1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
73 JH/01 Avoid costly startup code when not strictly needed. This reduces time
74 for some exim process initialisations. It does mean that the logging
75 of TLS configuration problems is only done for the daemon startup.
77 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
79 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
80 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
83 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
84 DKIM signing is being done. Previously a transport_filter would always
85 disable CHUNKING, falling back to traditional DATA.
87 JH/05 Regard command-line receipients as tainted.
89 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
91 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
92 PAM library frees one of the arguments given to it, despite the
93 documentation. Therefore a plain malloc must be used.
95 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
96 on-stack buffers were used, resulting in a taint trap when DSN information
97 copied from a received message was written into the buffer.
99 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
100 the ordering of its ARC headers. This caused a crash.
102 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
103 a new record was being constructed with information from the peer, a trap
106 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
107 installation would get error messages from DMARC verify, when it hit the
108 nonexistent file indicated by the default. Distros wanting DMARC enabled
109 should both provide the file and set the option.
110 Also enforce no DMARC verification for command-line sourced messages.
112 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
113 could, depending on the platform, hang at the STARTTLS response.
115 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
116 message on a connection. Previously if one message had ARC headers and
117 the following one did not, a crash could result when adding an
118 Authentication-Results: header.
120 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
121 Exim main code and Exim-related utities. The introduction of taint
122 tracking also did many adjustments to string handling. Since then, eximon
123 frequently terminated with an assert failure.
125 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
126 check for 452 responses. This slightly helps the inefficieny of doing
127 a large alias-expansion into a recipient-limited target. The max_rcpt
128 transport option still applies (and at the current default, will override
129 the new feature). The check is done for either cause of synch, and forces
130 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
131 connection. The new facility is not tunable at this time.
133 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
134 library live data was being used, so the results became garbage. Make
135 copies while it is still usable.
137 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
138 on delivery deferred (==) and failed (**) lines (if a delivery was
139 attemtped). Previously it was only on completion (=>) lines.
141 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
142 for the expansion of the server_scram_iter and server_scram_salt options.
144 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
145 are now specifically given a NO_DATA response without hitting the system
146 resolver. The library goes on to do the now-standard TXT lookup.
147 Use of dnsdb lookups is not affected.
149 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
150 only retrieve the errormessage once. Previously two calls to dlerror()
151 were used, and the second one (for mainlog/paniclog) retrieved null
154 JH/20 Taint checking: disallow use of tainted data for
155 - the appendfile transport file and directory options
156 - the pipe transport command
157 - the autoreply transport file, log and once options
158 - file names used by the redirect router (including filter files)
160 - paths used by single-key lookups
161 Previously this was permitted.
163 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
164 adjusted the size of a major service buffer; this failed because the
165 buffer was in use at the time. Change to a compile-time increase in the
166 buffer size, when this authenticator is compiled into exim.
168 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
169 previous fast-mode was untenable in the face of glibs using mmap to
170 support larger malloc requests.
172 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
173 New values supported, if defined on system where compiled:
174 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
175 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
177 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
178 running a limited number of proceses in parallel, a benefit is gained. The
179 amount varies with the platform hardware and load. The use of the option
180 queue_run_in_order means we cannot do this, as ordering becomes
183 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
184 had introduced a string-copy (for ensuring NUL-termination) which was not
185 appropriate for that case, which can include embedded NUL bytes in the
186 block of data. Investigation showed the copy to actually be needless, the
187 data being length-specified.
189 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
190 done during a receiving connection, and both used TLS, global info was
191 used rather than per-connection info for tracking the state of data
192 queued for transmission. This could result in a connection hang.
194 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
195 Previously, when delivering serveral messages down a single connection
196 only the first would provide a SIZE. This was due to the size information
197 not being properly tracked.
199 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
200 TAI (at 37 seconds currently), pretend to be in UTC for time-related
201 expansion and logging. Previously, spurious values such as a future
202 minute could be seen.
204 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
205 it could crash from a null-deref. This could also affect the
206 ${addresses: } operator and ${readsock } item.
208 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
209 message following a mime one, the variable was not reset.
