael Haardt.
2592 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2593 variable declaration deep within a block. Bug and patch from
2596 PP/15 lookups-Makefile IRIX compatibility coercion.
2598 PP/16 Make DISABLE_DKIM build knob functional.
2600 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2601 Patch by Simon Arlott
2603 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2604 variable argument macros. Our copy now differs from upstream.
2610 TF/01 Failure to get a lock on a hints database can have serious
2611 consequences so log it to the panic log.
2613 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2614 controlled using the smtp_confirmation log selector.
2616 TF/03 Include the error message when we fail to unlink a spool file.
2618 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2619 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2620 for maintaining out-of-tree patches for some time.
2622 PP/01 Bugzilla 139: Documentation and portability issues.
2623 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2624 Handle per-OS dynamic-module compilation flags.
2626 PP/02 Let /dev/null have normal permissions.
2627 The 4.73 fixes were a little too stringent and complained about the
2628 permissions on /dev/null. Exempt it from some checks.
2629 Reported by Andreas M. Kirchwitz.
2631 PP/03 Report version information for many libraries, including
2632 Exim version information for dynamically loaded libraries. Created
2633 version.h, now support a version extension string for distributors
2634 who patch heavily. Dynamic module ABI change.
2636 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2637 privilege escalation vulnerability whereby the Exim run-time user
2638 can cause root to append content of the attacker's choosing to
2641 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2644 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2645 If dropping privileges for untrusted macros, we disabled normal logging
2646 on the basis that it would fail; for the Exim run-time user, this is not
2647 the case, and it resulted in successful deliveries going unlogged.
2648 Fixed. Reported by Andreas Metzler.
2654 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2655 only prepend for the Resent-* case. Fixes regression introduced in
2656 Exim 4.70 by NM/22 for Bugzilla 607.
2658 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2659 increasing numbers of administrators be bitten by this.
2661 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2663 PP/03 Bugzilla 994: added openssl_options main configuration option.
2665 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2667 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2669 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2671 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2672 without a peer certificate, leading to a segfault because of an
2673 assumption that peers always have certificates. Be a little more
2674 paranoid. Problem reported by Martin Tscholak.
2676 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2677 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2678 NB: ClamAV planning to remove STREAM in "middle of 2010".
2679 CL also introduces -bmalware, various -d+acl logging additions and
2680 more caution in buffer sizes.
2682 PP/09 Implemented reverse_ip expansion operator.
2684 PP/10 Bugzilla 937: provide a "debug" ACL control.
2686 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2688 PP/12 Bugzilla 973: Implement --version.
2690 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2692 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2694 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2696 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2697 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2698 ignore trailing whitespace.
2700 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2702 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2705 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2706 Notification from Dr Andrew Aitchison.
2708 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2709 ExtendedDetectionInfo response format.
2710 Notification from John Horne.
2712 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2715 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2716 XSL and documented dependency on system catalogs, with examples of how
2719 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2722 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2723 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2724 configuration file which is writeable by the Exim user or group.
2726 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2727 of configuration files to cover files specified with the -C option if
2728 they are going to be used with root privileges, not just the default
2731 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2732 option (effectively making it always true).
2734 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2735 files to be used while preserving root privileges.
2737 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2738 that rogue child processes cannot use them.
2740 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2741 run-time user, instead of root.
2743 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2744 Exim run-time user without dropping privileges.
2746 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2747 result string, instead of calling string_vformat() twice with the same
2750 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2751 for other users. Others should always drop root privileges if they use
2752 -C on the command line, even for a whitelisted configure file.
2754 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2756 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2762 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2763 $data_path, and $header_path variables; fixed documentation bugs and
2766 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2767 exipick to access non-standard spools, including the "frozen" queue
2770 NM/01 Bugzilla 965: Support mysql stored procedures.
2771 Patch from Alain Williams
2773 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2775 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2776 Patch from Andreas Metzler
2778 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2779 Patch from Kirill Miazine
2781 NM/05 Bugzilla 671: Added umask to procmail example.
2783 JJ/03 installed exipick 20100323.0, fixing doc bug
2785 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2786 directory. Notification and patch from Dan Rosenberg.
2788 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2790 TK/02 Improve log output when DKIM signing operation fails.
2792 MH/01 Treat the transport option dkim_domain as a colon separated
2793 list, not as a single string, and sign the message with each element,
2794 omitting multiple occurences of the same signer.
2796 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2797 Bugzilla 985, 986. Patch by Simon Arlott
2799 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2800 Patch by Simon Arlott
2802 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2803 MBX locking. Notification from Dan Rosenberg.
2809 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2811 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2813 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2815 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2817 NM/04 Bugzilla 915: Flag broken perl installation during build.
2823 TK/01 Added patch by Johannes Berg that expands the main option
2824 "spamd_address" if it starts with a dollar sign.
2826 TK/02 Write list of recipients to X-Envelope-Sender header when building
2827 the mbox-format spool file for content scanning (suggested by Jakob
2830 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2831 (http://www.dcc-servers.net/) support via dccifd. Activated by
2832 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2834 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2835 by Mark Daniel Reidel <mr@df.eu>.
2837 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2838 When building exim an external PCRE library is now needed -
2839 PCRE is a system library on the majority of modern systems.
2840 See entry on PCRE_LIBS in EDITME file.
2842 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2843 conversation. Added nologin parameter to request.
2844 Patch contributed by Kirill Miazine.
2846 TF/01 Do not log submission mode rewrites if they do not change the address.
2848 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2850 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2851 log files in place. Contributed by Roberto Lima.
2853 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2855 TF/03 Bugzilla 615: When checking the local_parts router precondition
2856 after a local_part_suffix or local_part_prefix option, Exim now
2857 does not use the address's named list lookup cache, since this
2858 contains cached lookups for the whole local part.
2860 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2861 Robert Millan. Documentation is in experimental-spec.txt.
2863 TF/04 Bugzilla 668: Fix parallel build (make -j).
2865 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2867 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2868 Patch provided by Jan Srzednicki.
2870 TF/05 Leading white space used to be stripped from $spam_report which
2871 wrecked the formatting. Now it is preserved.
2873 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2874 that they are available at delivery time.
2876 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2878 TF/08 TLS error reporting now respects the incoming_interface and
2879 incoming_port log selectors.
2881 TF/09 Produce a more useful error message if an SMTP transport's hosts
2882 setting expands to an empty string.
2884 NM/06 Bugzilla 744: EXPN did not work under TLS.
2885 Patch provided by Phil Pennock.
2887 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2888 Patch provided by Richard Godbee.
2890 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2891 acl_smtp_notquit, added index entry.
2893 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2894 Patch provided by Eugene Bujak.
2896 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2897 accept(). Patch provided by Maxim Dounin.
2899 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2900 Patch provided by Phil Pennock.
2902 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2904 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2905 Patch provided by Brad "anomie" Jorsch.
2907 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2908 Patch provided by Dean Brooks.
2910 TK/05 Add native DKIM support (does not depend on external libraries).
2912 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2913 Patch provided by Graeme Fowler.
2915 NM/16 Bugzilla 851: Documentation example syntax fix.
2917 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2919 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2922 NM/19 Bugzilla 745: TLS version reporting.
2923 Patch provided by Phil Pennock.
2925 NM/20 Bugzilla 167: bool: condition support.
2926 Patch provided by Phil Pennock.
2928 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2929 clients. Patch provided by Phil Pennock.
2931 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2932 Patch provided by Brad "anomie" Jorsch.
2934 NM/23 Bugzilla 687: Fix misparses in eximstats.
2935 Patch provided by Heiko Schlittermann.
2937 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2938 Patch provided by Heiko Schlittermann.
2940 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2941 plus update to original patch.
2943 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2945 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2946 Patch provided by David Brownlee.
2948 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2950 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2952 NM/30 Bugzilla 888: TLS documentation bugfixes.
2954 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2956 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2957 Unlike the original bugzilla I have changed all shell scripts in src tree.
2959 NM/33 Bugzilla 898: Transport filter timeout fix.
2960 Patch by Todd Rinaldo.
2962 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2963 Patch by Serge Demonchaux.
2965 NM/35 Bugzilla 39: Base64 decode bug fixes.
2966 Patch by Jakob Hirsch.
2968 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2970 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2972 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2974 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2980 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2981 ALT-N's libdkim that I have put here:
2982 http://duncanthrax.net/exim-experimental/
2984 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2985 were called 'true' and it seems that C99 defines that as a reserved
2986 keyword to be used with 'bool' variable types. That means you could
2987 not include C99-style headers which use bools without triggering
2988 build errors in sieve.c.
2990 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2991 as mailq or other aliases. Changed the --help handling significantly
2992 to do whats expected. exim_usage() emits usage/help information.
2994 SC/01 Added the -bylocaldomain option to eximstats.
2996 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2998 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3000 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3006 PH/01 Another patch from the Sieve maintainer.
3008 PH/02 When an IPv6 address is converted to a string for single-key lookup
3009 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3010 dots are used instead of colons so that keys in lsearch files need not
3011 contain colons. This was done some time before quoting was made available
3012 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3013 (notated using the quote facility) so as to distinguish them from IPv4
3014 keys. This meant that lookups for IP addresses in host lists did not work
3015 for iplsearch lookups.
3017 This has been fixed by arranging for IPv6 addresses to be expressed with
3018 colons if the lookup type is iplsearch. This is not incompatible, because
3019 previously such lookups could never work.
3021 The situation is now rather anomalous, since one *can* have colons in
3022 ordinary lsearch keys. However, making the change in all cases is
3023 incompatible and would probably break a number of configurations.
3025 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3028 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3029 conversion specification without a maximum field width, thereby enabling
3030 a rogue spamd server to cause a buffer overflow. While nobody in their
3031 right mind would setup Exim to query an untrusted spamd server, an
3032 attacker that gains access to a server running spamd could potentially
3033 exploit this vulnerability to run arbitrary code as the Exim user.
3035 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3036 $primary_hostname instead of what libspf2 thinks the hosts name is.
3038 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3039 a directory entry by the name of the lookup key. Previously, if a
3040 symlink pointed to a non-existing file or a file in a directory that
3041 Exim lacked permissions to read, a lookup for a key matching that
3042 symlink would fail. Now it is enough that a matching directory entry
3043 exists, symlink or not. (Bugzilla 503.)
3045 PH/03 The body_linecount and body_zerocount variables are now exported in the
3048 PH/04 Added the $dnslist_matched variable.
3050 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3051 This means they are set thereafter only if the connection becomes
3054 PH/06 Added the client_condition to authenticators so that some can be skipped
3055 by clients under certain conditions.
3057 PH/07 The error message for a badly-placed control=no_multiline_responses left
3058 "_responses" off the end of the name.
3060 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3062 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3063 (without spaces) instead of just copying the configuration text.
3065 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3067 PH/11 Added $max_received_linelength.
3069 PH/12 Added +ignore_defer and +include_defer to host lists.
3071 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3072 way in which PCRE > 7.0 is built.
3074 PH/14 Implemented queue_only_load_latch.
3076 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3077 MAIL command. The effect was to mangle the value on 64-bit systems.
3079 PH/16 Another patch from the Sieve maintainer.
3081 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3083 PH/18 If a system quota error occurred while trying to create the file for
3084 a maildir delivery, the message "Mailbox is full" was not appended to the
3085 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3086 only to a quota excession during the actual writing of the file.
3088 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3089 characters?) which causes problems in log lines. The DN values are now
3090 passed through string_printing() before being added to log lines.
3092 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3093 and InterBase are left for another time.)
3095 PH/21 Added message_body_newlines option.
3097 PH/22 Guard against possible overflow in moan_check_errorcopy().
3099 PH/23 POSIX allows open() to be a macro; guard against that.
3101 PH/24 If the recipient of an error message contained an @ in the local part
3102 (suitably quoted, of course), incorrect values were put in $domain and
3103 $local_part during the evaluation of errors_copy.
3109 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3110 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3113 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3114 issue a MAIL command.
3116 PH/02 In an ACL statement such as
3118 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3120 if a client was not listed at all, or was listed with a value other than
3121 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3122 the condition was not true (as it should be), so access was not denied.
3123 The bug was that the ! inversion was incorrectly passed on to the second
3124 item. This has been fixed.
3126 PH/03 Added additional dnslists conditions == and =& which are different from
3127 = and & when the dns lookup returns more than one IP address.
3129 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3130 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3132 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3133 FSYNC, which compiles an option called disable_fsync that allows for
3134 bypassing fsync(). The documentation is heavily laced with warnings.
3136 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3138 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3139 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3140 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3141 including adding "make clean"; (3) Added -fPIC when compiling the test
3142 dynamically loaded module, to get rid of a warning.
3144 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3145 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3146 = 0s. The bug is otherwise harmless.
3148 PH/07 There was a bug in the dovecot authenticator such that the value of
3149 $auth1 could be overwritten, and so not correctly preserved, after a
3150 successful authentication. This usually meant that the value preserved by
3151 the server_setid option was incorrect.
3153 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3155 PH/09 Installed PCRE release 7.0.
3157 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3158 run for batched SMTP input. It is now run at the start of every message
3159 in the batch. While fixing this I discovered that the process information
3160 (output by running exiwhat) was not always getting set for -bs and -bS
3161 input. This is fixed, and it now also says "batched" for BSMTP.
3163 PH/11 Added control=no_pipelining.
3165 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3166 patch, slightly modified), and move the expansion of helo_data till after
3167 the connection is made in the smtp transport (so it can use these
3170 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3172 PH/14 Added log_selector = +pid.
3174 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3176 PH/16 Add ${if forany and ${if forall.
3178 PH/17 Added dsn_from option to vary the From: line in DSNs.
3180 PH/18 Flush SMTP output before performing a callout, unless control =
3181 no_callout_flush is set.
3183 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3184 was true (the default) a successful delivery failed to delete the retry
3185 item, thus causing premature timeout of the address. The bug is now
3188 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3190 PH/21 Long custom messages for fakedefer and fakereject are now split up
3191 into multiline responses in the same way that messages for "deny" and
3192 other ACL rejections are.
3194 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3195 with slight modification.
3197 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3198 draft, changing the syntax and factoring some duplicate code".
3200 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3201 for deliveries of the second and subsequent messages over the same SMTP
3204 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3205 ${reduce, with only minor "tidies".
3207 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3209 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3210 expansion side effects.
3212 PH/27 When a message times out after an over-quota error from an Exim-imposed
3213 quota, the bounce message says "mailbox is full". This message was not
3214 being given when it was a system quota that was exceeded. It now should
3217 MH/03 Made $recipients available in local_scan(). local_scan() already has
3218 better access to the recipient list through recipients_list[], but
3219 $recipients can be useful in postmaster-provided expansion strings.
3221 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3222 in the case of a MAIL command with additional options following the
3223 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3224 were accidentally chopped off.
3226 PH/29 SMTP synchronization checks are implemented when a command is read -
3227 there is a check that no more input is waiting when there shouldn't be
3228 any. However, for some commands, a delay in an ACL can mean that it is
3229 some time before the response is written. In this time, more input might
3230 arrive, invalidly. So now there are extra checks after an ACL has run for
3231 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3232 pipelining has not been advertised.
3234 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3236 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3237 returned in the SMTP response when smtp_return_error_details was false.
3238 This has been fixed.
3240 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3241 instead of the C I/O that was originally supplied, because problems were
3242 reported on Solaris.
3244 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3245 Exim which did not show up earlier: it was assuming that a call to
3246 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3247 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3248 was a macro that became an assignment, so it seemed to work. This has
3249 changed to a proper function call with a void return, hence the compile
3250 error. Exim's code has been fixed.
3252 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3255 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3257 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3259 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3260 because it turns out to be too big or there is a timeout) there is no
3261 "Completed" line in the log. When some messages of this type were
3262 selected by exigrep, they were listed as "not completed". Others were
3263 picked up by some special patterns. I have improved the selection
3264 criteria to be more general.
3266 PH/38 The host_find_failed option in the manualroute router can now be set
3267 to "ignore", to completely ignore a host whose IP address cannot be
3268 found. If all hosts are ignored, the behaviour is controlled by the new
3269 host_all_ignored option.
3271 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3272 homing or because of multiple MX records with /mx) generated more than
3273 one IP address, and the following item turned out to be the local host,
3274 all the secondary addresses of the first item were incorrectly removed
3275 from the list, along with the local host and any following hosts (which
3276 is what is supposed to happen).
3278 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3279 whoever called Exim into the -H file. In the case of the daemon it was
3280 behaving confusingly. When first started, it used values for whoever
3281 started the daemon, but after a SIGHUP it used the Exim user (because it
3282 calls itself on a restart). I have changed the code so that it now always
3285 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3286 message are rejected with the same error (e.g. no authentication or bad
3287 sender address), and a DATA command is nevertheless sent (as can happen
3288 with PIPELINING or a stupid MUA), the error message that was given to the
3289 RCPT commands is included in the rejection of the DATA command. This is
3290 intended to be helpful for MUAs that show only the final error to their
3293 PH/42 Another patch from the Sieve maintainer.
3295 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3296 Eximstats - Fixed some broken HTML links and added missing column headers
3298 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3299 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3301 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3303 PH/43 Yet another patch from the Sieve maintainer.
3305 PH/44 I found a way to check for a TCP/IP connection going away before sending
3306 the response to the final '.' that terminates a message, but only in the
3307 case where the client has not sent further data following the '.'
3308 (unfortunately, this is allowed). However, in many cases there won't be
3309 any further data because there won't be any more messages to send. A call
3310 to select() can be used: if it shows that the input is "ready", there is
3311 either input waiting, or the socket has been closed. An attempt to read
3312 the next input character can distinguish the two cases. Previously, Exim
3313 would have sent an OK response which the client would never have see.
3314 This could lead to message repetition. This fix should cure that, at
3315 least in a lot of common cases.
3317 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3318 advertised in response to EHLO.
3324 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3325 fixed by 4.65/MH/01 (is this a record?) are fixed:
3327 (i) An empty string was always treated as zero by the numeric comparison
3328 operators. This behaviour has been restored.
3330 (ii) It is documented that the numeric comparison operators always treat
3331 their arguments as decimal numbers. This was broken in that numbers
3332 starting with 0 were being interpreted as octal.
3334 While fixing these problems I realized that there was another issue that
3335 hadn't been noticed. Values of message_size_limit (both the global option
3336 and the transport option) were treated as octal if they started with 0.
3337 The documentation was vague. These values are now always treated as
3338 decimal, and I will make that clear in the documentation.
3344 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3345 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3348 MH/01 Don't check that the operands of numeric comparison operators are
3349 integers when their expansion is in "skipping" mode (fixes bug
3350 introduced by 4.64-PH/07).
3352 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3353 child addresses, Exim now panics and dies. Previously, because the count
3354 is held in a short int, deliveries were likely to be lost. As such a
3355 large number of recipients for a single message is ridiculous
3356 (performance will be very, very poor), I have chosen to impose a limit
3357 rather than extend the field.
3363 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3364 leftover -K file (the existence of which was triggered by #402).
3365 While we were at it, introduced process PID as part of the -K
3366 filename. This should rule out race conditions when creating
3369 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3370 processing considerably. Previous code took too long for large mails,
3371 triggering a timeout which in turn triggers #401.
3373 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3374 in the DK code in transports.c. sendfile() is not really portable,
3375 hence the _LINUX specificness.
3377 TF/01 In the add_headers option to the mail command in an Exim filter,
3378 there was a bug that Exim would claim a syntax error in any
3379 header after the first one which had an odd number of characters
3382 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3383 callout verification, Exim cached a "reject" for the entire domain. This
3384 is correct for most verifications, but it is not correct for a recipient
3385 verification with use_sender or use_postmaster set, because in that case
3386 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3387 case of MAIL FROM:<> rejection from other early rejections (e.g.
3388 rejection of HELO). When verifying a recipient using a non-null MAIL
3389 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3390 Whatever the result of the callout, the value of the domain cache is
3391 left unchanged (for any other kind of callout, getting as far as trying
3392 RCPT means that the domain itself is ok).
3394 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3397 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3398 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3399 session, the daemon ignores ECONNECT errors and logs others; it now
3400 ignores EPIPE as well.
3402 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3403 (quoted-printable decoding).
3405 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3406 later the small subsequent patch to fix an introduced bug.
3408 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3410 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3412 PH/08 An error is now given if message_size_limit is specified negative.
3414 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3415 to be given (somewhat) arbitrary names.
3417 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3420 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3421 miscellaneous code fixes
3423 PH/10 Added the log_reject_target ACL modifier to specify where to log
3426 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3427 hostname. This is wrong, because it relates to the incoming message (and
3428 probably the interface on which it is arriving) and not to the outgoing
3429 callout (which could be using a different interface). This has been
3430 changed to use the value of the helo_data option from the smtp transport
3431 instead - this is what is used when a message is actually being sent. If
3432 there is no remote transport (possible with a router that sets up host
3433 addresses), $smtp_active_hostname is used.
3435 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3436 tweaks were necessary in order to get it to work (see also 21 below):
3437 (a) The code assumed that strncpy() returns a negative number on buffer
3438 overflow, which isn't the case. Replaced with Exim's string_format()
3440 (b) There were several signed/unsigned issues. I just did the minimum
3441 hacking in of casts. There is scope for a larger refactoring.
3442 (c) The code used strcasecmp() which is not a standard C function.
3443 Replaced with Exim's strcmpic() function.
3444 (d) The code set only $1; it now sets $auth1 as well.
3445 (e) A simple test gave the error "authentication client didn't specify
3446 service in request". It would seem that Dovecot has changed its
3447 interface. Fortunately there's a specification; I followed it and
3448 changed what the client sends and it appears to be working now.
3450 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3453 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3454 address A is aliased to B and C, where B exists and C does not. Without
3455 -v the output is "A verified" because verification stops after a
3456 successful redirection if more than one address is generated. However,
3457 with -v the child addresses are also verified. Exim was outputting "A
3458 failed to verify" and then showing the successful verification for C,
3459 with its parentage. It now outputs "B failed to verify", showing B's
3460 parentage before showing the successful verification of C.
3462 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3463 look up a TXT record in a specific list after matching in a combined
3466 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3467 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3468 they consult the DNS. I had assumed they would set it the way they
3469 wanted; and indeed my experiments on Linux seem to show that in some
3470 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3471 To be on the safe side, however, I have now made the interface to
3472 host_find_byname() similar to host_find_bydns(), with an argument
3473 containing the DNS resolver options. The host_find_byname() function now
3474 sets these options at its start, just as host_find_bydns() does. The smtp
3475 transport options dns_qualify_single and dns_search_parents are passed to
3476 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3477 of host_find_byname() use the default settings of RES_DEFNAMES
3478 (qualify_single) but not RES_DNSRCH (search_parents).
3480 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3481 spool_read_header() do less string testing, by means of a preliminary
3482 switch on the second character of optional "-foo" lines. (This is
3483 overdue, caused by the large number of possibilities that now exist.
3484 Originally there were few.) While I was there, I also converted the
3485 str(n)cmp tests so they don't re-test the leading "-" and the first
3486 character, in the hope this might squeeze out yet more improvement.
3488 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3489 flag allowing group syntax was set by the header_syntax check but not
3490 turned off, possible causing trouble later; (2) The flag was not being
3491 set at all for the header_verify test, causing "group"-style headers to
3492 be rejected. I have now set it in this case, and also caused header_
3493 verify to ignore an empty address taken from a group. While doing this, I
3494 came across some other cases where the code for allowing group syntax
3495 while scanning a header line wasn't quite right (mostly, not resetting
3496 the flag correctly in the right place). These bugs could have caused
3497 trouble for malformed header lines. I hope it is now all correct.
3499 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3500 with the "reply" argument non-NULL. The code, however (which originally
3501 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3502 but it didn't always do it. This confused somebody who was copying the
3503 code for some other use. I have removed all the tests.
3505 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3506 feature that was used to support insecure browsers during the U.S. crypto
3507 embargo. It requires special client support, and Exim is probably the
3508 only MTA that supported it -- and would never use it because real RSA is
3509 always available. This code has been removed, because it had the bad
3510 effect of slowing Exim down by computing (never used) parameters for the
3511 RSA_EXPORT functionality.
3513 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3514 authenticator to fail if there's a tab character in the incoming data
3515 (there should never be unless someone is messing about, as it's supposed
3516 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3517 if the connection is using TLS or if the remote IP is the same as the
3518 local IP, and the "valid-client-cert option" if a client certificate has
3521 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3522 authenticators. This can be used for authorization after authentication
3523 succeeds. (In the case of plaintext, it servers for both authentication
3526 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3527 if any retry times were supplied.
3529 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3530 connection, where there is no client IP address to check. In this
3531 situation, the verify now always succeeds.
3533 PH/25 Applied John Jetmore's -Mset patch.
3535 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3537 PH/27 In a string expansion for a processed (not raw) header when multiple
3538 headers of the same name were present, leading whitespace was being
3539 removed from all of them, but trailing whitespace was being removed only
3540 from the last one. Now trailing whitespace is removed from each header
3541 before concatenation. Completely empty headers in a concatenation (as
3542 before) are ignored.
3544 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3545 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3547 PH/29 [Removed. This was a change that I later backed out, and forgot to
3548 correct the ChangeLog entry (that I had efficiently created) before
3549 committing the later change.]
3551 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3552 address errors (4xx response to RCPT) over the same connection as other
3553 messages routed to the same hosts. Such deliveries are always "forced",
3554 so retry times are not inspected. This resulted in far too many retries
3555 for the affected addresses. The effect occurred only when there were more
3556 hosts than the hosts_max_try setting in the smtp transport when it had
3557 the 4xx errors. Those hosts that it had tried were not added to the list
3558 of hosts for which the message was waiting, so if all were tried, there
3559 was no problem. Two fixes have been applied:
3561 (i) If there are any address or message errors in an SMTP delivery, none
3562 of the hosts (tried or untried) are now added to the list of hosts
3563 for which the message is waiting, so the message should not be a
3564 candidate for sending over the same connection that was used for a
3565 successful delivery of some other message. This seems entirely
3566 reasonable: after all the message is NOT "waiting for some host".
3567 This is so "obvious" that I'm not sure why it wasn't done
3568 previously. Hope I haven't missed anything, but it can't do any
3569 harm, as the worst effect is to miss an optimization.
3571 (ii) If, despite (i), such a delivery is accidentally attempted, the
3572 routing retry time is respected, so at least it doesn't keep
3573 hammering the server.
3575 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3576 in ${readsocket because some servers need this prod.
3578 PH/32 Added some extra debug output when updating a wait-xxx database.
3580 PH/33 The hint "could be header name not terminated by colon", which has been
3581 given for certain expansion errors for a long time, was not being given
3582 for the ${if def:h_colon_omitted{... case.
3584 PH/34 The spec says: "With one important exception, whenever a domain list is
3585 being scanned, $domain contains the subject domain." There was at least
3586 one case where this was not true.
3588 PH/35 The error "getsockname() failed: connection reset by peer" was being
3589 written to the panic log as well as the main log, but it isn't really
3590 panic-worthy as it just means the connection died rather early on. I have
3591 removed the panic log writing for the ECONNRESET error when getsockname()
3594 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3595 runs only) independently of the message's sender address. This meant
3596 that, if the 4xx error was in fact related to the sender, a different
3597 message to the same recipient with a different sender could confuse
3598 things. In particular, this can happen when sending to a greylisting
3599 server, but other circumstances could also provoke similar problems.
3600 I have changed the default so that the retry time for these errors is now
3601 based a combination of the sender and recipient addresses. This change
3602 can be overridden by setting address_retry_include_sender=false in the
3605 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3606 remote server are returned as part of bounce messages. This was not
3607 happening for LMTP over a pipe (the lmtp transport), but now it is the
3608 same for both kinds of LMTP.
3610 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3611 in header lines that were in fact CNAMEs. This is no longer the case.
3613 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3614 and queue runs started by the daemon processed all messages. This has
3615 been fixed so that -R and -S can now usefully be given with -q<time>.
3617 PH/40 Import PCRE release 6.7 (fixes some bugs).
3619 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3621 PH/42 Give an error if -q is specified more than once.
3623 PH/43 Renamed the variables $interface_address and $interface_port as
3624 $received_ip_address and $received_port, to make it clear that these
3625 values apply to message reception, and not to the outgoing interface when
3626 a message is delivered. (The old names remain recognized, of course.)
3628 PH/44 There was no timeout on the connect() call when using a Unix domain
3629 socket in the ${readsocket expansion. There now is.
3631 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3632 be meaningful with "accept".
3634 SC/01 Eximstats V1.43
3635 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3637 SC/02 Eximstats V1.44
3638 Use a glob alias rather than an array ref in the generated
3639 parser. This improves both readability and performance.
3641 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3642 Collect SpamAssassin and rejection statistics.
3643 Don't display local sender or destination tables unless
3644 there is data to show.
3645 Added average volumes into the top table text output.
3647 SC/04 Eximstats V1.46
3648 Collect data on the number of addresses (recipients)
3649 as well as the number of messages.
3651 SC/05 Eximstats V1.47
3652 Added 'Message too big' to the list of mail rejection
3653 reasons (thanks to Marco Gaiarin).
3655 SC/06 Eximstats V1.48
3656 Mainlog lines which have GMT offsets and are too short to
3657 have a flag are now skipped.
3659 SC/07 Eximstats V1.49 (Alain Williams)
3660 Added the -emptyok flag.
3662 SC/08 Eximstats V1.50
3663 Fixes for obtaining the IP address from reject messages.
3665 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3666 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3667 whitespace changes from 4.64-PH/27
3669 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3672 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3673 are found, allow negative numbers in numeric criteria)
3675 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3677 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3678 to match changes made in 4.64-PH/43
3680 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3682 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3683 "domain" is actually an IP address, and force a failure. This locks out
3684 those revolvers/nameservers that support "A-for-A" lookups, in
3685 contravention of the specifications.
3687 PH/48 When a host name was looked up from an IP address, and the subsequent
3688 forward lookup of the name timed out, the host name was left in
3689 $sender_host_name, contrary to the specification.
3691 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3692 restricted to single-key lookups, Exim was not diagnosing an error if
3693 * or *@ was used with a query-style lookup.
3695 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3697 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3698 long ago, but noone interested enough thought of it. Let's just say that
3699 the "1.1" means that there are some new functions that weren't there at
3700 some point in the past.
3702 PH/51 Error processing for expansion failure of helo_data from an smtp
3703 transport during callout processing was broken.
3705 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3706 tested/used via the -bh/-bhc/-bs options.
3708 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3709 bug, fixed in subsequent PCRE releases).
3711 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3712 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3714 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3720 SC/01 Use a glob alias rather than an array ref in eximstats generated
3721 parser. This improves both readability and performance.
3723 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3724 Don't display local sender or destination tables in eximstats unless
3725 there is data to show.
3726 Added average volumes into the eximstats top table text output.
3728 SC/03 Collect data on the number of addresses (recipients) as well
3729 as the number of messages in eximstats.
3731 TF/01 Correct an error in the documentation for the redirect router. Exim
3732 does (usually) call initgroups() when daemonizing.
3734 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3735 with consistent privilege compared to when running as a daemon.
3737 TF/03 Note in the spec that $authenticated_id is not set for local
3738 submissions from trusted users.
3740 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3741 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3743 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3744 by adding some example configuration directives to the default
3745 configuration file. A little bit of work is required to uncomment the
3746 directives and define how usernames and passwords are checked, but
3747 there is now a framework to start from.
3749 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3750 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3751 without this. I don't know how relevant this is to other LDAP libraries.
3753 PH/02 Add the verb name to the "unknown ACL verb" error.
3755 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3757 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3759 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3760 directory not expanded when it should be if an expanded home directory
3761 was set for the address (which is overridden by the transport).
3763 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3766 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3767 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3768 because it is too late at that time, and has no effect.
3770 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3771 security issue with \' (bugzilla #107). I could not use the
3772 PQescapeStringConn() function, because it needs a PGconn value as one of
3775 PH/08 When testing addresses using -bt, indicate those final addresses that
3776 are duplicates that would not cause an additional delivery. At least one
3777 person was confused, thinking that -bt output corresponded to deliveries.
3778 (Suppressing duplicates isn't a good idea as you lose the information
3779 about possibly different redirections that led to the duplicates.)
3781 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3782 systems where poll() doesn't work, in particular OS X.
3784 PH/10 Added more information to debugging output for retry time not reached.
3786 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3787 operations in malware.c.
3789 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3792 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3793 syslog_duplication set false, log lines that would normally be written
3794 both the the main log and to the reject log were not written to syslog at
3797 PH/14 In the default configuration, change the use of "message" in ACL warn
3798 statements to "add_header".
3800 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3801 not followed by a command (e.g. "seen endif").
3803 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3804 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3807 PH/17 Added extra conditions to the default value of delay_warning_condition
3811 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3812 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3813 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3816 The Auto-Submitted: and various List- headers are standardised, whereas I
3817 don't think Precedence: ever was.
3819 PH/18 Refactored debugging code in route_finduser() to show more information,
3820 in particular, the error code if getpwnam() issues one.
3822 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3823 This is apparently needed in addition to the PH/07 change above to avoid
3824 any possible encoding problems.
3826 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3827 but not after initializing Perl.
3829 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3830 output them only if debugging. By default they are written stderr,
3831 apparently, which is not desirable.
3833 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3836 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3839 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3841 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3842 authenticated or an ident call has been made. Suppress the default
3843 values for $authenticated_id and $authenticated_sender (but permit -oMai
3844 and -oMas) when testing with -bh.
3846 PH/24 Re-jigged the order of the tests in the default configuration so that the
3847 tests for valid domains and recipients precede the DNS black list and CSA
3848 tests, on the grounds that those ones are more expensive.
3850 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3851 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3852 This bug exists in every version of Exim that I still have, right back to
3855 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3856 However, an attempt to turn on case-sensitivity in a regex key by
3857 including (?-i) didn't work because the subject string was already
3858 lowercased, and the effects were non-intuitive. It turns out that a
3859 one-line patch can be used to allow (?-i) to work as expected.
3865 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3866 other effects) broke the use of negated acl sub-conditions.
3868 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3871 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3872 "Deny" causes Exim to reject the incoming connection with a 554 error.
3873 Unfortunately, if there is a major crisis, such as a disk failure,
3874 tcp-wrappers gives "deny", whereas what one would like would be some
3875 kind of temporary error. A kludge has been added to help with this.
3876 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3877 554 error is used if errno is still zero or contains ENOENT (which occurs
3878 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3881 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3883 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3884 errors. Otherwise a message that provokes a temporary error (when other
3885 messages do not) can cause a whole host to time out.
3887 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3888 addresses were routed directly to files or pipes from a redirect router.
3889 File deliveries just didn't batch; pipe deliveries might have suffered
3892 PH/06 A failure to get a lock for a hints database would erroneously always say
3893 "Failed to get write lock", even when it was really a read lock.
3895 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3896 of 0600. This has been changed to use the value of the lockfile_mode
3897 option (which defaults to 0600).
3899 PH/08 Applied small patch from the Sieve maintainer.
3901 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3902 folder from quota calculations, a direct delivery into this folder messed
3903 up the contents of the maildirsize file. This was because the regex was
3904 used only to exclude .Trash (or whatever) when the size of the mailbox
3905 was calculated. There was no check that a delivery was happening into an
3906 excluded directory. This bug has been fixed by ignoring all quota
3907 processing for deliveries into excluded directories.
3909 PH/10 Added the maildirfolder_create_regex option to appendfile.
3915 PH/01 The code for finding all the local interface addresses on a FreeBSD
3916 system running IPv6 was broken. This may well have applied to all BSD
3917 systems, as well as to others that have similar system calls. The broken
3918 code found IPv4 interfaces correctly, but gave incorrect values for the
3919 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3920 that it would not match correctly against @[] and not recognize the IPv6
3923 PH/02 The ipliteral router was not recognizing addresses of the form user@
3924 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3926 PH/03 Added disable_ipv6.
3928 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3929 decoded form, because it is most often used to construct To: headers
3930 lines in autoreplies, and the decoded form may well be syntactically
3931 invalid. However, $reply_address has leading white space removed, and all
3932 newlines turned into spaces so that the autoreply transport does not
3935 PH/05 If group was specified without a user on a router, and no group or user
3936 was specified on a transport, the group from the router was ignored.
3938 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3939 for visible compile-time settings that can be used to change these
3940 numbers, for those that want even more. Backwards compatibility with old
3941 spool files has been maintained. However, going back to a previous Exim
3942 release will lost any variables that are in spool files.
3944 PH/07 Two small changes when running in the test harness: increase delay when
3945 passing a TCP/IP connection to a new process, in case the original
3946 process has to generate a bounce, and remove special handling of
3947 127.0.0.2 (sic), which is no longer necessary.
3949 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3950 be the same on different OS.
3952 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3955 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3956 whether --show-vars was specified or not
3958 JJ/02 exipick: Added support for new ACL variable spool format introduced
3961 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3962 syntactically invalid From: or Reply-to: line, and a filter used this to
3963 generate an autoreply, and therefore failed to obtain an address for the
3964 autoreply, Exim could try to deliver to a non-existent relative file
3965 name, causing unrelated and misleading errors. What now happens is that
3966 it logs this as a hard delivery error, but does not attempt to create a
3969 PH/11 The exinext utility has a -C option for testing purposes, but although
3970 the given file was scanned by exinext itself; it wasn't being passed on
3971 when Exim was called.
3973 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3974 an end-of-file indication when reading a command response.
3976 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3977 compiled. In many other places in Exim, IPv6 addresses are always
3978 recognized, so I have changed this. It also means that IPv4 domain
3979 literals of the form [IPV4:n.n.n.n] are now always recognized.
3981 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3982 used if the router is not running as root, for example, when verifying at
3983 ACL time, or when using -bh. The debugging output from this situation was
3984 non-existent - all you got was a failure to exec. I have made two
3987 (a) Failures to set uid/gid, the current directory, or a process leader
3988 in a subprocess such as that created by queryprogram now generate
3989 suitable debugging output when -d is set.
3991 (b) The queryprogram router detects when it is not running as root,
3992 outputs suitable debugging information if -d is set, and then runs
3993 the subprocess without attempting to change uid/gid.
3995 PH/15 Minor change to Makefile for building test_host (undocumented testing
3998 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3999 additional section of a DNS packet that returns MX or SRV records.
4000 Instead, it always explicitly searches for A/AAAA records. This avoids
4001 major problems that occur when a DNS server includes only records of one
4002 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4003 fixed another bug: if SRV records were looked up and the corresponding
4004 address records were *not* found in the additional section, the port
4005 values from the SRV records were lost.
4007 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4008 using the correct key (the original address) when searching the retry
4009 rules in order to find which one to use for generating the retry hint.
4011 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4012 adding the default one. Similarly, if it contains a Reply-To: header, the
4013 errors_reply_to option, if set, is not used.
4015 PH/19 When calculating a retry time, Exim used to measure the "time since
4016 failure" by looking at the "first failed" field in the retry record. Now
4017 it does not use this if it is later than than the arrival time of the
4018 message. Instead it uses the arrival time. This makes for better
4019 behaviour in cases where some deliveries succeed, thus re-setting the
4020 "first failed" field. An example is a quota failure for a huge message
4021 when small messages continue to be delivered. Without this change, the
4022 "time since failure" will always be short, possible causing more frequent
4023 delivery attempts for the huge message than are intended.
4024 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4026 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4027 $1, $2, $3) because the numerical variables can be reset during some
4028 expansion items (e.g. "match"), thereby losing the authentication data.
4030 PH/21 Make -bV show the size of off_t variables so that the test suite can
4031 decide whether to run tests for quotas > 2G.
4033 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4034 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4035 filecount value is greater than 2G or if a quota value is greater than 2G
4036 on a system where the size of off_t is not greater than 4, a panic error
4039 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4040 never match. The debug and -bh output now contains an explicit error
4041 message indicating a malformed IPv4 address or mask.
4043 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4044 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4045 PH/23 above applies.
4047 PH/25 Do not write to syslog when running in the test harness. The only
4048 occasion when this arises is a failure to open the main or panic logs
4049 (for which there is an explicit test).
4051 PH/26 Added the /no_tell option to "control=freeze".
4053 PH/27 If a host name lookup failed very early in a connection, for example, if
4054 the IP address matched host_lookup and the reverse lookup yielded a name
4055 that did not have a forward lookup, an error message of the form "no IP
4056 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4057 could be logged. Now it outputs the IP address instead of "NULL".
4059 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4060 allows the sender and the authenticated sender to be set when
4061 submitting a message from within Exim. Since child_open_exim() is
4062 documented for local_scan(), the new function should be too.
4064 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4065 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4066 results in an empty string is now treated as unset.
4068 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4070 PH/31 Added sender_verify_fail logging option.
4072 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4073 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4074 tidied the source and removed it altogether.
4076 PH/33 When a queue run was abandoned because the load average was too high, a
4077 log line was always written; now it is written only if the queue_run log
4078 selector is set. In addition, the log line for abandonment now contains
4079 information about the queue run such as the pid. This is always present
4080 in "start" and "stop" lines but was omitted from the "abandon" line.
4082 PH/34 Omit spaces between a header name and the colon in the error message that
4083 is given when verify = headers_syntax fails (if there are lots of them,
4084 the message gets confusing).