211 JH/30 When an pipelined-connect fails at the first response, assume incorrect
212 cached capability (perhaps the peer reneged?) and immediately retry in
215 JH/31 Fix spurious detection of timeout while writing to transport filter.
217 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
218 an attempt to copy the string was made before checking it.
220 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
221 taint of the lookup key was maintained; we now regard the presence in the
222 filesystem as sufficient validation.
224 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
225 argument is supplied.
227 JH/35 The dsearch lookup now requires that the directory is an absolute path.
228 Previously this was not checked, and nonempty relative paths made an
229 access under Exim's current working directory.
231 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
232 Previously no event was raised.
234 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
235 parameter supplied by the sender MAIL FROM command. Previously it was
236 ignored, and only the check_spool_space option value for the required
239 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
240 the size of the signing public-key. Previously it was instead giving
241 the size of the signature hash.
243 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
244 the default. See the (new) dkim_verify_min_keysizes option.
246 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
247 and an ACL use a lookup for checking either the local_part or domain,
248 stale data could be accessed. Ensure that variable references are
249 dropped between messages.
251 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
252 by the client was not checked as pointing within response data before
253 being used. A malicious client could thus cause an out-of-bounds read and
254 possibly gain authentication. Fix by adding the check.
256 JH/42 Internationalisation: change the default for downconversion in the smtp
257 transport to be "if needed". Previously it was "as previously set" for
258 the message, which usually meant "if needed" for message-submission but
259 "no" for everything else. However, MTAs have been seen using SMTPUTF8
260 even when the envelope addresses did not need it, resulting in forwarding
261 failures to non-supporting MTAs. A downconvert in such cases will be
262 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
263 transport. The change does mean that addresses needing conversion will
264 be converted when previously a delivery failure would occur.
266 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
267 response was received it would be used unchecked in a fail-DSN, violating
268 standards on line-length limits. Truncate if needed.
270 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
271 doesn't work. (Additionally add a new main config option to configure the
278 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
279 side, to decode a TLS 1.3 packet capture.
281 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
282 Previously the default library behaviour applied, sending two, each in
285 JH/03 Debug output for ACL now gives the config file name and line number for
288 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
290 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
292 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
293 buffer overrun for (non-chunking) other transports.
295 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
296 TLS1.3, means that a server rejecting a client certificate is not visible
297 to the client until the first read of encrypted data (typically the
298 response to EHLO). Add detection for that case and treat it as a failed
299 TLS connection attempt, so that the normal retry-in-clear can work (if
300 suitably configured).
302 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
303 and/or domain. Found and fixed by Jason Betts.
305 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
306 configuration). If a CNAME target was not a wellformed name pattern, a
309 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
310 the OS reports them interleaved with other addresses.
312 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
313 used both for input and for a verify callout, both encrypted, SMTP
314 responses being sent by the server could be lost. This resulted in
315 dropped connections and sometimes bounces generated by a peer sending
318 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
319 string. Previously it was possible to cause undefined behaviour in a
320 library routine (usually a crash). Found by "zerons".
322 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
325 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
326 API was removed, so update to use the newer ones.
328 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
329 any timeout set, is taking a long time. Previously we would hang on to a
330 rotated logfile "forever" if the input was arriving with long gaps
331 (a previous attempt to fix addressed lack, for a long time, of initial
334 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
335 shared (NFS) environment. The length of the tempfile name is now
336 4 + 16 ("hdr.$message_exim_id") which might break on file
337 systems which restrict the file name length to lower values.
340 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
341 shared (NFS) environment.
343 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
344 did for all versions <4.90). Notably -M, -m, --invert, -I may be
347 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
348 on some platforms for bit 31.
350 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
351 to changes apparently associated with TLS1.3 handling some of the APIs
352 previously used were either nonfunctional or inappropriate. Strings
353 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
354 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
355 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
356 This affects log line X= elements, the $tls_{in,out}_cipher variables,
357 and the use of specific cipher names in the encrypted= ACL condition.
359 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
361 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
362 verification result was not updated unless hosts_require_ocsp applied.
364 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
365 queue_list_requires_admin set to false, non-admin users were denied the
368 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
369 directory-of-certs mode. Previously they were advertised despite the
372 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
373 A single TCP connection by a client will now hold a TLS connection open
374 for multiple message deliveries, by default. Previously the default was to
377 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
378 default. If built with the facility, DANE will be used. The facility
379 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
381 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
382 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
383 must be defined and you must still, unless you define DISABLE_TLS, manage
384 the the include-dir and library-file requirements that go with that
385 choice. Non-TLS builds are still supported.