4086 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4087 names, as there are now some PTR records that contain slashes. This check
4088 is only to protect against broken name servers that fall over on strange
4089 characters, so the fact that it applies to all lookups doesn't matter.
4091 PH/36 Now that the new test suite is complete, we can remove some of the
4092 special code in Exim that was needed for the old test suite. For example,
4093 sorting DNS records because real resolvers return them in an arbitrary
4094 order. The new test suite's fake resolver always returns records in the
4097 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4098 bounces) except when queue_only is set, to avoid logging races between
4099 the different processes.
4101 PH/38 Panic-die if .include specifies a non-absolute path.
4103 PH/39 A tweak to the "H" retry rule from its user.
4105 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4106 a label. They prevented compilation on older perls.
4108 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4109 a warning to be raised on newish perls.
4111 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4112 on queue. Changes to match documented behaviour of showing count of
4113 messages matching specified criteria.
4115 PH/40 Changed the default ident timeout from 30s to 5s.
4117 PH/41 Added support for the use of login_cap features, on those BSD systems
4118 that have them, for controlling the resources used by pipe deliveries.
4120 PH/42 The content-scanning code uses fopen() to create files in which to put
4121 message data. Previously it was not paying any attention to the mode of
4122 the files. Exim runs with umask(0) because the rest of the code creates
4123 files with open(), and sets the required mode explicitly. Thus, these
4124 files were ending up world-writeable. This was not a big issue, because,
4125 being within the spool directory, they were not world-accessible. I have
4126 created a function called modefopen, which takes an additional mode
4127 argument. It sets umask(777), creates the file, chmods it to the required
4128 mode, then resets the umask. All the relevant calls to fopen() in the
4129 content scanning code have been changed to use this function.
4131 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4132 to 24 hours. This avoids potential overflow problems when processing G
4133 and H retry rules. I suspect nobody ever tinkers with this value.
4135 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4137 PH/45 When the plaintext authenticator is running as a client, the server's
4138 challenges are checked to ensure they are valid base64 strings. By
4139 default, the authentication attempt is cancelled if an invalid string is
4140 received. Setting client_ignore_invalid_base64 true ignores these errors.
4141 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4142 they are received. Thus, the responses can be made to depend on the
4143 challenges. If an invalid string is ignored, an empty string is placed in
4146 PH/46 Messages that are created by the autoreply transport now contains a
4147 References: header, in accordance with RFCs 2822 and 3834.
4149 PH/47 Added authenticated_sender_force to the smtp transport.
4151 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4153 PH/49 Installed latest patch from the Sieve maintainer.
4155 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4156 was also set, the appendfile transport was unnecessarily scanning a
4157 directory of message files (e.g. for maildir delivery) to find the count
4158 of files (along with the size), even though it did not need this
4159 information. It now does the scan only if it needs to find either the
4160 size of the count of files.
4162 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4164 PH/52 Two bugs concerned with error handling when the smtp transport is
4167 (i) Exim was not creating retry information for temporary errors given
4168 for individual recipients after the DATA command when the smtp transport
4169 was used in LMTP mode. This meant that they could be retried too
4170 frequently, and not timed out correctly.
4172 (ii) Exim was setting the flag that allows error details to be returned
4173 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4174 recipients that were returned after the DATA command.
4176 PH/53 This is related to PH/52, but is more general: for any failing address,
4177 when detailed error information was permitted to be returned to the
4178 sender, but the error was temporary, then after the final timeout, only
4179 "retry timeout exceeded" was returned. Now it returns the full error as
4180 well as "retry timeout exceeded".
4182 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4183 do this, and (what is worse) MTAs that accept it.
4185 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4186 will now be deprecated.
4188 PH/56 New os.c-cygwin from the Cygwin maintainer.
4190 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4191 formats (previously only available in exim formats via -bpr, -bpru,
4192 and -bpra. Now also available in native and exiqgrep formats)
4194 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4195 with very large, slow to parse queues
4197 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4199 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4201 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4202 responses to authentication challenges, though it was showing the
4203 challenges; (ii) I've removed the CR characters from the debug output for
4206 PH/58 Allow for the insertion of a newline as well as a space when a string
4207 is turned into more than one encoded-word during RFC 2047 encoding. The
4208 Sieve code now uses this.
4210 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4211 data_4xx, lost_connection, tls_required.
4213 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4214 message was being sent as an SMTP response.
4216 PH/61 Add -l and -k options to exicyclog.
4218 PH/62 When verifying, if an address was redirected to one new address, so that
4219 verification continued, and the new address failed or deferred after
4220 having set something in $address_data, the value of $address_data was not
4221 passed back to the ACL. This was different to the case when no
4222 redirection occurred. The value is now passed back in both cases.
4224 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4225 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4226 use login_cap.h, so on its own it isn't the distinguishing feature. The
4227 new name refers directly to the setclassresources() function.
4229 PH/65 Added configuration files for NetBSD3.
4231 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4233 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4234 is preferred over IPv4.
4236 PH/68 The bounce_return_message and bounce_return_body options were not being
4237 honoured for bounces generated during the reception of non-SMTP messages.
4238 In particular, this applied to messages rejected by the ACL. This bug has
4239 been fixed. However, if bounce_return_message is true and bounce_return_
4240 body is false, the headers that are returned for a non-SMTP message
4241 include only those that have been read before the error was detected.
4242 (In the case of an ACL rejection, they have all been read.)
4244 PH/69 The HTML version of the specification is now built in a directory called
4245 spec_html instead of spec.html, because the latter looks like a path with
4246 a MIME-type, and this confuses some software.
4248 PH/70 Catch two compiler warnings in sieve.c.
4250 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4251 function verify_get_ident() calls ip_connect() to connect a socket, but
4252 if the "connect()" function timed out, ip_connect() used to close the
4253 socket. However, verify_get_ident() also closes the socket later, and in
4254 between Exim writes to the log, which may get opened at this point. When
4255 the socket was closed in ip_connect(), the log could get the same file
4256 descriptor number as the socket. This naturally causes chaos. The fix is
4257 not to close the socket in ip_connect(); the socket should be closed by
4258 the function that creates it. There was only one place in the code where
4259 this was missing, in the iplookup router, which I don't think anybody now
4260 uses, but I've fixed it anyway.
4262 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4263 well as to direct DNS lookups. Otherwise the handling of names in host
4264 lists is inconsistent and therefore confusing.
4270 PH/01 Two changes to the default runtime configuration:
4272 (1) Move the checks for relay_from_hosts and authenticated clients from
4273 after to before the (commented out) DNS black list checks.
4275 (2) Add control=submission to the relay_from_hosts and authenticated
4276 clients checks, on the grounds that messages accepted by these
4277 statements are most likely to be submissions.
4279 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4281 (1) Generate an error if the third argument for the ${prvs expansion is
4284 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4287 (3) Reset the variables that are obtained from the first argument of
4288 ${prvscheck and used in the second argument before leaving the code,
4289 because their memory is reclaimed, so using them afterwards may do
4292 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4293 one (it's much easier than Tom thought :-).
4295 (5) Because of (4), we can now allow for the use of $prvscheck_result
4296 inside the third argument.
4298 PH/03 For some reason, the default setting of PATH when running a command from
4299 a pipe transport was just "/usr/bin". I have changed it to
4302 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4303 anything to be listed in the output from -bV.
4305 PH/05 When a filter generated an autoreply, the entire To: header line was
4306 quoted in the delivery log line, like this:
4308 => >A.N.Other <ano@some.domain> <original@ddress> ...
4310 This has been changed so that it extracts the operative address. There
4311 may be more than one such address. If so, they are comma-separated, like
4314 => >ano@some.domain,ona@other.domain <original@ddress> ...
4316 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4317 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4318 not being looked up in the rDNS to get a host name, Exim was showing the
4319 IP address twice in Received: lines, even though the IP addresses were
4320 identical. For example:
4322 Received: from [1.2.3.4] (helo=[1.2.3.4])
4324 However, if the real host name was known, it was omitting the HELO data
4325 if it matched the actual IP address. This has been tidied up so that it
4326 doesn't show the same IP address twice.
4328 PH/07 When both +timestamp and +memory debugging was on, the value given by
4329 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4330 called by the debug printing, thereby overwriting the timestamp buffer.
4331 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4333 PH/08 When the original message was included in an autoreply transport, it
4334 always said "this is a copy of the message, including all the headers",
4335 even if body_only or headers_only was set. It now gives an appropriate
4338 PH/09 Applied a patch from the Sieve maintainer which:
4340 o fixes some comments
4341 o adds the (disabled) notify extension core
4342 o adds some debug output for the result of if/elsif tests
4343 o points to the current vacation draft in the documentation
4344 and documents the missing references header update
4348 o fixes a bug in processing the envelope test (when testing
4349 multiple envelope elements, the last element determined the
4352 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4353 Electronic Mail") by including:
4355 Auto-submitted: auto-generated
4357 in the messages that it generates (bounce messages and others, such as
4358 warnings). In the case of bounce messages for non-SMTP messages, there was
4359 also a typo: it was using "Auto_submitted" (underscore instead of
4360 hyphen). Since every message generated by Exim is necessarily in response
4361 to another message, thes have all been changed to:
4363 Auto-Submitted: auto-replied
4365 in accordance with these statements in the RFC:
4367 The auto-replied keyword:
4369 - SHOULD be used on messages sent in direct response to another
4370 message by an automatic process,
4372 - MUST NOT be used on manually-generated messages,
4374 - MAY be used on Delivery Status Notifications (DSNs) and Message
4375 Disposition Notifications (MDNs),
4377 - MUST NOT be used on messages generated by automatic or periodic
4378 processes, except for messages which are automatic responses to
4381 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4382 to the default Received: header definition.
4384 PH/12 Added log selector acl_warn_skipped (default on).
4386 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4387 variables because (a) they are in the wrong storage pool and (b) even if
4388 they were copied, it wouldn't work properly because of the caching.
4390 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4391 checking when decoding. Apparently there are clients that generate
4392 overlong encoded strings. Why am I not surprised?
4394 PH/15 If the first argument of "${if match_address" was not empty, but did not
4395 contain an "@" character, Exim crashed. Now it writes a panic log message
4396 and treats the condition as false.
4398 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4400 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4401 extension "envelope-auth". The code is finished and in agreement with
4402 other implementations, but there is no documentation so far and in fact,
4403 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4404 not changing the active code.
4406 Print executed "if" and "elsif" statements when debugging is used. This
4407 helps a great deal to understand what a filter does.
4409 Document more things not specified clearly in RFC3028. I had all this
4410 sorted out, when out of a sudden new issues came to my mind. Oops."
4412 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4415 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4416 canonical form (as documented). However, after a host name lookup from
4417 the IP address, check_host() was doing a simple string comparison with
4418 addresses acquired from the DNS when checking that the found name did
4419 have the original IP as one of its addresses. Since any found IPv6
4420 addresses are likely to be in abbreviated form, the comparison could
4421 fail. Luckily, there already exists a function for doing the comparison
4422 by converting both addresses to binary, so now that is used instead of
4423 the text comparison.
4425 PH/20 There was another similar case to PH/19, when a complete host name was
4426 given in a host list; looking up its IP address could give an abbreviated
4427 form, whereas the current host's name might or might not be abbreviated.
4428 The same fix has been applied.
4434 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4435 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4438 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4439 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4441 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4443 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4444 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4445 Also, for "Auto-Submitted", treat anything other than "no" as
4446 non-personal, in accordance with RFC 3834. (Previously it treated
4447 anything starting "auto-" as non-personal.)
4449 TF/01 The control=submission/name=... option had a problem with syntax
4450 errors if the name included a slash character. The /name= option
4451 now slurps the rest of the string, so it can include any characters
4452 but it must come last in the list of options (after /sender_retain
4455 PH/05 Some modifications to the interface to the fake nameserver for the new
4463 TK/01 Added the "success_on_redirect" address verification option. See
4464 NewStuff for rationale and an example.
4466 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4468 PH/02 Patch to exigrep to allow it to work on syslog lines.
4470 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4471 fread() to skip over the body file's header line, because in Cygwin the
4472 header line is locked and is inaccessible.
4474 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4475 co-exist for some time) to make it clear that it is the Exim ID that is
4476 referenced, not the Message-ID: header line.
4478 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4479 string_format() function, because snprintf() does not exist on all
4482 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4483 ${stat: expansion item.
4485 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4486 protocol synchronization error", to keep the pedants happy.
4488 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4489 well as for IRIX systems, when gcc is being used. See the host.c source
4492 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4494 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4497 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4498 transport and to the smtp transport in LMTP mode.
4500 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4502 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4503 run a filter in a subprocess. This could lead to confusion in subsequent
4504 lookups in the parent process. There should also be a search_tidyup() at
4505 the end of the subprocess.
4507 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4508 only if the host matched helo_try_verify_hosts, which caused the
4509 verification to occur when the EHLO/HELO command was issued. The ACL just
4510 tested the remembered result. Now, if a previous verification attempt has
4511 not happened, "verify = helo" does it there and then.
4513 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4515 TK/03 Fix log output including CR from clamd.
4517 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4518 exist provoked a memory error which could cause a segfault.
4520 PH/15 Installed PCRE 6.2
4522 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4524 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4525 of the problem. Specifically, suggested +O2 rather than +O1 for the
4528 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4530 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4531 record, the port was not correctly logged when the outgoing_port log
4532 selector was set (it logged the transort's default port).
4534 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4535 fallback_hosts, and "hosts" in the smtp transport.
4537 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4538 host errors such as "Connection refused".
4540 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4541 authentication with radiusclient 0.4.9:
4543 - Error returned from rc_read_config was caught wrongly
4544 - Username/password not passed on to radius server due to wrong length.
4546 The presumption is that some radiusclient API changes for 4.51/PH/17
4547 were not taken care of correctly. The code is still untested by me (my
4548 Linux distribution still has 0.3.2 of radiusclient), but it was
4549 contributed by a Radius user.
4551 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4552 expanding the "port" option of the smtp transport.
4554 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4555 while reading a MIME header. Thanks to Tom Hughes for a patch.
4557 PH/24 Include config.h inside local_scan.h so that configuration settings are
4560 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4561 that in an ACL for RCPT (for example), you can examine exactly what was
4564 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4565 commands, but it was not correctly comparing the address with the actual
4566 client host address. Thus, it would show the EHLO address in Received:
4567 header lines when this was not necessary.
4569 PH/27 Added the % operator to ${eval:}.
4571 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4572 it should be ignoring failures (because with -C, for example, it has lost
4573 privilege). It wasn't ignoring creation failures other than "already
4576 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4579 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4580 because an input line was too long, either on its own, or by virtue of
4581 too many continuations, the temporary file was not being removed, and the
4582 return code was incorrect.
4584 PH/31 Missing "BOOL" in function definition in filtertest.c.
4586 PH/32 Applied Sieve patches from the maintainer.
4588 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4590 PH/33 Added "verify = not_blind".
4592 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4593 Local/Makefile (with some defaults set). These are used in built scripts
4594 such as exicyclog, but they have never been used in the exim_install
4595 script (though there are many overriding facilities there). I have
4596 arranged that the exim_install script now takes note of these two
4599 PH/35 Installed configuration files for Dragonfly.
4601 PH/36 When a locally submitted message by a trusted user did not contain a
4602 From: header, and the sender address was obtained from -f or from an SMTP
4603 MAIL command, and the trusted user did not use -F to supply a sender
4604 name, $originator_name was incorrectly used when constructing a From:
4605 header. Furthermore, $originator_name was used for submission mode
4606 messages from external hosts without From: headers in a similar way,
4607 which is clearly wrong.
4609 PH/37 Added control=suppress_local_fixups.
4611 PH/38 When log_selector = +received_sender was set, and the addition of the
4612 sender made the log line's construction buffer exactly full, or one byte
4613 less than full, an overflow happened when the terminating "\n" was
4616 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4617 when the result of a list match is failure because a DNS lookup failed.
4619 PH/40 RM_COMMAND is now used in the building process.
4621 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4622 the "build-* directories that it finds.
4624 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4625 address in a domain literal was a prefix of an interface address.
4627 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4628 when verifying a sender address, unless rewrite_headers is false.
4630 PH/44 Wrote a long comment about why errors_to addresses are verified as
4631 recipients, not senders.
4633 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4634 the ratelimit ACL was added.
4636 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4638 PH/46 Added extra information about PostgreSQL errors to the error string.
4640 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4641 avoiding the need to install special zones in a real server. This is
4642 backwards compatible; if it can't find the fake resolver, it drops back.
4643 Thus, both old and new test suites can be run.
4645 TF/02 Added util/ratelimit.pl
4647 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4648 clock is set back in time.
4650 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4651 Candler <B.Candler@pobox.com>.
4653 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4654 OK for addresses that are the result of redirections.
4656 PH/48 A number of further additions for the benefit of the new test suite,
4657 including a fake gethostbyname() that interfaces to the fake DNS resolver
4660 TF/06 The fix for widen_domains has also been applied to qualify_single and
4661 search_parents which are the other dnslookup options that can cause
4664 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4667 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4669 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4670 service" if it encountered an MX record with an empty target hostname.
4671 The message is now "an MX or SRV record indicated no SMTP service".
4673 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4674 if the DNS of the sending site is misconfigured. This is quite a
4675 common situation. This change restores the behaviour of treating a
4676 helo verification defer as a failure.
4678 PH/51 If self=fail was set on a router, the bounce message did not include the
4679 actual error message.
4685 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4687 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4688 command itself ended in error, the underlying message about the transport
4689 filter timeout was being overwritten with the pipe command error. Now the
4690 underlying error message should be appended to the second error message.
4692 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4694 PH/02 Reduce the amount of output that "make" produces by default. Full output
4695 can still be requested.
4697 PH/03 The warning log line about a condition test deferring for a "warn" verb
4698 was being output only once per connection, rather than after each
4699 occurrence (because it was using the same function as for successful
4700 "warn" verbs). This seems wrong, so I have changed it.
4702 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4703 it should not have, which might have caused a crash in the right
4704 circumstances, but probably never did.
4706 PH/04 Installed a modified version of Tony Finch's patch to make submission
4707 mode fix the return path as well as the Sender: header line, and to
4708 add a /name= option so that you can make the user's friendly name appear
4711 TF/03 Added the control = fakedefer ACL modifier.
4713 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4714 Mark Lowes for thorough testing.
4716 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4718 TK/03 Merged latest SRS patch from Miles Wilton.
4720 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4721 with the definition in sysexits.h (which is #included earlier).
4722 Fortunately, Exim does not actually use EX_OK. The code used to try to
4723 preserve the sysexits.h value, by assuming that macro definitions were
4724 scanned for macro replacements. I have been disabused of this notion,
4725 so now the code just undefines EX_OK before #including unistd.h.
4727 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4728 in the smtp transport. When a block could not be written in a single
4729 write() function, the timeout was being re-applied to each part-write.
4730 This seems wrong - if the receiver was accepting one byte at a time it
4731 would take for ever. The timeout is now adjusted when this happens. It
4732 doesn't have to be particularly precise.
4734 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4735 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4737 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4738 without a domain if the check to <postmaster@domain> fails.
4740 SC/01 Eximstats: added -xls and the ability to specify output files
4741 (patch written by Frank Heydlauf).
4743 SC/02 Eximstats: use FileHandles for outputting results.
4745 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4747 SC/04 Eximstats: fixed display of large numbers with -nvr option
4749 SC/05 Eximstats: fixed merging of reports with empty tables.
4751 SC/06 Eximstats: added the -include_original_destination flag
4753 SC/07 Eximstats: removed tabs and trailing whitespace.
4755 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4757 TK/06 MBOX spool code: Add real "From " MBOX separator line
4758 so the .eml file is really in mbox format (even though
4759 most programs do not really care). Patch from Alex Miller.
4761 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4762 The latter is generated from $received_to and is only set if the
4763 message has one envelope recipient. SA can use these headers,
4764 obviously out-of-the-box. Patch from Alex Miller.
4766 PH/08 The ${def test on a variable was returning false if the variable's
4767 value was "0", contrary to what the specification has always said!
4768 The result should be true unless the variable is empty.
4770 PH/09 The syntax error of a character other than { following "${if
4771 def:variable_name" (after optional whitespace) was not being diagnosed.
4772 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4773 accidental colon was present, for example, could give incorrect results.
4775 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4776 result is used (not including appendfile, where other changes are about
4779 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4780 This involved changing a lot of size variables from int to off_t. It
4781 should work with maildirs and everything.
4783 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4784 spamd dying while we are connected to it.
4786 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4789 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4790 function for BDB 4.3.
4792 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4794 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4795 the sources for PCRE have been re-arranged and more files are now
4798 PH/15 The code I had for printing potentially long long variables in PH/11
4799 above was not the best (it lost precision). The length of off_t variables
4800 is now inspected at build time, and an appropriate printing format (%ld
4801 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4802 to be "long long int" or "long int". This is needed for the internal
4803 formatting function string_vformat().
4805 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4806 the configuration file to be ":syslog", then the script "guesses" where
4807 the logs files are, rather than using the compiled in default. In our
4808 case the guess is not the same as the compiled default, so the script
4809 suddenly stopped working when I started to use syslog. The patch checks
4810 to see if log_file_path is "". If so, it attempts to read it from exim
4811 with no configuration file to get the compiled in version, before it
4812 falls back to the previous guessing code."
4814 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4815 implementing BATV in an Exim configuration. See NewStuff for the gory
4818 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4819 Makefile that are specific to HP-UX.
4821 PH/18 If the "use_postmaster" option was set for a recipient callout together
4822 with the "random" option, the postmaster address was used as the MAIL
4823 FROM address for the random test, but not for the subsequent recipient
4824 test. It is now used for both.
4826 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4827 patch removes a few documentation additions to RFC 3028, because the
4828 latest draft now contains them. It adds the new en;ascii-case comparator
4829 and a new error check for 8bit text in MIME parts. Comparator and
4830 require names are now matched exactly. I enabled the subaddress
4831 extension, but it is not well tested yet (read: it works for me)."
4833 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4834 rework some of the code of TK/09 above to avoid the hardwired use of
4835 "%lld" and "long long". Replaced the call to snprintf() with a call to
4838 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4839 records point to non-existent hosts", "retry timeout exceeded", and
4840 "retry time not reached for any host after a long failure period".
4842 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4843 experimental DomainKeys support:
4845 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4846 (2) On an error such as an illegally used "control", the wrong name for
4847 the control was given.
4849 These problems did NOT occur unless DomainKeys support was compiled.
4851 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4853 PH/24 Added ${if match_ip condition.
4855 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4856 databases so that it will be absolutely obvious if a crash occurs in the
4857 DB library. This is a regular occurrence (often caused by mis-matched
4860 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4861 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4862 picked up on a user's system that detects such things. There doesn't seem
4863 to be a gcc warning option for this - only an attribute that has to be
4864 put on the function's prototype. It seems that in Fedora Core 4 they have
4865 set this on a number of new functions. No doubt there will be more in due
4868 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4869 specify a transport. However, if an address that was verified by such a
4870 router was the subject of a callout, Exim crashed because it tried to
4871 read the rcpt_include_affixes from the non-existent transport. Now it
4872 just assumes that the setting of that option is false. This bug was
4873 introduced by 4.51/PH/31.
4875 PH/28 Changed -d+all to exclude +memory, because that information is very
4876 rarely of interest, but it makes the output a lot bigger. People tend to
4877 do -d+all out of habit.
4879 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4880 code in os-type was giving problems when libc.so lives in lib64, like on
4883 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4884 aren't the modern standard, and it seems that some systems' include files
4885 don't always have them. Exim was already checking for some of the newer
4886 ones like T_AAAA, and defining it itself. I've added checks for all the
4887 record types that Exim uses.
4889 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4890 not automatically generating a new one, as it is supposed to. This
4891 prevented TLS from working. If the file did exist, but contained invalid
4892 data, a new version was generated, as expected. It was only the case of a
4893 non-existent file that was broken.
4895 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4896 with a change in libdomainkeys > 0.64.
4898 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4899 from DNS. If the selector record carries the flag, it now has
4900 precedence over the domain-wide flag.
4902 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4904 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4905 the use of a port name that isn't defined in /etc/services) occurred, the
4906 message was deferred as in a normal delivery, and thus remained on the
4907 spool, instead of being failed because of the mua_wrapper setting. This
4908 is now fixed, and I tidied up some of the mua_wrapper messages at the
4911 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4912 the messages in a hash of arrays rather than using individual hashes.
4913 This is a bit cleaner and results in dramatic memory savings, albeit
4914 at a slight CPU cost.
4916 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4917 as requested by Marc Sherman.
4919 SC/10 Eximstats: added histograms for user specified patterns as requested
4922 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4924 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4925 fopen() in the content-scanning modules that did not already have it.
4931 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4932 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4934 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4936 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4938 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4939 to test Sieve filters that use "vacation".
4941 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4942 that changes the way the GnuTLS parameters are stored in the cache file.
4943 The new format can be generated externally. For backward compatibility,
4944 if the data in the cache doesn't make sense, Exim assumes it has read an
4945 old-format file, and it generates new data and writes a new file. This
4946 means that you can't go back to an older release without removing the
4949 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4950 work if there are any delivery delays because "one_time" forces the
4951 parent to be marked "delivered", so its unseen clone is never tried
4952 again. For this reason, Exim now forbids the simultaneous setting of
4955 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4956 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4957 entry if you want to know the details. The fix, however, neglected to
4958 consider the case when local delivery batching is involved. The test for
4959 "previously delivered" was not happening when checking to see if an
4960 address could be batched with a previous (undelivered) one; under
4961 certain circumstances this could lead to multiple deliveries to the same
4964 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4965 in its include files, and this causes problems building Exim.
4967 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4968 header_syntax) but Exim was just ignoring anything given after a slash.
4969 In particular, this caused confusion with an attempt to use "verify =
4970 reverse_host_lookup/defer_ok". An error is now given when options are
4971 supplied for verify items that do not have them. (Maybe reverse_host_
4972 lookup should have a defer_ok option, but that's a different point.)
4974 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4975 defined by RFC 821) to 2048, because there were problems with some AUTH
4976 commands, and RFC 1869 says the size should be increased for extended
4977 SMTP commands that take arguments.
4979 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4982 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4983 "unknown" error; now it says that the functionality isn't in the binary.
4985 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4986 an address' error message when a string expansion fails (syntax or
4987 whatever). Otherwise the password may appear in the log. Following change
4988 PH/42 below, there is no longer a chance of it appearing in a bounce
4991 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4993 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4994 of its addresses was ever tried. (Bugzilla bug #2.)
4996 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4997 the result incorrectly in the debug output. (It correctly added a newline
4998 to what was transported.)
5000 TF/01 Added $received_time.
5002 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5003 commented out examples of how to interface to a virus scanner and to
5004 SpamAssassin. Also added commented examples of av_scanner and
5005 spamd_address settings.
5007 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5008 and controls are allowed in which ACLs. There were a couple of minor
5009 errors. Some of the entries in the conditions table (which is a table of
5010 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5011 negation of where the condition IS allowed.
5013 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5015 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5016 header file does not have a version number, so I've had to invent a new
5017 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5018 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5019 radiusclient), but it was contributed by a Radius user.
5021 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5022 files or directories, for OpenSSL.
5024 PH/19 When an Exim process that is running as root has to create an Exim log
5025 file, it does so in a subprocess that runs as exim:exim so as to get the
5026 ownership right at creation (otherwise, other Exim processes might see
5027 the file with the wrong ownership). There was no test for failure of this
5028 fork() call, which would lead to the process getting stuck as it waited
5029 for a non-existent subprocess. Forks do occasionally fail when resources
5030 run out. I reviewed all the other calls to fork(); they all seem to check
5033 PH/20 When checking for unexpected SMTP input at connect time (before writing
5034 the banner), Exim was not dealing correctly with a non-positive return
5035 from the read() function. If the client had disconnected by this time,
5036 the result was a log entry for a synchronization error with an empty
5037 string after "input=" when read() returned zero. If read() returned -1
5038 (an event I could not check), uninitialized data bytes were printed.
5039 There were reports of junk text (parts of files, etc) appearing after
5042 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5044 PH/22 Added support for macro redefinition, and (re)definition in between
5045 driver and ACL definitions.
5047 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5048 forgetting to use the resulting value; it was using the unexpanded value.
5050 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5051 hadn't been configured. The fix is from Juergen Kreileder, who
5052 understands it better than I do:
5054 "Here's what I see happening with three configured cyrus_sasl
5055 authenticators configured (plain, login, cram-md5):
5057 On startup auth_cyrus_sasl_init() gets called for each of these.
5058 This means three calls to sasl_listmech() without a specified mech_list.
5059 => SASL tests which mechs of all available mechs actually work
5060 => three warnings about OTP not working
5061 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5063 With the patch, sasl_listmech() also gets called three times. But now
5064 SASL's mech_list option is set to the server_mech specified in the the
5065 authenticator. Or in other words, the answer from sasl_listmech()
5066 gets limited to just the mech you're testing for (which is different
5068 => the return list contains just 'plain' or 'login', 'cram-md5' or
5069 nothing depending on the value of ob->server_mech.
5071 I've just tested the patch: Authentication still works fine,
5072 unavailable mechs specified in the exim configuration are still
5073 caught, and the auth.log warnings about OTP are gone."
5075 PH/25 When debugging is enabled, the contents of the command line are added
5076 to the debugging output, even when log_selector=+arguments is not
5079 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5080 answer is "GNU", and only if the return is "GNU/something" is the answer
5083 PH/27 $acl_verify_message is now set immediately after the failure of a
5084 verification in an ACL, and so is available in subsequent modifiers. In
5085 particular, the message can be preserved by coding like this:
5087 warn !verify = sender
5088 set acl_m0 = $acl_verify_message
5090 Previously, $acl_verify_message was set only while expanding "message"
5091 and "log_message" when a very denied access.
5093 PH/28 Modified OS/os.c-Linux with
5095 -#ifndef OS_LOAD_AVERAGE
5096 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5098 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5099 nomenclature these days.)
5101 PH/29 Installed patch from the Sieve maintainer that adds the options
5102 sieve_useraddress and sieve_subaddress to the redirect router.
5104 PH/30 In these circumstances:
5105 . Two addresses routed to the same list of hosts;
5106 . First host does not offer TLS;
5107 . First host accepts first address;
5108 . First host gives temporary error to second address;
5109 . Second host offers TLS and a TLS session is established;
5110 . Second host accepts second address.
5111 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5112 and peerdn, if requested) that were in fact used only for the second
5115 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5116 attention to any local part prefix or suffix that was matched by the
5117 router that accepted the address. It now behaves in the same way as it
5118 does for delivery: the affixes are removed from the local part unless
5119 rcpt_include_affixes is set on the transport.
5121 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5122 timeout during the DATA phase of an incoming message.
5124 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5125 applied a patch sanctioned by the Sieve maintainer.
5127 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5128 the uid or gid is negative. A case of a negative gid caused this to be
5129 noticed. The fix allows for either to be negative.
5131 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5132 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5135 PH/36 The change PH/12 above was broken. Fixed it.
5137 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5138 the grounds that routing the same address twice would always produce the
5139 same answer. This might have been true once, but it is certainly no
5140 longer true now. Routing a child address may depend on the previous
5141 routing that produced that child. Some complicated redirection strategies
5142 went wrong when messages had multiple recipients, and made Exim's
5143 behaviour dependent on the order in which the addresses were given.
5145 I have moved the duplicate checking until after the routing is complete.
5146 Exim scans the addresses that are assigned to local and remote
5147 transports, and removes any duplicates. This means that more work will be
5148 done, as duplicates will always all be routed, but duplicates are
5149 presumably rare, so I don't expect this is of any significance.
5151 For deliveries to pipes, files, and autoreplies, the duplicate checking
5152 still happens during the routing process, since they are not going to be
5155 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5156 It corrects a timeout issue with spamd. This is Ian's comment: "The
5157 background is that sometimes spamd either never reads data from a
5158 connection it has accepted, or it never writes response data. The exiscan
5159 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5160 blindly assumes that writes won't block so it may never time out."
5162 PH/39 Allow G after quota size as well as K and M.
5164 PH/40 The value set for $authenticated_id in an authenticator may not contain
5165 binary zeroes or newlines because the value is written to log lines and
5166 to spool files. There was no check on this. Now the value is run through
5167 the string_printing() function so that such characters are converted to
5168 printable escape sequences.
5170 PH/41 $message_linecount is a new variable that contains the total number of
5171 lines in the message. Compare $body_linecount, which is the count for the
5174 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5175 bounce and delay warning messages, except in certain special cases, which
5178 (a) An SMTP error message from a remote host;
5179 (b) A message specified in a :fail: redirection;
5180 (c) A message specified in a "fail" command in a system filter;
5181 (d) A message specified in a FAIL return from the queryprogram router;
5182 (e) A message specified by the cannot_route_message router option.
5184 In these cases only, Exim does include the error details in bounce and
5185 warning messages. There are also a few cases where bland messages such
5186 as "unrouteable address" or "local delivery error" are given.
5188 PH/43 $value is now also set for the "else" part of a ${run expansion.
5190 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5191 being worked on, but at least Exim now implements the latest version to
5194 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5195 process to complete was treated as a delivery failure, a timeout while
5196 writing the message to the pipe was logged, but erroneously treated as a
5197 successful delivery. Such timeouts include transport filter timeouts. For
5198 consistency with the overall process timeout, these timeouts are now
5199 treated as errors, giving rise to delivery failures by default. However,
5200 there is now a new Boolean option for the pipe transport called
5201 timeout_defer, which, if set TRUE, converts the failures into defers for
5202 both kinds of timeout. A transport filter timeout is now identified in
5205 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5206 systems where "make" and "gmake" are different, calling "gmake" at top
5207 level broke things. I've arranged for the value of $(MAKE) to be passed
5208 from the Makefile to this script so that it can call the same version of
5212 A note about Exim versions 4.44 and 4.50
5213 ----------------------------------------
5215 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5216 changes of various kinds. As a consequence, a big documentation update was
5217 needed. This delayed the release for rather longer than seemed good, especially
5218 in the light of a couple of (minor) security issues. Therefore, the changes
5219 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5220 release. So 4.44 and 4.50 are in effect two different branches that both start
5223 I have left the 4.50 change log unchanged; it contains all the changes since
5224 4.43. The change log for 4.44 is below; many of its items are identical to
5225 those for 4.50. This seems to be the most sensible way to preserve the
5226 historical information.
5232 1. Minor wording change to the doc/README.SIEVE file.
5234 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5235 computation of the current number of files was incorrect.
5237 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5238 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5241 4. Give more explanation in the error message when the command for a transport
5242 filter fails to execute.
5244 5. There are several places where Exim runs a non-Exim command in a
5245 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5246 was being done only for the command run by the queryprogram router. It is
5247 now done for all such subprocesses. The other cases are: ${run, transport
5248 filters, and the commands run by the lmtp and pipe transports.
5250 6. Added CONFIGURE_GROUP build-time option.
5252 7. Some older OS have a limit of 256 on the maximum number of file
5253 descriptors. Exim was using setrlimit() to set 1000 as a large value
5254 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5255 systems. I've change it so that if it can't get 1000, it tries for 256.
5257 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5258 was an oversight, and furthermore, ever since the addition of extra
5259 controls (e.g. 4.43/32), the checks on when to allow different forms of
5260 "control" were broken. There should now be diagnostics for all cases when a
5261 control that does not make sense is encountered.
5263 9. Added the /retain_sender option to "control=submission".
5265 10. $recipients is now available in the predata ACL (oversight).
5267 11. Tidy the search cache before the fork to do a delivery from a message
5268 received from the command line. Otherwise the child will trigger a lookup
5269 failure and thereby defer the delivery if it tries to use (for example) a
5270 cached ldap connection that the parent has called unbind on.
5272 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5273 of $address_data from the recipient verification was clobbered by the
5274 sender verification.
5276 13. The value of address_data from a sender verification is now available in
5277 $sender_address_data in subsequent conditions in the ACL statement.
5279 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5281 15. Added a new option "connect=<time>" to callout options, to set a different
5284 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5285 was its contents. (It was OK if the option was not defined at all.)
5287 17. A "Completed" log line is now written for messages that are removed from
5288 the spool by the -Mrm option.
5290 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5291 information about exactly what failed.
5293 19. Added -dd to debug only the daemon process.
5295 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5296 handles timeouts, both on the server side and network timeouts. Renamed the
5297 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5299 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5300 It is now set to "smtps".
5302 22. $host_address is now set to the target address during the checking of
5303 ignore_target_hosts.
5305 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5306 being passed; this would have caused $sender_host_name to have been used if
5307 matching the list had actually called for a host name (not very likely,
5308 since this list is usually IP addresses). A host name is now passed as
5311 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5312 code that specifies a non-restarting handler (typically sigaction() in
5313 modern systems) in an attempt to fix a rare and obscure crash bug.
5315 25. Narrowed the window for a race in the daemon that could cause it to ignore
5316 SIGCHLD signals. This is not a major problem, because they are used only to
5317 wake it up if nothing else does.
5319 26. A malformed maildirsize file could cause Exim to calculate negative values
5320 for the mailbox size or file count. Odd effects could occur as a result.
5321 The maildirsize information is now recalculated if the size or filecount
5324 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5325 support for a long time. Removed HAVE_SYS_VFS_H.
5327 28. Installed the latest version of exipick from John Jetmore.
5329 29. In an address list, if the pattern was not a regular expression, an empty
5330 subject address (from a bounce message) matched only if the pattern was an
5331 empty string. Non-empty patterns were not even tested. This was the wrong
5332 because it is perfectly reasonable to use an empty address as part of a
5333 database query. An empty address is now tested by patterns that are
5334 lookups. However, all the other forms of pattern expect the subject to
5335 contain a local part and a domain, and therefore, for them, an empty
5336 address still always fails if the pattern is not itself empty.
5338 30. Exim went into a mad DNS loop when attempting to do a callout where the
5339 host was specified on an smtp transport, and looking it up yielded more
5340 than one IP address.
5342 31. Re-factored the code for checking spool and log partition space into a
5343 function that finds that data and another that does the check. The former
5344 is then used to implement four new variables: $spool_space, $log_space,
5345 $spool_inodes, and $log_inodes.
5347 32. The RFC2047 encoding function was originally intended for short strings
5348 such as real names; it was not keeping to the 75-character limit for
5349 encoded words that the RFC imposes. It now respects the limit, and
5350 generates multiple encoded words if necessary. To be on the safe side, I
5351 have increased the buffer size for the ${rfc2047: expansion operator from
5354 33. It is now permitted to omit both strings after an "if" condition; if the
5355 condition is true, the result is "true". As before, when the second string
5356 is omitted, a false condition yields an empty string. This makes it less
5357 cumbersome to write custom ACL and router conditions.
5359 34. Failure to deliver a bounce message always caused it to be frozen, even if
5360 there was an errors_to setting on the router. The errors_to setting is now
5363 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5364 canonical form (fully expanded) before being placed in
5365 $sender_host_address.
5367 36. The table in the code that translates DNS record types into text (T_A to
5368 "A" for instance) was missing entries for NS and CNAME. It is just possible
5369 that this could have caused confusion if both these types were looked up
5370 for the same domain, because the text type is used as part of Exim's
5371 per-process caching. But the chance of anyone hitting this buglet seems
5374 37. The dnsdb lookup has been extended in a number of ways.
5376 (1) There is a new type, "zns", which walks up the domain tree until it
5377 finds some nameserver records. It should be used with care.
5379 (2) There is a new type, "mxh", which is like "mx" except that it returns
5380 just the host names, not the priorities.
5382 (3) It is now possible to give a list of domains (or IP addresses) to be
5383 looked up. The behaviour when one of the lookups defers can be
5384 controlled by a keyword.
5386 (4) It is now possible to specify the separator character for use when
5387 multiple records are returned.
5389 38. The dnslists ACL condition has been extended: it is now possible to supply
5390 a list of IP addresses and/or domains to be looked up in a particular DNS
5393 39. Added log_selector=+queue_time_overall.
5395 40. When running the queue in the test harness, wait just a tad after forking a
5396 delivery process, to get repeatability of debugging output.
5398 41. Include certificate and key file names in error message when GnuTLS fails
5399 to set them up, because the GnuTLS error message doesn't include the name
5400 of the failing file when there is a problem reading it.
5402 42. Allow both -bf and -bF in the same test run.
5404 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5406 44. The "Exiscan patch" is now merged into the mainline Exim source.
5408 45. Sometimes the final signoff response after QUIT could fail to get
5409 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5410 < 0 before doing a fflush(). This bug looks as though it goes back to the
5411 introduction of TLS in release 3.20, but "sometimes" must have been rare
5412 because the tests only now provoked it.
5414 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5415 (this can affect the format of dates).
5417 47. exim_tidydb, when checking for the continued existence of a message for
5418 which it has found a message-specific retry record, was not finding
5419 messages that were in split spool directories. Consequently, it was
5420 deleting retry records that should have stayed in existence.
5422 48. Steve fixed some bugs in eximstats.
5424 49. The SPA authentication driver was not abandoning authentication and moving
5425 on to the next authenticator when an expansion was forced to fail,
5426 contradicting the general specification for all authenticators. Instead it
5427 was generating a temporary error. It now behaves as specified.