387 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
390 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
391 default. If the platform supports and has the facility enabled, it will
392 be requested on all coneections.
394 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
395 controlled by the build-time option SUPPORT_PIPE_CONNECT.
397 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
399 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
400 success-DSN messages. Previously the From: header was always the default
401 one for these; the option was ignored.
403 JH/28 Fix the timeout on smtp response to apply to the whole response.
404 Previously it was reset for every read, so a teergrubing peer sending
405 single bytes within the time limit could extend the connection for a
406 long time. Credit to Qualsys Security Advisory Team for the discovery.
408 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
409 delivery address, which leaked information of the results of local
410 forwarding. Change to the original envelope recipient address, per
413 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
414 requested. Previously not bounce was generated and a log entry of
415 error ignored was made.
417 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
419 JH/32 Introduce a general tainting mechanism for values read from the input
420 channel, and values derived from them. Refuse to expand any tainted
421 values, to catch one form of exploit.
423 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
424 was unused and the unexpanded text used for the test. Found and
425 fixed by Ruben Jenster.
427 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
428 an attempt to use a TLS library read routine dereffed a nul pointer,
431 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
432 them in our smtp response.
434 JH/36 Have the general router option retry_use_local_part default to true when
435 any of the restrictive preconditions are set (to anything). Previously it
436 was only for check_local user. The change removes one item of manual
437 configuration which is required for proper retries when a remote router
438 handles a subset of addresses for a domain.
440 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
441 link count into consideration.
443 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
444 caused the extension of big_buffer, the following lines were ignored.
446 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
447 accordance with RFC 2308. Previously there was no expiry, so a longlived
448 receive process (eg. due to ACL delays) versus a short SOA value could
451 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
453 JH/39 Promote DMARC support to mainline.
455 JH/40 Bug 2452: Add a References: header to DSNs.
457 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
458 parameters. The relevant library call is documented as "Deprecated: This
459 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
460 3.6.0, DH parameters are negotiated following RFC7919."
462 HS/06 Change the default of dnssec_request_domains to "*"
464 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
465 carried on and emitted a BDAT command, even when PIPELINING was not
468 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
469 buffer was used for the filename, resulting in a trap when tainted
470 arguments (eg. $domain) were used.
472 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
473 recommended to avoid a possible server-load attack. The feature can be
474 re-enabled via the openssl_options main cofiguration option.
476 JH/45 local_scan API: documented the current smtp_printf() call. This changed
477 for version 4.90 - adding a "more data" boolean to the arguments.
478 Bumped the ABI version number also, this having been missed previously;
479 release versions 4.90 to 4.92.3 inclusive were effectively broken in
480 respect of usage of smtp_printf() by either local_scan code or libraries
481 accessed via the ${dlfunc } expansion item. Both will need coding
482 adjustment for any calls to smtp_printf() to match the new function
483 signature; a FALSE value for the new argument is always safe.
485 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
486 the file-offset (which the Linux syscall does, and exim expects); this
487 resulted in an indefinite loop.
489 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
490 to do ARC verification. The Authentication-Results: header line added
491 by the configuration then had no ARC item.
497 JH/01 Remove code calling the customisable local_scan function, unless a new
498 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
500 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
501 non-signal-safe functions being used.
503 JH/03 Bug 2269: When presented with a received message having a stupidly large
504 number of DKIM-Signature headers, disable DKIM verification to avoid
505 a resource-consumption attack. The limit is set at twenty.
507 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
508 report of oldest_pass in ${authres } in consequence, and separate out
509 some descriptions of reasons for verification fail.
511 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
512 files in the spool were present and unlocked. A queue-runner could spot
513 them, resulting in a duplicate delivery. Fix that by doing the unlock
514 after the unlink. Investigation by Tim Stewart. Take the opportunity to
515 add more error-checking on spoolfile handling while that code is being
518 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
519 No known attacks, no CVE, this is defensive hardening.
521 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
522 a queue-runner could start a delivery while other operations were ongoing.