5429 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5430 (the order specifies the preference for clients). The order is now AES256,
5431 AES128, 3DES, ARCFOUR128.
5433 51. Small patch to Sieve code - explicitly set From: when generating an
5436 52. Exim crashed if a remote delivery caused a very long error message to be
5437 recorded - for instance if somebody sent an entire SpamAssassin report back
5438 as a large number of 550 error lines. This bug was coincidentally fixed by
5439 increasing the size of one of Exim's internal buffers (big_buffer) that
5440 happened as part of the Exiscan merge. However, to be on the safe side, I
5441 have made the code more robust (and fixed the comments that describe what
5444 53. Now that there can be additional text after "Completed" in log lines (if
5445 the queue_time_overall log selector is set), a one-byte patch to exigrep
5446 was needed to allow it to recognize "Completed" as not the last thing in
5449 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5450 patch that reportedly fixes this has been added. I am not expert enough to
5451 create a test for it. This is what the patch creator wrote:
5453 "I found a little strange behaviour of ldap code when working with
5454 Windows 2003 AD Domain, where users was placed in more than one
5455 Organization Units. When I tried to give exim partial DN, the exit code
5456 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5457 But simultaneously result of request was absolutely normal ldap result,
5458 so I produce this patch..."
5460 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5461 REFERENCE, so I have modified the code to exclude the patch when that macro
5464 55. Some experimental protocols are using DNS PTR records for new purposes. The
5465 keys for these records are domain names, not reversed IP addresses. The
5466 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5467 leaves it alone. Component reversal etc. now happens only for IP addresses.
5470 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5472 57. Double the size of the debug message buffer (to 2048) so that more of very
5473 long debug lines gets shown.
5475 58. The exicyclog utility now does better if the number of log files to keep
5476 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5478 59. Two changes related to the smtp_active_hostname option:
5480 (1) $smtp_active_hostname is now available as a variable.
5481 (2) The default for smtp_banner uses $smtp_active_hostname instead
5482 of $primary_hostname.
5484 60. The host_aton() function is supposed to be passed a string that is known
5485 to be a valid IP address. However, in the case of IPv6 addresses, it was
5486 not checking this. This is a hostage to fortune. Exim now panics and dies
5487 if the condition is not met. A case was found where this could be provoked
5488 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5489 components; fortuitously, this particular loophole had already been fixed
5490 by change 4.50/55 above.
5492 If there are any other similar loopholes, the new check in host_aton()
5493 itself should stop them being exploited. The report I received stated that
5494 data on the command line could provoke the exploit when Exim was running as
5495 exim, but did not say which command line option was involved. All I could
5496 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5497 running as the user.
5500 61. There was a buffer overflow vulnerability in the SPA authentication code
5501 (which came originally from the Samba project). I have added a test to the
5502 spa_base64_to_bits() function which I hope fixes it.
5505 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5506 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5508 63. The daemon start-up calls getloadavg() while still root for those OS that
5509 need the first call to be done as root, but it missed one case: when
5510 deliver_queue_load_max is set with deliver_drop_privilege. This is
5511 necessary for the benefit of the queue runner, because there is no re-exec
5512 when deliver_drop_privilege is set.
5514 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5515 This has been fixed.
5517 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5518 was in use, was not putting the data itself into the right store pool;
5519 consequently, it could be overwritten for a subsequent message in the same
5520 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5523 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5525 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5526 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5527 the function were treating the return as a boolean value, which happened to
5528 work because 0=false and not-0=true, but is not correct code.
5530 68. The host_aton() function was not handling scoped IPv6 addresses (those
5531 with, for example, "%eth0" on the end) correctly.
5533 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5534 negated items (that is, ~something) in unsigned ints. Some compilers
5535 apparently mutter when there is no cast.
5537 70. If an address verification called from an ACL failed, and did not produce a
5538 user-specific message (i.e. there was only a "system" message), nothing was
5539 put in $acl_verify_message. In this situation, it now puts the system
5542 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5543 session; change 4.31/43 added the unwanted input to the log line - except
5544 that it did not do this in the start of session case. It now does.
5546 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5547 This is wrong and can cause the other end to generate a synchronization
5548 error if it is another Exim or anything else that does the synchronization
5549 check. A QUIT command is no longer sent after a timeout.
5551 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5552 during host lookups.
5554 74. The defer_ok option of callout verification was not working if it was used
5555 when verifying addresses in header lines, that is, for this case:
5557 verify = header_sender/callout=defer_ok
5559 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5560 those file descriptors could be used for SMTP connections. If anything
5561 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5562 could be sent to the SMTP client, causing chaos. The daemon now opens
5563 stdin, stdout, and stderr to /dev/null when it puts itself into the
5566 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5567 log by default. The user can override this with suitable Perl magic.
5569 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5570 the log message when discard triggers, was not working for the DATA ACL or
5571 for the non-SMTP ACL.
5573 78. Error message wording change in sieve.c.
5575 79. If smtp_accept_max_per_host was set, the number of connections could be
5576 restricted to fewer than expected, because the daemon was trying to set up
5577 a new connection before checking whether the processes handling previous
5578 connections had finished. The check for completed processes is now done
5579 earlier. On busy systems, this bug wouldn't be noticed because something
5580 else would have woken the daemon, and it would have reaped the completed
5583 80. If a message was submitted locally by a user whose login name contained one
5584 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5585 It caused a spool format error. I have fixed the spool reading code. A
5586 related problem was that the "from" clause in the Received: line became
5587 illegal because of the space(s). It is now covered by ${quote_local_part.
5589 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5592 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5594 83. Patch from Sieve maintainer to fix unterminated string problem in
5595 "vacation" handling.
5597 84. Some minor changes to the Linux configuration files to help with other
5598 OS variants using glibc.
5600 85. One more patch for Sieve to update vacation handling to latest spec.
5603 ----------------------------------------------------
5604 See the note above about the 4.44 and 4.50 releases.
5605 ----------------------------------------------------
5611 1. Change 4.43/35 introduced a bug that caused file counts to be
5612 incorrectly computed when quota_filecount was set in an appendfile
5615 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5616 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5619 3. Give more explanation in the error message when the command for a transport
5620 filter fails to execute.
5622 4. There are several places where Exim runs a non-Exim command in a
5623 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5624 was being done only for the command run by the queryprogram router. It is
5625 now done for all such subprocesses. The other cases are: ${run, transport
5626 filters, and the commands run by the lmtp and pipe transports.
5628 5. Some older OS have a limit of 256 on the maximum number of file
5629 descriptors. Exim was using setrlimit() to set 1000 as a large value
5630 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5631 systems. I've change it so that if it can't get 1000, it tries for 256.
5633 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5634 was an oversight, and furthermore, ever since the addition of extra
5635 controls (e.g. 4.43/32), the checks on when to allow different forms of
5636 "control" were broken. There should now be diagnostics for all cases when a
5637 control that does not make sense is encountered.
5639 7. $recipients is now available in the predata ACL (oversight).
5641 8. Tidy the search cache before the fork to do a delivery from a message
5642 received from the command line. Otherwise the child will trigger a lookup
5643 failure and thereby defer the delivery if it tries to use (for example) a
5644 cached ldap connection that the parent has called unbind on.
5646 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5647 of $address_data from the recipient verification was clobbered by the
5648 sender verification.
5650 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5651 was its contents. (It was OK if the option was not defined at all.)
5653 11. A "Completed" log line is now written for messages that are removed from
5654 the spool by the -Mrm option.
5656 12. $host_address is now set to the target address during the checking of
5657 ignore_target_hosts.
5659 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5660 being passed; this would have caused $sender_host_name to have been used if
5661 matching the list had actually called for a host name (not very likely,
5662 since this list is usually IP addresses). A host name is now passed as
5665 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5666 code that specifies a non-restarting handler (typically sigaction() in
5667 modern systems) in an attempt to fix a rare and obscure crash bug.
5669 15. Narrowed the window for a race in the daemon that could cause it to ignore
5670 SIGCHLD signals. This is not a major problem, because they are used only to
5671 wake it up if nothing else does.
5673 16. A malformed maildirsize file could cause Exim to calculate negative values
5674 for the mailbox size or file count. Odd effects could occur as a result.
5675 The maildirsize information is now recalculated if the size or filecount
5678 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5679 support for a long time. Removed HAVE_SYS_VFS_H.
5681 18. Updated exipick to current release from John Jetmore.
5683 19. Allow an empty sender to be matched against a lookup in an address list.
5684 Previously the only cases considered were a regular expression, or an
5687 20. Exim went into a mad DNS lookup loop when doing a callout where the
5688 host was specified on the transport, if the DNS lookup yielded more than
5691 21. The RFC2047 encoding function was originally intended for short strings
5692 such as real names; it was not keeping to the 75-character limit for
5693 encoded words that the RFC imposes. It now respects the limit, and
5694 generates multiple encoded words if necessary. To be on the safe side, I
5695 have increased the buffer size for the ${rfc2047: expansion operator from
5698 22. Failure to deliver a bounce message always caused it to be frozen, even if
5699 there was an errors_to setting on the router. The errors_to setting is now
5702 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5703 canonical form (fully expanded) before being placed in
5704 $sender_host_address.
5706 24. Updated eximstats to version 1.33
5708 25. Include certificate and key file names in error message when GnuTLS fails
5709 to set them up, because the GnuTLS error message doesn't include the name
5710 of the failing file when there is a problem reading it.
5712 26. Expand error message when OpenSSL has problems setting up cert/key files.
5715 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5716 (this can affect the format of dates).
5718 28. exim_tidydb, when checking for the continued existence of a message for
5719 which it has found a message-specific retry record, was not finding
5720 messages that were in split spool directories. Consequently, it was
5721 deleting retry records that should have stayed in existence.
5723 29. eximstats updated to version 1.35
5724 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5725 1.35 - bugfix such that pie charts by volume are generated correctly
5727 30. The SPA authentication driver was not abandoning authentication and moving
5728 on to the next authenticator when an expansion was forced to fail,
5729 contradicting the general specification for all authenticators. Instead it
5730 was generating a temporary error. It now behaves as specified.
5732 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5733 (the order specifies the preference for clients). The order is now AES256,
5734 AES128, 3DES, ARCFOUR128.
5736 31. Small patch to Sieve code - explicitly set From: when generating an
5739 32. Exim crashed if a remote delivery caused a very long error message to be
5740 recorded - for instance if somebody sent an entire SpamAssassin report back
5741 as a large number of 550 error lines. This bug was coincidentally fixed by
5742 increasing the size of one of Exim's internal buffers (big_buffer) that
5743 happened as part of the Exiscan merge. However, to be on the safe side, I
5744 have made the code more robust (and fixed the comments that describe what
5747 33. Some experimental protocols are using DNS PTR records for new purposes. The
5748 keys for these records are domain names, not reversed IP addresses. The
5749 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5750 leaves it alone. Component reversal etc. now happens only for IP addresses.
5753 34. The host_aton() function is supposed to be passed a string that is known
5754 to be a valid IP address. However, in the case of IPv6 addresses, it was
5755 not checking this. This is a hostage to fortune. Exim now panics and dies
5756 if the condition is not met. A case was found where this could be provoked
5757 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5758 components; fortuitously, this particular loophole had already been fixed
5759 by change 4.50/55 or 4.44/33 above.
5761 If there are any other similar loopholes, the new check in host_aton()
5762 itself should stop them being exploited. The report I received stated that
5763 data on the command line could provoke the exploit when Exim was running as
5764 exim, but did not say which command line option was involved. All I could
5765 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5766 running as the user.
5769 35. There was a buffer overflow vulnerability in the SPA authentication code
5770 (which came originally from the Samba project). I have added a test to the
5771 spa_base64_to_bits() function which I hope fixes it.
5774 36. The daemon start-up calls getloadavg() while still root for those OS that
5775 need the first call to be done as root, but it missed one case: when
5776 deliver_queue_load_max is set with deliver_drop_privilege. This is
5777 necessary for the benefit of the queue runner, because there is no re-exec
5778 when deliver_drop_privilege is set.
5780 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5781 was in use, was not putting the data itself into the right store pool;
5782 consequently, it could be overwritten for a subsequent message in the same
5783 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5786 38. Sometimes the final signoff response after QUIT could fail to get
5787 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5788 < 0 before doing a fflush(). This bug looks as though it goes back to the
5789 introduction of TLS in release 3.20, but "sometimes" must have been rare
5790 because the tests only now provoked it.
5796 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5797 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5798 Now it returns an integer. A number of calls were still expecting a T/F
5799 return. Fortuitously, in all cases, the tests worked in OK situations,
5800 which is the norm. However, things would have gone wrong on any write
5801 failures on the smtp file descriptor. This function is used when sending
5802 messages over SMTP and also when doing verify callouts.
5804 2. When Exim is called to do synchronous delivery of a locally submitted
5805 message (the -odf or -odi options), it no longer closes stderr before doing
5808 3. Implemented the mua_wrapper option.
5810 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5812 5. Implemented the functions header_remove(), header_testname(),
5813 header_add_at_position(), and receive_remove_recipient(), and exported them
5816 6. If an ACL "warn" statement specified the addition of headers, Exim already
5817 inserted X-ACL-Warn: at the start if there was no header name. However, it
5818 was not making this test for the second and subsequent header lines if
5819 there were newlines in the string. This meant that an invalid header could
5820 be inserted if Exim was badly configured.
5822 7. Allow an ACL "warn" statement to add header lines at the start or after all
5823 the Received: headers, as well as at the end.
5825 8. Added the rcpt_4xx retry error code.
5827 9. Added postmaster_mailfrom=xxx to callout verification option.
5829 10. Added mailfrom=xxxx to the callout verification option, for verify=
5832 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5833 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5835 12. Inserted some casts to stop certain compilers warning when using pointer
5836 differences as field lengths or precisions in printf-type calls (mostly
5837 affecting debugging statements).
5839 13. Added optional readline() support for -be (dynamically loaded).
5841 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5842 same clock tick as a message's arrival, so that its received time was the
5843 same as the "first fail" time on the retry record, and that message
5844 remained on the queue past the ultimate address timeout, every queue runner
5845 would try a delivery (because it was past the ultimate address timeout) but
5846 after another failure, the ultimate address timeout, which should have then
5847 bounced the address, did not kick in. This was a "< instead of <=" error;
5848 in most cases the first failure would have been in the next clock tick
5849 after the received time, and all would be well.
5851 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5852 being recognized when the domain list was tested by the match_domain
5853 condition in an expansion string.
5855 16. Added the ${str2b64: operator.
5857 17. Exim was always calling setrlimit() to set a large limit for the number of
5858 processes, without checking whether the existing limit was already
5859 adequate. (It did check for the limit on file descriptors.) Furthermore,
5860 errors from getrlimit() and setrlimit() were being ignored. Now they are
5861 logged to the main and panic logs, but Exim does carry on, to try to do its
5862 job under whatever limits there are.
5864 18. Imported PCRE 5.0.
5866 19. Trivial typo in log message " temporarily refused connection" (the leading
5869 20. If the log selector return_path_on_delivery was set and an address was
5870 redirected to /dev/null, the delivery process crashed because it assumed
5871 that a return path would always be set for a "successful" delivery. In this
5872 case, the whole delivery is bypassed as an optimization, and therefore no
5875 21. Internal re-arrangement: the function for sending a challenge and reading
5876 a response while authentication was assuming a zero-terminated challenge
5877 string. It's now changed to take a pointer and a length, to allow for
5878 binary data in such strings.
5880 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5882 23. Exim was not respecting finduser_retries when seeking the login of the
5883 uid under which it was called; it was always trying 10 times. (The default
5884 setting of finduser_retries is zero.) Also, it was sleeping after the final
5885 failure, which is pointless.
5887 24. Implemented tls_on_connect_ports.
5889 25. Implemented acl_smtp_predata.
5891 26. If the domain in control=submission is set empty, Exim assumes that the
5892 authenticated id is a complete email address when it generates From: or
5893 Sender: header lines.
5895 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5896 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5897 chown and chgrp in /bin and hostname in /usr/bin.
5899 28. Exim was keeping the "process log" file open after each use, just as it
5900 does for the main log. This opens the possibility of it remaining open for
5901 long periods when the USR1 signal hits a daemon. Occasional processlog
5902 errors were reported, that could have been caused by this. Anyway, it seems
5903 much more sensible not to leave this file open at all, so that is what now
5906 29. The long-running daemon process does not normally write to the log once it
5907 has entered its main loop, and it closes the log before doing so. This is
5908 so that log files can straightforwardly be renamed and moved. However,
5909 there are a couple of unusual error situations where the daemon does write
5910 log entries, and I had neglected to close the log afterwards.
5912 30. The text of an SMTP error response that was received during a remote
5913 delivery was being truncated at 512 bytes. This is too short for some of
5914 the long messages that one sometimes sees. I've increased the limit to
5917 31. It is now possible to make retry rules that apply only when a message has a
5918 specific sender, in particular, an empty sender.
5920 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5921 it possible to be selective about when SMTP synchronization is enforced.
5923 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5925 32. Implemented hosts_connection_nolog.
5927 33. Added an ACL for QUIT.
5929 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5932 35. Added mailbox_size and mailbox_filecount to appendfile.
5934 36. Added control = no_multiline_responses to ACLs.
5936 37. There was a bug in the logic of the code that waits for the clock to tick
5937 in the case where the clock went backwards by a substantial amount such
5938 that the microsecond fraction of "now" was more than the microsecond
5939 fraction of "then" (but the whole seconds number was less).
5941 38. Added support for the libradius Radius client library this is found on
5942 FreeBSD (previously only the radiusclient library was supported).
5948 1. When certain lookups returned multiple values in the form name=value, the
5949 quoting of the values was not always being done properly. Specifically:
5950 (a) If the value started with a double quote, but contained no whitespace,
5952 (b) If the value contained whitespace other than a space character (i.e.
5953 tabs or newlines or carriage returns) it was not quoted.
5954 This fix has been applied to the mysql and pgsql lookups by writing a
5955 separate quoting function and calling it from the lookup code. The fix
5956 should probably also be applied to nisplus, ibase and oracle lookups, but
5957 since I cannot test any of those, I have not disturbed their existing code.
5959 2. A hit in the callout cache for a specific address caused a log line with no
5960 reason for rejecting RCPT. Now it says "Previous (cached) callout
5961 verification failure".
5963 3. There was an off-by-one bug in the queryprogram router. An over-long
5964 return line was truncated at 256 instead of 255 characters, thereby
5965 overflowing its buffer with the terminating zero. As well as fixing this, I
5966 have increased the buffer size to 1024 (and made a note to document this).
5968 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5969 when Exim is waiting for an SMTP response from a remote server, Exim
5970 restarts its select() call on the socket, thereby resetting its timeout.
5971 This is not a problem when such interrupts are rare. Somebody set up a cron
5972 job to run exiwhat every 2 minutes, which is less than the normal select()
5973 timeout (5 or 10 minutes). This meant that the select() timeout never
5974 kicked in because it was always reset. I have fixed this by comparing the
5975 time when an interrupt arrives with the time at the start of the first call
5976 to select(). If more time than the timeout has elapsed, the interrupt is
5977 treated as a timeout.
5979 5. Some internal re-factoring in preparation for the addition of Sieve
5980 extensions (by MH). In particular, the "personal" test is moved to a
5981 separate function, and given an option for scanning Cc: and Bcc: (which is
5982 not set for Exim filters).
5984 6. When Exim created an email address using the login of the caller as the
5985 local part (e.g. when creating a From: or Sender: header line), it was not
5986 quoting the local part when it contained special characters such as @.
5988 7. Installed new OpenBSD configuration files.
5990 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5991 try to make them clearer.
5993 9. Callout options, other than the timeout value, were being ignored when
5994 verifying sender addresses in header lines. For example, when using
5996 verify = header_sender/callout=no_cache
5998 the cache was (incorrectly) being used.
6000 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6001 only the Cygwin environment.
6003 11. When return_path_on_delivery was set as a log selector, if different remote
6004 addresses in the same message used different return paths and parallel
6005 remote delivery occurred, the wrong values would sometimes be logged.
6006 (Whenever a remote delivery process finished, the return path value from
6007 the most recently started remote delivery process was logged.)
6009 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6010 lines when AUTH and/or TLS are in use. This is the "received protocol"
6011 field. Exim used to use "asmtp" for authenticated SMTP, without any
6012 indication (in the protocol name) for TLS use. Now it follows the RFC and
6013 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6014 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6015 names appear in log lines as well as in Received: header lines.
6017 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6018 extensions, and comparison tests, and to fix some bugs.
6020 14. Changes to the "personal" filter test:
6022 (1) The test was buggy in that it was just doing the equivalent of
6023 "contains" tests on header lines. For example, if a user's address was
6024 anne@some.where, the "personal" test would incorrectly be true for
6026 To: susanne@some.where
6028 This test is now done by extracting each address from the header in turn,
6029 and checking the entire address. Other tests that are part of "personal"
6030 are now done using regular expressions (for example, to check local parts
6031 of addresses in From: header lines).
6033 (2) The list of non-personal local parts in From: addresses has been
6034 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6035 taken from the Sieve specification recommendations.
6037 (3) If the message contains any header line starting with "List-" it is
6038 treated as non-personal.
6040 (4) The test for "circular" in the Subject: header line has been removed
6041 because it now seems ill-conceived.
6043 15. Minor typos in src/EDITME comments corrected.
6045 16. Installed latest exipick from John Jetmore.
6047 17. If headers_add on a router specified a text string that was too long for
6048 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6049 of string_sprintf() is now avoided.
6051 18. $message_body_size was not set (it was always zero) when running the DATA
6052 ACL and the local_scan() function.
6054 19. For the "mail" command in an Exim filter, no default was being set for
6055 the once_repeat time, causing a random time value to be used if "once" was
6056 specified. (If the value happened to be <= 0, no repeat happened.) The
6057 default is now 0s, meaning "never repeat". The "vacation" command was OK
6058 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6059 (I found it when inspecting the code).
6061 20. There is now an overall timeout for performing a callout verification. It
6062 defaults to 4 times the callout timeout, which applies to individual SMTP
6063 commands during the callout. The overall timeout applies when there is more
6064 than one host that can be tried. The timeout is checked before trying the
6065 next host. This prevents very long delays if there are a large number of
6066 hosts and all are timing out (e.g. when the network connections are timing
6067 out). The value of the overall timeout can be changed by specifying an
6068 additional sub-option for "callout", called "maxwait". For example:
6070 verify = sender/callout=5s,maxwait=20s
6072 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6073 to the end before writing, but this should make it even safer).
6075 22. Exim was forgetting that it had advertised PIPELINING for the second and
6076 subsequent messages on an SMTP connection. It was also not resetting its
6077 memory on STARTTLS and an internal HELO.
6079 23. When Exim logs an SMTP synchronization error within a session, it now
6080 records whether PIPELINING has been advertised or not.
6082 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6083 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6084 rather than long int.
6086 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6088 26. Added the never_mail option to autoreply.
6094 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6095 crash if the getsockname() call failed; this can happen if a connection is
6096 closed very soon after it is established. The problem was simply in the
6097 order in which certain operations were done, causing Exim to try to write
6098 to the SMTP stream before it had set up the file descriptor. The bug has
6099 been fixed by making things happen in the correct order.
6105 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6106 before the connection was closed, thus losing the rejection response.
6108 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6109 some early Solaris releases, but causes trouble in current releases where
6110 socklen_t is defined.
6112 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6115 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6118 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6119 that releasing the top of it at the end releases what was used for sub-
6120 expansions (unless the block got too big). However, discard this block if
6121 the first thing is a variable or header, so that we can use its block when
6122 it is dynamic (useful for very large $message_headers, for example).
6124 6. Lookups now cache *every* query, not just the most recent. A new, separate
6125 store pool is used for this. It can be recovered when all lookup caches are
6126 flushed. Lookups now release memory at the end of their result strings.
6127 This has involved some general refactoring of the lookup sources.
6129 7. Some code has been added to the store_xxx() functions to reduce the amount
6130 of flapping under certain conditions.
6132 8. log_incoming_interface used to affect only the <= reception log lines. Now
6133 it causes the local interface and port to be added to several more SMTP log
6134 lines, for example "SMTP connection from", and rejection lines.
6136 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6138 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6140 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6141 could be overwritten at the end of the current message (or the start of a
6142 new message if it was set in a HELO ACL). The value is now preserved for
6143 the duration of the SMTP connection.
6145 12. If a transport had a headers_rewrite setting, and a matching header line
6146 contained an unqualified address, that address was qualified, even if it
6147 did not match any rewriting rules. The underlying bug was that the values
6148 of the flags that permit the existence of unqualified sender and recipient
6149 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6150 non-local messages, and by -bnq for local messages) were not being
6151 preserved with the message after it was received.
6153 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6154 "next input=" as part of the text comprising the host identity instead of
6155 the correct text. The code was using the same buffer for two different
6156 strings. However, depending on which order the printing function evaluated
6157 its arguments, the bug did not always show up. Under Linux, for example, my
6158 test suite worked just fine.
6160 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6161 doesn't work with some older versions of Perl. It has been changed to "my",
6162 which in any case is probably the better facility to use.
6164 15. A really picky compiler found some instances of statements for creating
6165 error messages that either had too many or two few arguments for the format
6168 16. The size of the buffer for calls to the DNS resolver has been increased
6169 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6170 for addresses that have a lot of PTR records. This alleviates a problem; it
6171 does not fully solve it.
6173 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6174 buffer now truncates the list and logs the incident, which is the same
6175 action as happens when Exim is looking up a host name and its aliases.
6176 Previously in this situation something unpredictable would happen;
6177 sometimes it was "internal error: store_reset failed".
6179 18. If a server dropped the connection unexpectedly when an Exim client was
6180 using GnuTLS and trying to read a response, the client delivery process
6181 crashed while trying to generate an error log message.
6183 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6184 string, for example:
6186 warn message = H1: something\nH2: something
6188 the text was added as a single header line from Exim's point of view
6189 though it ended up OK in the delivered message. However, searching for the
6190 second and subsequent header lines using $h_h2: did not work. This has been
6191 fixed. Similarly, if a system filter added multiple headers in this way,
6192 the routers could not see them.
6194 20. Expanded the error message when iplsearch is called with an invalid key to
6195 suggest using net-iplsearch in a host list.
6197 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6198 ACLs are no longer actually imposed (and a message to that effect is
6201 22. If a "gecos" field in a passwd entry contained escaped characters, in
6202 particular, if it contained a \" sequence, Exim got it wrong when building
6203 a From: or a Sender: header from that name. A second bug also caused
6204 incorrect handling when an unquoted " was present following a character
6205 that needed quoting.
6207 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6208 was not being matched caselessly.
6210 24. Arranged for all hyphens in the exim.8 source to be escaped with
6213 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6214 properly. Recipient callout cache records were still being keyed to include
6215 the sender, even when use_sender was set false. This led to far more
6216 callouts that were necessary. The sender is no longer included in the key
6217 when use_sender is false.
6219 26. Added "control = submission" modifier to ACLs.
6221 27. Added the ${base62d: operator to decode base 62 numbers.
6223 28. dnsdb lookups can now access SRV records.
6225 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6226 the configuration file.
6228 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6229 (-v) mode. This makes the output for a verbose queue run more intelligible.
6231 31. Added a use_postmaster feature to recipient callouts.
6233 32. Added the $body_zerocount variable, containing the number of binary zero
6234 bytes in the message body.
6236 33. The time of last modification of the "new" subdirectory is now used as the
6237 "mailbox time last read" when there is a quota error for a maildir
6240 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6242 35. Added +ignore_unknown as a special item in host lists.
6244 36. Code for decoding IPv6 addresses in host lists is now included, even if
6245 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6246 address was recognized as an IP address, but was then not correctly decoded
6247 into binary, causing unexpected and incorrect effects when compared with
6254 1. Very minor rewording of debugging text in manualroute to say "list of
6255 hosts" instead of "hostlist".
6257 2. If verify=header_syntax was set, and a header line with an unqualified
6258 address (no domain) and a large number of spaces between the end of the
6259 name and the colon was received, the reception process suffered a buffer
6260 overflow, and (when I tested it) crashed. This was caused by some obsolete
6261 code that should have been removed. The fix is to remove it!
6263 3. When running in the test harness, delay a bit after writing a bounce
6264 message to get a bit more predictability in the log output.
6266 4. Added a call to search_tidyup() just before forking a reception process. In
6267 theory, someone could use a lookup in the expansion of smtp_accept_max_
6268 per_host which, without the tidyup, could leave open a database connection.
6270 5. Added the variables $recipient_data and $sender_data which get set from a
6271 lookup success in an ACL "recipients" or "senders" condition, or a router
6272 "senders" option, similar to $domain_data and $local_part_data.
6274 6. Moved the writing of debug_print from before to after the "senders" test
6277 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6278 problems for message scanning, either using a data ACL, or using
6279 local_scan() because the Received: header was not generated till after they
6280 were called (in order to set the time as the time of reception completion).
6281 I have revised the way this works. The header is now generated after the
6282 body is received, but before the ACL or local_scan() are called. After they
6283 are run, the timestamp in the header is updated.
6289 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6290 before starting a queue runner without re-exec. This happened only when
6291 deliver_drop_privilege was set or when the Exim user was set to root. The
6292 effect of the bug was that timeouts during subsequent deliveries caused
6293 crashes instead of being properly handled. The handler is now left at its
6294 default (and expected) setting.
6296 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6297 message, again when deliver_drop_privilege is set or Exim is run as root.
6298 The bug described in (1) was not present in this case, but the tidying up
6299 of the other signals was missing. I have made the two cases consistent.
6301 3. The ignore_target_hosts setting on a manualroute router was being ignored
6302 for hosts that were looked up using the /MX notation.
6304 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6307 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6308 operated on the sender address. After changing the $sender_address to <>
6309 for the sender address verify, Exim was re-instated it as the original
6310 (before rewriting) address, but remembering that it had rewritten it, so it
6311 wasn't rewriting it again. This bug also had the effect of breaking the
6312 sender address verification caching when the sender address was rewritten.
6314 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6315 This has been changed so that if the ip literal address matches
6316 ignore_target_hosts, the router declines.
6318 7. Added expansion conditions match_domain, match_address, and match_local_
6319 part (NOT match_host).
6321 8. The placeholder for the Received: header didn't have a length field set.
6323 9. Added code to Exim itself and to exim_lock to test for a specific race
6324 condition that could lead to file corruption when using MBX delivery. The
6325 issue is with the lockfile that is created in /tmp. If this file is removed
6326 after a process has opened it but before that process has acquired a lock,
6327 there is the potential for a second process to recreate the file and also
6328 acquire a lock. This could lead to two Exim processes writing to the file
6329 at the same time. The added code performs the same test as UW imapd; it
6330 checks after acquiring the lock that its file descriptor still refers to
6331 the same named file.
6333 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6334 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6337 11. Added the smtp_active_hostname option. If used, this will typically be made
6338 to depend on the incoming interface address. Because $interface_address is
6339 not set up until the daemon has forked a reception process, error responses
6340 that can happen earlier (such as "too many connections") no longer contain
6343 12. If an expansion in a condition on a "warn" statement fails because a lookup
6344 defers, the "warn" statement is abandoned, and the next ACL statement is
6345 processed. Previously this caused the whole ACL to be aborted.
6347 13. Added the iplsearch lookup type.
6349 14. Added ident_timeout as a log selector.
6351 15. Added tls_certificate_verified as a log selector.
6353 16. Added a global option tls_require_ciphers (compare the smtp transport
6354 option of the same name). This controls incoming TLS connections.
6356 17. I finally figured out how to make tls_require_ciphers do a similar thing
6357 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6358 before starting the TLS session.
6360 18. Tabs are now shown as \t in -bP output.
6362 19. If the log selector return_path_on_delivery was set, Exim crashed when
6363 bouncing a message because it had too many Received: header lines.
6365 20. If two routers both had headers_remove settings, and the first one included
6366 a superfluous trailing colon, the final name in the first list and the
6367 first name in the second list were incorrectly joined into one item (with a
6368 colon in the middle).
6374 1. Added -C and -D options to the exinext utility, mainly to make it easier
6375 to include in the automated testing, but these could be helpful when
6376 multiple configurations are in use.
6378 2. The exinext utility was not formatting the output nicely when there was
6379 an alternate port involved in the retry record key, nor when there was a
6380 message id as well (for retries that were specific to a specific message
6381 and a specific host). It was also confused by IPv6 addresses, because of
6382 the additional colons they contain. I have fixed the IPv4 problem, and
6383 patched it up to do a reasonable job for IPv6.
6385 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6386 delivery, the log line now contains "pipelined" if PIPELINING was used.
6388 4. An SMTP transport process used to panic and die if the bind() call to set
6389 an explicit outgoing interface failed. This has been changed; it is now
6390 treated in the same way as a connect() failure.
6392 5. A reference to $sender_host_name in the part of a conditional expansion
6393 that was being skipped was still causing a DNS lookup. This no longer
6396 6. The def: expansion condition was not recognizing references to header lines
6397 that used bh_ and bheader_.
6399 7. Added the _cache feature to named lists.
6401 8. The code for checking quota_filecount in the appendfile transport was
6402 allowing one more file than it should have been.
6404 9. For compatibility with Sendmail, the command line option
6412 and sets the incoming protocol and host name (for trusted callers). The
6413 host name and its colon can be omitted when only the protocol is to be set.
6414 Note the Exim already has two private options, -pd and -ps, that refer to
6415 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6416 "s", but I don't think that's a major issue.
6418 10. A number of refactoring changes to the code, none of which should affect
6421 (a) The number of logging options was getting close to filling up the
6422 32-bit word that was used as a bit map. I have split them into two classes:
6423 those that are passed in the argument to log_write(), and those that are
6424 only ever tested independently outside of that function. These are now in
6425 separate 32-bit words, so there is plenty of room for expansion again.
6426 There is no change in the user interface or the logging behaviour.
6428 (b) When building, for example, log lines, the code previously used a
6429 macro that called string_cat() twice, in order to add two strings. This is
6430 not really sufficiently general. Furthermore, there was one instance where
6431 it was actually wrong because one of the argument was used twice, and in
6432 one call a function was used. (As it happened, calling the function twice
6433 did not affect the overall behaviour.) The macro has been replaced by a
6434 function that can join an arbitrary number of extra strings onto a growing
6437 (c) The code for expansion conditions now uses a table and a binary chop
6438 instead of a serial search (which was left over from when there were very
6439 few conditions). Also, it now recognizes conditions like "pam" even when
6440 the relevant support is not compiled in: a suitably worded error message is
6441 given if an attempt is made to use such a condition.
6443 11. Added ${time_interval:xxxxx}.
6445 12. A bug was causing one of the ddress fields not to be passed back correctly
6446 from remote delivery subprocesses. The field in question was not being
6447 subsequently used, so this caused to problems in practice.
6449 13. Added new log selectors queue_time and deliver_time.
6451 14. Might have fixed a bug in maildirsizefile handling that threw up
6452 "unexpected character" debug warnings, and recalculated the data
6453 unnecessarily. In any case, I expanded the warning message to give more
6456 15. Added the message "Restricted characters in address" to the statements in
6457 the default ACL that block characters like @ and % in local parts.
6459 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6460 Three changes have been made:
6462 (a) There was a serious bug; a negative response to MAIL caused the whole
6463 recipient domain to be cached as invalid, thereby blocking all messages
6464 to all local parts at the same domain, from all senders. This bug has
6465 been fixed. The domain is no longer cached after a negative response to
6466 MAIL if the sender used is not empty.
6468 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6471 (c) A new callout option, "use_sender" has been added for people who want
6472 the modified behaviour.
6478 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6481 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6482 indeed breaks things for older releases.
6484 3. Added additional logging to the case where there is a problem reading data
6485 from a filter that is running in a subprocess using a pipe, in order to
6486 try to track down a specific problem.
6488 4. Testing facility fudge: when running in the test harness and attempting
6489 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6490 getting "No route to host". Convert this to a timeout.
6492 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6495 6. Some OS don't have socklen_t but use size_t instead. This affects the
6496 fifth argument of getsockopt() amongst other things. This is now
6497 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6498 can be set for individual OS. I have set it for SunOS5, OSF1, and
6499 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6500 some earlier ones do not.
6502 7. Change 4.30/15 was not doing the test caselessly.
6504 8. The standard form for an IPv6 address literal was being rejected by address
6505 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6506 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6507 this, as well as the form without the "IPv6" on the front (but only when
6508 address literals are enabled, of course).
6510 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6512 10. Exim crashed if a message with an empty sender address specified by -f
6513 encountered a router with an errors_to setting. This could be provoked only
6514 by a command such as
6518 where an empty string was supplied; "<>" did not hit this bug.
6520 11. Installed PCRE release 4.5.
6522 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6523 remained set. It is now erased.
6525 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6526 times from message ids (which are base 36 rather than the normal 62).
6528 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6529 were being counted as actual protocol errors, and logged if the log
6530 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6531 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6532 and DATA following a set of rejected RCPTs do not count as protocol errors.
6533 In other words, Exim assumes they were pipelined, though this may not
6534 actually be the case. Of course, in all cases the client gets an
6535 appropriate error code.
6537 15. If a lookup fails in an ACL condition, a message about the failure may
6538 be available; it is used if testing the ACL cannot continue, because most
6539 such messages specify what the cause of the deferral is. However, some
6540 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6541 that caused an old message to be retained and used if a later statement
6542 caused a defer, replacing the real cause of the deferral.
6544 16. If an IP address had so many PTR records that the DNS lookup buffer
6545 was not large enough to hold them, Exim could crash while trying to process
6546 the truncated data. It now detects and logs this case.
6548 17. Further to 4.21/58, another change has been made: if (and only if) the
6549 first line of a message (the first header line) ends with CRLF, a bare LF
6550 in a subsequent header line has a space inserted after it, so as not to
6551 terminate the header.
6553 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6554 unnecessarily, used atoi() instead of strtol(), and didn't check the
6555 termination when getting file sizes from file names by regex.
6557 19. Completely re-implemented the support for maildirsize files, in the light
6558 of a number of problems with the previous contributed implementation
6559 (4.30/29). In particular:
6561 . If the quota is zero, the maildirsize file is maintained, but no quota is
6564 . If the maildir directory does not exist, it is created before any attempt
6565 to write a maildirsize file.
6567 . The quota value in the file is just a cache; if the quota is changed in
6568 the transport, the new value overrides.
6570 . A regular expression is available for excluding directories from the
6573 20. The autoreply transport checks the characters in options that define the
6574 message's headers; it allows continued headers, but it was checking with
6575 isspace() after an embedded newline instead of explicitly looking for a
6578 21. If all the "regular" hosts to which an address was routed had passed their
6579 expiry times, and had not reached their retry times, the address was
6580 bounced, even if fallback hosts were defined. Now Exim should go on to try
6583 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6584 equivalent code in the SMTP transport. Some hosts send humungous responses
6585 to HELO/EHLO, more than 1024 it seems.
6587 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6588 gives compiler warnings in some environments. I've now done it "properly",
6591 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6592 (because of problems with the built-in one) was declared to return uschar *
6593 instead of char *, causing compiler failure.
6595 25. Fixed a file descriptor leak when processing alias/forward files.
6597 26. Fixed a minor format string issue in dbfn.c.
6599 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6601 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6602 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6605 29. When a sender address is verified, it is cached, to save repeating the test
6606 when there is more than one recipient in a message. However, when the
6607 verification involves a callout, it is possible for different callout
6608 options to be set for different recipients. It is too complicated to keep
6609 track of this in the cache, so now Exim always runs a verification when a
6610 callout is required, relying on the callout cache for the optimization.
6611 The overhead is duplication of the address routing, but this should not be
6614 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6615 to be verified with callout=postmaster, and the main callout worked but the
6616 postmaster check failed, the verification correctly failed. However, if a
6617 subsequent RCPT command asked for sender verification *without* the
6618 postmaster check, incorrect caching caused this verification also to fail,
6621 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6622 it was not caching the DNS options (qualify_single, search_parents) that
6623 were used when the lookup failed. A subsequent lookup with different
6624 options therefore always gave the same answer, though there were cases
6625 where it should not have. (Example: a "domains = !$mx_any" option on a
6626 dnslookup router: the "domains" option is always processed without any
6627 widening, but the router might have qualify_single set.) Now Exim uses the
6628 cached value only when the same options are set.
6630 32. Added John Jetmore's "exipick" utility to the distribution.
6632 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6633 than a timeout (e.g. a certificate is required, and is not provided), an
6634 Exim server now closes the connection immediately. Previously it waited for
6635 the client to close - but if the client is SSL, it seems that they each
6636 wait for each other, leading to a delay before one of them times out.
6638 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6639 maintained 0.8.x compatibility because I don't think many are using it, and
6640 it is clearly obsolete.
6642 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6645 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6646 client certificate was expired. A simple patch fixes this, though I don't
6647 understand the full logic of why the verify callback is called multiple
6650 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6651 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6652 which causes problems with some clients (such as the Certicom SSL Plus
6653 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6654 disables the coutermeasure allowing Eudora to connect."
6656 38. Exim was not checking that a write() to a log file succeeded. This could
6657 lead to Bad Things if a log got too big, in particular if it hit a file
6658 size limit. Exim now panics and dies if it cannot write to a log file, just
6659 as it does if it cannot open a log file.
6661 39. Modified OS/Makefile-Linux so that it now contains
6663 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6665 The two -D definitions ensure that Exim is compiled with large file
6666 support, which makes it possible to handle log files that are bigger than
6669 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6670 instance) a domain was checked against a named list that involved a lookup,
6671 causing $domain_data to be set, then another domain was checked against the
6672 same list, then the first domain was re-checked, the value of $domain_data
6673 after the final check could be wrong. In particular, if the second check
6674 failed, it could be set empty. This bug probably also applied to
6677 41. The strip_trailing_dot option was not being applied to the address given
6678 with the -f command-line option.
6680 42. The code for reading a message's header from the spool was incrementing
6681 $received_count, but never initializing it. This meant that the value was
6682 incorrect (doubled) while delivering a message in the same process in which
6683 it was received. In the most common configuration of Exim, this never
6684 happens - a fresh exec is done - but it can happen when
6685 deliver_drop_privilege is set.
6687 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6688 it now includes up to 150 characters of the unexpected data in the log
6691 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6692 and building data strings. The size of both of these buffers was 10 000
6693 bytes - far larger than anybody would *ever* want, thought I. Needless to
6694 say, somebody hit the limit. I have increased the maximum line length to
6695 20 000 and the maximum data length of concatenated lines to 100 000. I have
6696 also fixed two bugs, because there was no checking on these buffers. Tsk,
6697 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6698 buffer is too small.
6700 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6701 lsearch lookups. Now it does.
6703 46. When parsing a route_list item in a manualroute router, a fixed-length
6704 buffer was used for the list of hosts. I made this 1024 bytes long,
6705 thinking that nobody would ever have a list of hosts that long. Wrong.
6706 Somebody had a whole pile of complicated expansion conditions, and the
6707 string was silently truncated, leading to an expansion error. It turns out
6708 that it is easier to change to an unlimited length (owing to other changes
6709 that have happened since this code was originally written) than to build
6710 structure for giving a limitation error. The length of the item that
6711 expands into the list of hosts is now unlimited.
6713 47. The lsearch lookup could not handle data where the length of text line was
6714 more than 4095 characters. Such lines were truncated, leading to shortened
6715 data being returned. It should now handle lines of any length.
6717 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6718 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6721 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6722 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6723 no longer happens in comment lines. A list of replacements is now placed
6724 at the head of all of the source files, except those whose only change is
6725 to replace PERL_COMMAND in the very first #! line.
6727 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6728 messages on the queue, with a bottom-up merge sort, using code contributed
6729 by Michael Haardt. This should make operations like -bp somewhat faster on
6730 large queues. It won't affect queue runners, except when queue_run_in_order
6733 51. Installed eximstats 1.31 in the distribution.
6735 52. Added support for SRV lookups to the dnslookup router.
6737 53. If an ACL referred to $message_body or $message_body_end, the value was not
6738 reset for any messages that followed in the same SMTP session.
6740 54. The store-handling optimization for building very long strings was not
6741 differentiating between the different store pools. I don't think this
6742 actually made any difference in practice, but I've tidied it.
6744 55. While running the routers to verify a sender address, $sender_address
6745 was still set to the sender address. This is wrong, because when routing to
6746 send a bounce to the sender, it would be empty. Therefore, I have changed
6747 it so that, while verifying a sender address, $sender_address is set to <>.
6748 (There is no change to what happens when verifying a recipient address.)
6750 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6751 target A or AAAA records (if not already returned) without resetting the
6752 qualify_single or search_parents options of the DNS resolver. These are
6753 inappropriate in this case because the targets of MX and SRV records must
6754 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6755 target that, when qualified, matched something in the local domain. These
6756 two options are now turned off when doing these lookups.
6758 57. It seems that at least some releases of Reiserfs (which does not have the
6759 concept of a fixed number of inodes) returns zero and not -1 for the
6760 number of available inodes. This interacted badly with check_spool_inodes,
6761 which assumed that -1 was the "no such thing" setting. What I have done is
6762 to check that the total number of inodes is greater than zero before doing
6763 the test of how many are available.
6765 58. When a "warn" ACL statement has a log_message modifier, the message is
6766 remembered, and not repeated. This is to avoid a lot of repetition when a
6767 message has many recipients that cause the same warning to be written.
6768 However, Exim was preserving the list of already written lines for an
6769 entire SMTP session, which doesn't seem right. The memory is now reset if a
6770 new message is started.
6772 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6773 debugging output unless log_rewrite was also set.
6775 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6776 of (int)(handle) when we know that handle contains (void *)(-1).
6778 61. The Exim daemon panic-logs an error return when it closes the incoming
6779 connection. However "connection reset by peer" seems to be common, and
6780 isn't really an error worthy of noting specially, so that particular error
6783 62. When Exim is trying to find all the local interfaces, it used to panic and
6784 die if the ioctl to get the interface flags failed. However, it seems that
6785 on at least one OS (Solaris 9) it is possible to have an interface that is
6786 included in the list of interfaces, but for which you get a failure error
6787 for this call. This happens when the interface is not "plumbed" into a
6788 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6789 failure of the "get flags" call assumes that the interface is down.
6791 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6792 makes life easier for people who are doing arithmetic on fields extracted
6793 from dates, where you often get leading zeros that should not be
6794 interpreted as octal.
6796 64. Added qualify_domain to the redirect router, to override the global
6799 65. If a pathologically long header line contained very many addresses (the
6800 report of this problem mentioned 10 000) and each of them was rewritten,
6801 Exim could use up a very large amount of memory. (It kept on making new
6802 copies of the header line as it rewrote, and never released the old ones.)
6803 At the expense of a bit more processing, the header rewriting function has
6804 been changed so that it no longer eats memory in this way.
6806 66. The generation of the Received: header has been moved from the time that a
6807 message starts to be received, to the time that it finishes. The timestamp
6808 in the Received: header should now be very close to that of the <= log
6809 line. There are two side-effects of this change:
6811 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6812 logged header lines no longer include the local Received: line, because
6813 it has not yet been created. The same applies to a copy of the message
6814 that is returned to a non-SMTP sender when a message is rejected.
6816 (b) When a filter file is tested using -bf, no additional Received: header
6817 is added to the test message. After some thought, I decided that this
6820 This change does not affect the value of $received_for. It is still set
6821 after address rewriting, but before local_scan() is called.
6823 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6825 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6826 gave an unhelpful panic error message, and a defer error. I have managed to
6827 change this behaviour so that it now rejects any supplied certificate,
6828 which seems right, as the list of acceptable certificates is empty.
6830 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6831 gave an unhelpful defer error. I have not managed to make this reject any
6832 supplied certificates, but the error message it gives is "no certificate
6833 supplied", which is not helpful.
6835 70. exigrep's output now also includes lines that are not associated with any
6836 message, but which match the given pattern. Implemented by a patch from
6837 Martin Sluka, which also tidied up the Perl a bit.
6839 71. Recipient callout verification, like sender verification, was using <> in
6840 the MAIL FROM command. This isn't really the right thing, since the actual
6841 sender may affect whether the remote host accepts the recipient or not. I
6842 have changed it to use the actual sender in the callout; this means that
6843 the cache record is now keyed on a recipient/sender pair, not just the
6844 recipient address. There doesn't seem to be a real danger of callout loops,
6845 since a callout by the remote host to check the sender would use <>.
6846 [SEE ABOVE: changed after hitting problems.]
6848 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6849 temporary errors. However, in the case of such a code being given after
6850 the end of a data transmission (i.e. after ".") Exim was failing to write
6851 a retry record for the message. (Yes, there was some broken host that was
6852 actually sending 8xx at this point.)
6854 73. An unknown lookup type in a host list could cause Exim to panic-die when
6855 the list was checked. (An example that provoked this was putting <; in the
6856 middle of a list instead of at the start.) If this happened during a DATA
6857 ACL check, a -D file could be left lying around. This kind of configuration
6858 error no longer causes Exim to die; instead it causes a defer error. The
6859 incident is still logged to the main and panic logs.
6861 74. Buglet left over from Exim 3 conversion. The message "too many messages
6862 in one connection" was written to the rejectlog but not the mainlog, except
6863 when address rewriting (yes!) was being logged.
6865 75. Added write_rejectlog option.
6867 76. When a system filter was run not as root (that is, when system_filter_user
6868 was set), the values of the $n variables were not being returned to the
6869 main process; thus, they were not subsequently available in the $sn
6872 77. Added +return_path_on_delivery log selector.
6874 78. A connection timeout was being treated differently from recipients deferred
6875 when testing hosts_max_try with a message that was older than the host's
6876 retry timeout. (The host should not be counted, thus allowing all hosts to
6877 be tried at least once before bouncing.) This may have been the cause of an
6878 occasionally reported bug whereby a message would remain on the queue
6879 longer than the retry timeout, but would be bounced if a delivery was
6880 forced. I say "may" because I never totally pinned down the problem;
6881 setting up timeout/retry tests is difficult. See also the next item.
6883 79. The ultimate address timeout was not being applied to errors that involved
6884 a combination of host plus message (for example, a timeout on a MAIL
6885 command). When an address resolved to a number of possible hosts, and they
6886 were not all tried for each delivery (e.g. because of hosts_max_try), a
6887 message could remain on the queue longer than the retry timeout.
6889 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6892 81. Fixed an obscure SMTP outgoing bug which required at least the following
6893 conditions: (a) there was another message waiting for the same server;
6894 (b) the server returned 5xx to all RCPT commands in the first message so
6895 that the message was not completed; (c) the server dropped the connection
6896 or gave a negative response to the RSET that Exim sends to abort the
6897 transaction. The observed case was a dropped connection after DATA that had
6898 been sent in pipelining mode. That is, the server had advertised PIPELINING
6899 but was not implementing it correctly. The effect of the bug was incorrect
6900 behaviour, such as trying another host, and this could lead to a crash.
6906 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6907 and daemon.c were passed as pointers to ints; they should have been
6908 pointers to socklen_t variables (which are typically unsigned ints).
6910 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6913 3. Fixed a really odd bug that affected only the testing scheme; patching a
6914 certain fixed string in the binary changed the value of another string that
6915 happened to be identical to the end of the original first string.
6917 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6918 name", it returns that address as the IP address. On some operating
6919 systems (e.g. Solaris), it also passes back the IP address string as the
6920 "host name". However, on others (e.g. Linux), it passes back an empty
6921 string. Exim wasn't checking for this, and was changing the host name to an
6922 empty string, assuming it had been canonicalized.
6924 5. Although rare, it is permitted to have more than one PTR record for a given
6925 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6926 all the names associated with an address, because they do in Solaris.
6927 However, it seems that they do not in Linux for data that comes from the
6928 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6929 I found this out when I moved to a new Linux workstation and tried to run
6930 the Exim test suite.
6932 To get round this problem I have changed the code so that it now does its
6933 own call to the DNS to look up PTR records when searching for a host name.
6934 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6935 addresses that are only in /etc/hosts are still found.
6937 This behaviour is, however, controlled by an option called host_lookup_
6938 order, which defaults to "bydns:byaddr". If people want to use the other
6939 order, or indeed, just use one or the other means of lookup, they can
6940 specify it in this variable.
6942 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6943 some operating systems, this comes back from gethostbyaddr() as an empty
6944 string, and this is what Exim used to test for. However, it seems that in
6945 other systems, "." is yielded. Exim now tests for this case too.
6947 7. The values of check_spool_space and check_log_space are now held internally
6948 as a number of kilobytes instead of an absolute number of bytes. If a
6949 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6950 kilobyte. This means that much larger values can be stored.
6952 8. Exim monitor: an attempt to get the action menu when not actually pointing
6953 at a message produces an empty menu entitled "No message selected". This
6954 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6955 no entries in it ("Shell widget menu has zero width and/or height"). So I
6956 have added a single, blank menu entry in this case.
6958 9. Added ${quote_local_part.
6960 10. MIME decoding is now applied to the contents of Subject: header lines when
6963 11. Now that a reference to $sender_host_address automatically causes a reverse
6964 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6965 host lookup before query-style lookups in lists that might use this
6966 variable. This has therefore been abolished, and the "net-" prefix is no
6967 longer necessary for query-style lookups.
6969 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6970 have been a typo for LFLAGS, so it has been changed.
6972 13. The install script calls Exim with "-C /dev/null" in order to find the
6973 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6974 to be output. However, since Exim outputs its version number before the
6975 error, it didn't break the script. It just looked ugly. I fixed this by
6976 always allowing "-C /dev/null" if the caller is root.
6978 14. Ignore overlarge ACL variable number when reading spool file - insurance
6979 against a later release with more variables having written the file.
6981 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6982 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6983 this, as well as the form without the "IPv6" on the front.
6985 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6986 OS/Makefile-Darwin file.
6988 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6989 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6991 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6992 variable. However, some OS use ints for inodes. I've added cast to long int
6993 to get rid of the compiler warning.
6995 19. I had forgotten to lock out "/../" in configuration file names when
6996 ALT_CONFIG_PREFIX was set.
6998 20. Routers used for verification do not need to specify transports. However,
6999 if such a router generated a host list, and callout was configured, Exim
7000 crashed, because it could not find a port number from the (non-existent)
7001 transport. It now assumes port 25 in this circumstance.
7003 21. Added the -t option to exigrep.
7005 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7006 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7007 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7008 them is set, LOOKUP_LSEARCH is forced.
7010 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7012 24. Added sender and host information to the "rejected by local_scan()" log
7013 line; previously there was no indication of these.
7015 25. Added .include_if_exists.
7017 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7018 Linux. It turns out that not all file systems support this. Apparently some
7019 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7020 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7021 not supported on the file descriptor, is now ignored when Exim is trying to
7022 sync a directory. This applies only to Linux.
7024 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7026 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7027 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7030 29. Added support for maildirsize files from supplied patch (modified a bit).
7032 30. The use of :fail: followed by an empty string could lead Exim to respond to
7033 sender verification failures with (e.g.):
7035 550 Verification failed for <xxx>
7036 550 Sender verify failed
7038 where the first response line was missing the '-' that indicates it is not
7039 the final line of the response.
7041 31. The loop for finding the name of the user that called Exim had a hardwired
7042 limit of 10; it now uses the value of finduser_retries, which is used for
7043 all other user lookups.
7045 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7048 33. Exim was neglecting to zero errno before one call of strtol() when
7049 expanding a string and expecting an integer value. On some systems this
7050 resulted in spurious "integer overflow" errors. Also, it was casting the
7051 result into an int without checking.
7053 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7054 did not work. The code looks as if it has *never* worked, though it appears
7055 to have been documented since at least release 1.62. I have made it work.
7057 35. The "timeout_DNS" error in retry rules, also documented since at least
7058 1.62, also never worked. As it isn't clear exactly what this means, and
7059 clearly it isn't a major issue, I have abolished the feature by treating it
7060 as "timeout", and writing a warning to the main and panic logs.
7062 36. The display of retry rules for -brt wasn't always showing the error code
7065 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7066 timeout_connect_A, timeout_connect_MX.
7068 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7069 to the empty sender.
7071 39. The daemon was not analysing the content of -oX till after it had closed
7072 stderr and disconnected from the controlling terminal. This meant that any
7073 syntax errors were only noted on the panic log, and the return code from
7074 the command was 0. By re-arranging the code a little, I've made the
7075 decoding happen first, so such errors now appear on stderr, and the return
7076 code is 1. However, the actual setting up of the sockets still happens in
7077 the disconnected process, so errors there are still only recorded on the
7080 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7081 connections (as happens on some IP stacks) was logged at start up time as
7082 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7083 it from "IPv6 and IPv4", which means that two separate sockets are being
7086 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7087 says whether AF_INET or AF_INET6 was passed as an argument.
7089 42. Exiwhat output was messed up when time zones were included in log
7092 43. Exiwhat now gives more information about the daemon's listening ports,
7093 and whether -tls-on-connect was used.
7095 44. The "port" option of the smtp transport is now expanded.
7097 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7098 silently ignored. Now an error message is written to the main and panic
7101 46. There's a new ACL modifier called "logwrite" which writes to a log file
7102 as soon as it is encountered.
7104 47. Added $local_user_uid and $local_user_gid at routing time.
7106 48. Exim crashed when trying to verify a sender address that was being
7109 49. Exim was recognizing only a space character after ".include". It now also
7110 recognizes a tab character.
7112 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7113 extracting the relevant information from the specification. The man page no
7114 longer contains scrambled data for the -d option, and I've added a section
7115 at the front about calling Exim under different names.
7117 51. Added "extra_headers" argument to the "mail" command in filter files.
7119 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7122 53. Installed eximstats 1.29.
7124 54. Added transport_filter_timeout as a generic transport option.
7126 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7127 Cc: header lines. This was required by RFC 822, but it not required by RFC
7130 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7131 incoming messages that did not have them. Now it does so only if the
7132 message originates locally, that is, if there is no associated remote host
7133 address. When Resent- header lines are present, this applies to the Resent-
7134 lines rather than the non-Resent- lines.
7136 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7137 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7139 58. Messages for configuration errors now include the name of the main
7140 configuration file - useful now that there may be more than one file in a
7141 list (.included file names were always shown).
7143 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7144 for those rare installations that do not start the daemon as root or run it
7145 setuid root. I've cut out the call to initgroups() if the daemon is not
7148 60. The Exim user and group can now be bound into the binary as text strings
7149 that are looked up at the start of Exim's processing.
7151 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7153 62. Added $mailstore_basename variable.
7155 63. Installed patch to sieve.c from Michael Haardt.
7157 64. When Exim failed to open the panic log after failing to open the main log,
7158 the original message it was trying to log was written to stderr and debug
7159 output, but if they were not available (the usual case in production), it
7160 was lost. Now it is written to syslog before the two lines that record the
7161 failures to open the logs.
7163 65. Users' Exim filters run in subprocesses under the user's uid. It is
7164 possible for a "deliver" command or an alias in a "personal" command to
7165 provoke an address rewrite. If logging of address rewriting is configured,
7166 this fails because the process is not running as root or exim. There may be
7167 a better way of dealing with this, but for the moment (because 4.30 needs
7168 to be released), I have disabled address rewrite logging when running a
7169 filter in a non-root, non-exim process.
7175 1. The buildconfig auxiliary program wasn't quoting the value set for
7176 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7177 not defined. This bug was masked in 4.22 by the effect that was fixed in
7180 2. Some messages that were rejected after a message id was allocated were
7181 shown as "incomplete" by exigrep. It no longer does this for messages that
7182 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7184 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7185 have allow_domain_literals set, the ID did not get logged in the <= line.
7186 Domain literals are now always recognized in Message-ID: header lines.
7188 4. The first argument for a ${extract expansion item is the key name or field
7189 number. Leading and trailing spaces in this item were not being ignored,
7190 causing some misleading effects.
7192 5. When deliver_drop_privilege was set, single queue runner processes started
7193 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7194 same command in the process it spins off) were not dropping privilege.
7196 6. When the daemon running as "exim" started a queue runner, it always
7197 re-executed Exim in the spun-off process. This is a waste of effort when
7198 deliver_drop_privilege is set. The new process now just calls the
7199 queue-runner function directly.
7205 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7208 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7209 ignored. Though the use of -oP was forcing the writing of a pid file, it
7210 was always written to the default place.
7212 3. If the message "no IP address found for host xxxx" is generated during
7213 incoming verification, it is now followed by identification of the incoming
7214 connection (so you can more easily find what provoked it).
7216 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7218 5. Added some features to "harden" Exim a bit more against certain attacks:
7220 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7221 be put in Local/Makefile. This is like the never_users runtime option,
7222 but it cannot be overridden. The default setting is "root".
7224 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7225 prefix string with which any file named in a -C command line option
7228 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7229 is retained for -C and -D only if the caller of Exim is root. Without
7230 it, the exim user may also use -C and -D and retain privilege.
7232 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7233 command line option is disabled.
7235 6. Macro names set by the -D option must start with an upper case letter, just
7236 like macro names defined in the configuration file.
7238 7. Added "dereference=" facility to LDAP.
7240 8. Two instances of the typo "uknown" in the source files are fixed.
7242 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7243 the Configure-Makefile script screwed up while processing it.
7245 10. Incorporated PCRE 4.4.
7247 11. The SMTP synchronization check was not operating right at the start of an
7248 SMTP session. For example, it could not catch a HELO sent before the client
7249 waited for the greeting. There is now a check for outstanding input at the
7250 point when the greeting is written. Because of the duplex, asynchronous
7251 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7252 way, but not yet received, when the check is performed.
7254 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7255 on TCP/IP sockets, because this apparently causes some broken clients to
7258 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7259 unchanged) from the Cygwin maintainer.
7261 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7262 was supported instead of testing for mbx. Effectively a typo.
7264 15. The spa authenticator server code was not checking that the input it
7265 received was valid base64.
7267 16. The debug output line for the "set" modifier in ACLs was not showing the
7268 name of the variable that was being set.
7270 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7272 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7273 lookup on its own. Something else was needed to trigger the lookup. For
7274 example, a match in host_lookup or the need for a host name in a host list.
7275 Now, if $sender_host_name is referenced and the host name has not yet been
7276 looked up, a lookup is performed. If the lookup fails, the variable remains
7277 empty, and $host_lookup_failed is set to "1".
7279 19. Added "eqi" as a case-independent comparison operator.
7281 20. The saslauthd authentication condition could segfault if neither service
7282 nor realm was specified.
7284 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7285 error message that was logged was misleading, and incoming SMTP
7286 connections were dropped. The message is now more accurate, and temporary
7287 errors are given to SMTP connections.
7289 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7290 (except RSET) with a 421 error, until QUIT is received. However, it was
7291 failing to send a response to QUIT.
7293 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7294 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7295 ACL is now run after the helo_try_verify_hosts code.
7297 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7298 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7299 case-independent, so other case variants are also recognized). Apparently
7300 some systems use these upper case variants.
7302 25. If more than two messages were waiting for the same host, and a transport
7303 filter was specified for the transport, Exim sent two messages over the
7304 same TCP/IP connection, and then failed with "socket operation on non-
7305 socket" when it tried to send the third.
7307 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7309 27. The extern definition of crypt16() in expand.c was not being excluded when
7310 the OS had its own crypt16() function.
7312 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7313 as a preferred synonym for return_size_limit, both as an option and as an
7316 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7318 30. Changed the default configuration ACL to relax the local part checking rule
7319 for addresses that are not in any local domains. For these addresses,
7320 slashes and pipe symbols are allowed within local parts, but the sequence
7321 /../ is explicitly forbidden.
7323 31. SPA server authentication was not clearing the challenge buffer before
7326 32. log_message in a "warn" ACL statement was writing to the reject log as
7327 well as to the main log, which contradicts the documentation and doesn't
7328 seem right (because no rejection is happening). So I have stopped it.
7330 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7331 However, I am unable to do any testing of this.
7333 34. Fixed an infelicity in the appendfile transport. When checking directories
7334 for a mailbox, to see if any needed to be created, it was accidentally
7335 using path names with one or more superfluous leading slashes; tracing
7336 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7338 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7339 contents are added to the log line that is written for every discarded
7340 recipient. (Previously a log_message setting was ignored.)
7342 36. The ${quote: operator now quotes the string if it is empty.
7344 37. The install script runs exim in order to find its version number. If for
7345 some reason other than non-existence or emptiness, which it checks, it
7346 could not run './exim', it was installing it with an empty version number,
7347 i.e. as "exim-". This error state is now caught, and the installation is
7350 38. An argument was missing from the function that creates an error message
7351 when Exim fails to connect to the socket for saslauthd authentication.
7352 This could cause Exim to crash, or give a corrupted message.
7354 39. Added isip, isip4, and isip6 to ${if conditions.
7356 40. The ACL variables $acl_xx are now saved with the message, and can be
7357 accessed later in routers, transports, and filters.
7359 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7360 strings in the file are not string-expanded.
7362 42. If a MAIL command specified a SIZE value that was too large to fit into an
7363 int variable, the check against message_size_limit failed. Such values are
7364 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7365 day this will have to be increased, but I don't think I want to be around
7366 when emails are that large.
7373 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7374 iconv() is not standard in FreeBSD.
7376 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7377 IPv6 enabled. The observed symptom was a segmentation fault on return from
7378 the function os_common_find_running_interfaces() in src/os.c.
7380 3. In the check_special_case() function in daemon.c I had used "errno" as an
7381 argument name, which causes warnings on some systems. This was basically a
7382 typo, since it was named "eno" in the comments!
7384 4. The code that waits for the clock to tick (at a resolution of some fraction
7385 of a second) so as to ensure message-id uniqueness was always waiting for
7386 at least one whole tick, when it could have waited for less. [This is
7387 almost certainly not relevant at current processor speeds, where it is
7388 unlikely to ever wait at all. But we try to future-proof.]
7390 5. The function that sleeps for a time interval that includes fractions of a
7391 second contained a race. It did not block SIGALRM between setting the
7392 timer, and suspending (a couple of lines later). If the interval was short
7393 and the sigsuspend() was delayed until after it had expired, the suspension
7394 never ended. On busy systems this could lead to processes getting stuck for
7397 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7398 process, before it forks any delivery processes. The open lookup caching
7399 mechanism meant that the open file or database connection was passed into
7400 the delivery process. The problem was that delivery processes always tidy
7401 up cached lookup data. This could cause a problem for the next delivery
7402 process started by the queue runner, because the external queue runner
7403 process does not know about the closure. So the next delivery process
7404 still has data in the lookup cache. In the case of a file lookup, there was
7405 no problem because closing a file descriptor in a subprocess doesn't affect
7406 the parent. However, if the lookup was caching a connection to a database,
7407 the connection was closed, and the second delivery process was likely to
7408 see errors such as "PGSQL: query failed: server closed the connection
7409 unexpectedly". The problem has been fixed by closing all cached lookups
7410 in a queue runner before running a delivery process.
7412 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7413 seem to have the "const" qualifier which it has on other OS. I've
7416 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7417 *of the same type* (client or server) with the same public name that an
7418 error should be diagnosed.
7420 9. When Exim looked up a host name for an IP address, but failed to find the
7421 original IP address when looking up the host name (a safety check), it
7422 output the message "<ip address> does not match any IP for NULL", which was
7423 confusing, to say the least. The bug was that the host name should have
7424 appeared instead of "NULL".
7426 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7427 user that is built into the binary, and the -C or -D options is used, root
7428 privilege is dropped before the configuration file is read. In addition,
7429 logging is switched to stderr instead of the normal log files. If the
7430 configuration then re-defines the Exim user, the unprivileged environment
7431 is probably not what is expected, so Exim logs a panic warning message (but
7434 However, if deliver_drop_privilege is set, the unprivileged state may well
7435 be exactly what is intended, so the warning has been cut out in that case,
7436 and Exim is allowed to try to write to its normal log files.
7442 1. smtp_return_error_details was not giving details for temporary sender
7443 or receiver verification errors.
7445 2. Diagnose a configuration error if two authenticators have the same public
7448 3. Exim used not to create the message log file for a message until the first
7449 delivery attempt. This could be confusing when incoming messages were held
7450 for policy or load reasons. The message log file is now created at the time
7451 the message is received, and an initial "Received" line is written to it.
7453 4. The automatically generated man page for command line options had a minor
7454 bug that caused no ill effects; however, a more serious problem was that
7455 the procedure for building the man page automatically didn't always
7456 operate. Consequently, release 4.20 contains an out-of-date version. This
7457 shouldn't happen again.
7459 5. When building Exim with embedded Perl support, the script that builds the
7460 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7461 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7463 6. The freeze_tell option was not being used for messages that were frozen on
7464 arrival, either by an ACL or by local_scan().
7466 7. Added the smtp_incomplete_transaction log selector.
7468 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7469 was accepting AUTH without a new EHLO.
7471 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7472 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7475 10. Logging of TCP/IP connections (when configured) now happens in the main
7476 daemon process instead of the child process, so that the TCP/IP connection
7477 count is more accurate (but it can never be perfect).
7479 11. The use of "drop" in a nested ACL was not being handled correctly in the
7480 outer ACL. Now, if condition failure induced by the nested "drop" causes
7481 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7482 or "require"), the connection is dropped.
7484 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7485 that yield "discard" can now be used with an "accept" or a "discard" verb,
7486 but an error is generated for any others (because I can't see a useful way
7487 to define what should happen).
7489 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7490 are now processed in the same way as lines in the Exim configuration file.
7491 In particular, continuation lines are supported.
7493 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7495 15. Added -ti meaning -t -i.
7497 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7498 domains, and warn by logging if others are found.
7500 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7501 returned from GIFCONF when Exim is trying to find the list of interfaces on
7502 a host. The code in os.c has been modified to copy each ifreq to an aligned
7503 structure in all cases.
7505 Also, in some cases, the returned ifreq's were being copied to a 'struct
7506 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7507 means the last couple of bytes of an IPv6 address could be chopped if the
7508 ifreq contained only a normal sockaddr (14 bytes storage).
7510 18. Named domain lists were not supported in the hosts_treat_as_local option.
7511 An entry such as +xxxx was not recognized, and was treated as a literal
7514 19. Ensure that header lines added by a DATA ACL are included in the reject log
7515 if the ACL subsequently rejects the message.
7517 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7518 MD5 (which is deprecated).
7520 21. When testing a filter file using -bf, Exim was writing a message when it
7521 took the sender from a "From " line in the message, but it was not doing so
7522 when it took $return_path from a Return-Path: header line. It now does.
7524 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7525 with a valid header line field name (a series of printing characters
7526 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7528 23. Changed "disc" in the source to "disk" to conform to the documentation and
7529 the book and for uniformity.
7531 24. Ignore Sendmail's -Ooption=value command line item.
7533 25. When execve() failed while trying to run a command in a pipe transport,
7534 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7535 could be confused with a return value of 69 from the command itself. This
7536 has been changed to 127, the value the shell returns if it is asked to run
7537 a non-existent command. The wording for the related log line suggests a
7538 non-existent command as the problem.
7540 26. If received_header_text expands to an empty string, do not add a Received:
7541 header line to the message. (Well, it adds a token one on the spool, but
7542 marks it "old" so that it doesn't get used or transmitted.)
7544 27. Installed eximstats 1.28 (addition of -nt option).
7546 28. There was no check for failure on the call to getsockname() in the daemon
7547 code. This can fail if there is a shortage of resources on the system, with
7548 ENOMEM, for example. A temporary error is now given on failure.
7550 29. Contrary to the C standard, it seems that in some environments, the
7551 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7552 program. Exim now does this explicitly; it affects the formatting of
7553 timestamps using strftime().
7555 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7556 complaints. I've now initialized all the variables, to avoid this.
7558 32. Header lines added by a system filter were not being "seen" during
7559 transport-time rewrites.
7561 33. The info_callback() function passed to OpenSSL is set up with type void
7562 (*)(SSL *, int, int), as described somewhere. However, when calling the
7563 function (actually a macro) that sets it up, the type void(*)() is
7564 expected. I've put in a cast to prevent warnings from picky compilers.
7566 34. If a DNS black list lookup found a CNAME record, but there were no A
7567 records associated with the domain it pointed at, Exim crashed.
7569 35. If a DNS black list lookup returned more than one A record, Exim ignored
7570 all but the first. It now scans all returned addresses if a particular IP
7571 value is being sought. In this situation, the contents of the
7572 $dnslist_value variable are a list of all the addresses, separated by a
7575 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7576 to accept a host name and all its aliases if the forward lookup for any of
7577 them yielded the IP address of the incoming connection. Now it accepts only
7578 those names whose forward lookup yields the correct IP address. Any other
7579 names are discarded. This closes a loophole whereby a rogue DNS
7580 administrator could create reverse DNS records to break through a
7581 wildcarded host restriction in an ACL.
7583 37. If a user filter or a system filter that ran in a subprocess used any of
7584 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7585 the wrong values were passed to the pipe command ($thisaddress had the
7586 value of $0, $0 had the value of $1, etc). This bug was introduced by
7587 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7589 38. Improved the line breaking for long SMTP error messages from ACLs.
7590 Previously, if there was no break point between 40 and 75 characters, Exim
7591 left the rest of the message alone. Two changes have been made: (a) I've
7592 reduced the minimum length to 35 characters; (b) if it can't find a break
7593 point between 35 and 75 characters, it looks ahead and uses the first one
7594 that it finds. This may give the occasional overlong line, but at least the
7595 remaining text gets split now.
7597 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7598 file descriptors might be low, and that setting 1000 would always raise it.
7599 It turns out that in some environments, the limit is already over 1000 and
7600 that lowering it causes trouble. So now Exim takes care not to decrease it.
7602 40. When delivering a message, the value of $return_path is set to $sender_
7603 address at the start of routing (routers may change the value). By an
7604 oversight, this default was not being set up when an address was tested by
7605 -bt or -bv, which affected the outcome if any router or filter referred to
7608 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7609 log only when "message" or "log_message" are set. However, if one of the
7610 conditions was an address verification, or a call to a nested ACL, the
7611 messages generated by the underlying test were being passed through. This
7612 no longer happens. The underlying message is available in $acl_verify_
7613 message for both "message" and "log_message" expansions, so it can be
7614 passed through if needed.
7616 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7617 new expansion $bh_ to give the encoded byte string without charset
7618 translation. Translation happens only if iconv() is available; HAVE_ICONV
7619 indicates this at build time. HEADERS_CHARSET gives the charset to
7620 translate to; headers_charset can change it in the configuration, and
7621 "headers charset" can change it in an individual filter file.
7623 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7624 that creates RFC 2047 encoded "words" labels them as that charset instead
7625 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7626 expansion operator; (ii) when Exim creates a From: line for a local
7627 message; (iii) when a header line is rewritten to include a "phrase" part.
7629 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7630 buggy, causing it to skip the first lines of messages whose message ID
7631 ended in 'D'. This would not have bitten before Exim release 4.14, because
7632 message IDs were unlikely to end in 'D' before then. The effect was to have
7633 incorrect size information for certain domains.
7635 45. #include "config.h" was missing at the start of the crypt16.c module. This
7636 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7639 46. If there was a timeout during a "random" callout check, Exim treated it as
7640 a failure of the random address, and carried on sending RSET and the real
7641 address. If the delay was just some slowness somewhere, the response to the
7642 original RCPT would be taken as a response to RSET and so on, causing
7643 mayhem of various kinds.
7645 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7646 when I implemented it. It didn't allow for the fact that some option values
7647 may legitimately be negative (e.g. size_addition), and it didn't even do
7648 the right test for positive values.
7650 48. Domain names in DNS records are case-independent. Exim always looks them up
7651 in lower case. Some resolvers return domain names in exactly the case they
7652 appear in the zone file, that is, they may contain uppercase letters. Not
7653 all resolvers do this - some return always lower case. Exim was treating a
7654 change of case by a resolver as a change of domain, similar to a widening
7655 of a domain abbreviation. This triggered its re-routing code and so it was
7656 trying to route what was effectively the same domain again. This normally
7657 caused routing to fail (because the router wouldn't handle the domain
7658 twice). Now Exim checks for this case specially, and just changes the
7659 casing of the domain that it ultimately uses when it transmits the message
7662 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7665 50. If a filter generated a file delivery with a non-absolute name (possible if
7666 no home directory exists for the router), the forbid_file option was not
7669 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7670 the existing equality matching.
7672 52. Exim was using ints instead of ino_t variables in some places where it was
7673 dealing with inode numbers.
7675 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7676 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7677 called TMPDIR, and if it finds it is different, it changes its value.
7679 54. The smtp_printf() function is now made available to local_scan() so
7680 additional output lines can be written before returning. There is also an
7681 smtp_fflush() function to enable the detection of a dropped connection.
7682 The variables smtp_input and smtp_batched_input are exported to
7685 55. Changed the default runtime configuration: the message "Unknown user"
7686 has been removed from the ACL, and instead placed on the localuser router,
7687 using the cannot_route_message feature. This means that any verification
7688 failures that generate their own messages won't get overridden. Similarly,
7689 the "Unrouteable address" message that was in the ACL for unverifiable
7690 relay addresses has also been removed.
7692 56. Added hosts_avoid_esmtp to the smtp transport.
7694 57. The exicyclog script was not checking for the esoteric option
7695 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7696 will work only if exicyclog is run under the appropriate euid.
7698 58. Following a discussion on the list, the rules by which Exim recognises line
7699 endings on incoming messages have been changed. The -dropcr and drop_cr
7700 options are now no-ops, retained only for backwards compatibility. The
7701 following line terminators are recognized: LF CRLF CR. However, special
7702 processing applies to CR:
7704 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7705 nor a local message in the state where . is a terminator.
7707 (ii) If a bare CR is encountered in a header line, an extra space is added
7708 after the line terminator so as not to end the header. The reasoning
7709 behind this is that bare CRs in header lines are most likely either
7710 to be mistakes, or people trying to play silly games.
7712 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7713 was being incorrectly given as 18 bytes larger than it should have been.
7714 This is a VOB (very old bug).
7716 60. This may never have affected anything current, but just in case it has:
7717 When the local host is found other than at the start of a list of hosts,
7718 the local host, those with the same MX, and any that follow, are discarded.
7719 When the list in question was part of a longer list of hosts, the following
7720 hosts (not currently being processed) were also being discarded. This no
7721 longer happens. I'm not sure if this situation could ever has previously
7724 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7727 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7728 header. This is something that is recommended in a new Internet Draft, and
7729 is something that is documented as being done by Sendmail. There are two
7730 possible values. For messages generated by the autoreply transport, Exim
7733 Auto-Submitted: auto-replied
7735 whereas for all other generated messages (e.g. bounces) it adds
7737 Auto-Submitted: auto-generated
7739 63. The "personal" condition in filters now includes a test for the
7740 Auto-Submitted: header. If it contains the string "auto-" the message it
7741 not considered personal.
7743 64. Added rcpt_include_affixes as a generic transport option.
7745 65. Added queue_only_override (default true).
7747 66. Added the syslog_duplication option.
7749 67. If what should have been the first header line of a message consisted of
7750 a space followed by a colon, Exim was mis-interpreting it as a header line.
7751 It isn't of course - it is syntactically invalid and should therefore be
7752 treated as the start of the message body. The misbehaviour could have
7753 caused a number of strange effects, including loss of data in subsequent
7754 header lines, and spool format errors.
7756 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7757 client host had authenticated. This control can now be exercised by an ACL
7758 for more flexibility.
7760 69. By default, callouts do not happen when testing with -bh. There is now a
7761 variant, -bhc, which does actually run the callout code, including
7762 consulting and updating the callout cache.
7764 70. Added support for saslauthd authentication, courtesy of Alexander
7767 71. If statvfs() failed on the spool or log directories while checking their
7768 size for availability, Exim confusingly gave the error "space shortage".
7769 Furthermore, in debugging mode it crashed with a floating point exception.
7770 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7771 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7772 serious problem, Exim now writes to the main and panic logs when this
7773 happens, with details of the failure. It then refuses to accept the
7774 incoming message, giving the message "spool directory problem" or "log
7775 directory problem" with a 421 code for SMTP messages.
7777 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7778 0, 1, and 2 exist, because some OS complain for execs without them (see
7779 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7780 descriptors. However, the code omitted to check that the open succeeded,
7781 causing mysterious errors if for some reason the permissions on /dev/null
7782 got screwed. Now Exim writes a message to the main and panic logs, and
7783 bombs out if it can't open /dev/null.
7785 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7786 interact so that it is all more flexible. It is supposed to remain
7787 backwards compatible. Also added extra_local_interfaces.
7789 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7790 to bomb out with an assertion failure - to the client this appears as a
7791 connection drop. This problem occurs in the part of the code that was taken
7792 from the Samba project. Fortunately, the assertion is in a very simple
7793 function, so I have fixed this by reproducing the function inline in the
7794 one place where it is called, and arranging for authentication to fail
7795 instead of killing the process with assert().
7797 75. The SPA client code was not working when the server requested OEM rather
7798 than Unicode encoding.
7800 76. Added code to make require_files with a specific uid setting more usable in
7801 the case where statting the file as root fails - usually a non-root-mounted
7802 NFS file system. When this happens and the failure is EACCES, Exim now
7803 forks a subprocess and does the per-uid checking as the relevant uid.
7805 77. Added process_log_path.
7807 78. If log_file_path was not explicitly set, a setting of check_log_space or
7808 check_log_inodes was ignored.
7810 79. If a space check for the spool or log partitions fails, the incident is now
7811 logged. Of course, in the latter case the data may get lost...
7813 80. Added the %p formatting code to string_format() so that it can be used to
7814 print addresses in debug_print(). Adjusted all the address printing in the
7815 debugging in store.c to use %p rather than %d.
7817 81. There was a concern that a line of code in smtp_in.c could overflow a
7818 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7819 initially expressed, the concern was not well-founded, because trailing
7820 spaces are removed early. However, if the trailing spaces were followed by
7821 a NULL, they did not get removed, so the overflow was possible. Two fixes
7824 (a) I re-wrote the offending code in a cleaner fashion.
7825 (b) If an incoming SMTP command contains a NULL character, it is rejected
7828 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7829 runs initgroups(), so that if the Exim user is in any additional groups,
7830 they will be used during message reception.
7836 The change log for 4.20 and earlier releases has been archived.