523 Cutthrough delivery was a common victim, resulting in duplicate delivery.
524 Found and investigated by Tim Stewart. Fix by using the open message data
525 file handle rather than opening another, and not locally closing it (which
526 releases a lock) for that case, while creating the temporary .eml format
527 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
529 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
530 $sender_verify_failure/$recipient_verify_failure to "random".
532 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
535 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
536 Previously this would segfault.
538 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
541 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
542 like zero, since the resolver should be doing this for us, But we need one
543 as a CNAME but no MX presence gets the CNAME returned; we need to check
544 that doesn't point to an MX to declare it "no MX returned" rather than
545 "error, loop". A new main option is added so the older capability of
546 following some limited number of chain links is maintained.
548 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
550 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
551 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
552 now-deprecated earlier definitions used only specified the range up to TLS
553 1.2 (in the older-version library docs).
555 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
557 JH/15 Rework TLS client-side context management. Stop using a global, and
558 explicitly pass a context around. This enables future use of TLS for
559 connections to service-daemons (eg. malware scanning) while a client smtp
560 connection is using TLS; with cutthrough connections this is quite likely.
562 JH/16 Fix ARC verification to do AS checks in reverse order.
564 JH/17 Support a "tls" option on the ${readsocket } expansion item.
566 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
567 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
568 Previously the "utf8" would be re-prepended for every additional message.
570 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
571 Previously thery were accepted, resulting in issues when attempting to
572 forward messages to a non-supporting MTA.
574 PP/02 Let -n work with printing macros too, not just options.
576 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
577 one parent address was copied, and bogus data was used at delivery-logging
578 time. Either a crash (after delivery) or bogus log data could result.
579 Discovery and analysis by Tim Stewart.
581 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
582 Previously if the string ended mid-character, we did not insert the
583 promised '?' replacement.
585 PP/04 Documentation: current string operators work on bytes, not codepoints.
587 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
588 should pack well giving a smaller memory footprint so better caching and
589 therefore performance. Group the declarations where this can't be done so
590 that the byte-sized flag variables are not interspersed among pointer
591 variables, giving a better chance of good packing by the compiler.
593 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
594 non-null, to avoid issues with sites running BATV. Previously reports were
595 sent with an empty envelope sender so looked like bounces.
597 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
598 The ignore_error flag wasn't being returned from the filter subprocess so
599 was not set for later routers. Investigation and fix by Matthias Kurz.
601 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
602 and a msg:complete for the whole, when a message is manually removed using
603 -Mrm. Developement by Matthias Kurz, hacked on by JH.
605 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
606 a "Gnu special" function, asprintf() in the DB utility binary builds; I
607 hope that is portable enough.
609 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
610 requiring a known-CA anchor certificate; make it now rely entirely on the
611 TLSA as an anchor. Checking the name on the leaf cert against the name
612 on the A-record for the host is still done for TA (but not for EE mode).
614 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
615 included in delivery lines for non-proxied connections, when compiled with
616 SUPPORT_SOCKS and running with proxy logging enabled.
618 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
619 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
620 move the existing event to fire before the normal logging of message
621 failure so that custom logging is bracketed by normal logging.
623 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
624 msg:fail:internal event. Developement by Matthias Kurz.
626 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
627 far too small for todays use of crypto signatures stored there. Go all
628 the way to the max DNS message size of 64kB, even though this might be
629 overmuch for IOT constrained device use.
631 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
632 copy a string over itself. The library routine is documented as not
633 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
635 JH/32 For main options check_spool_space and check_inode_space, where the
636 platform supports 64b integers, support more than the previous 2^31 kB
637 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
638 the previous G, M, k.
640 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
641 $authenticated_fail_id variable on authentication failure. Previously
644 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
645 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
646 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
647 GNUTLS_SEC_PARAM_MEDIUM.
649 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
650 side. Previously we would continue as if no SNI had been received.
652 JH/36 Harden the handling of string-lists. When a list consisted of a sole
653 "<" character, which should be a list-separator specification, we walked
654 off past the nul-terimation.
656 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
657 causes) even when the retry time is not yet met. Previously they were
658 not, meaning that when (say) an account was over-quota and temp-rejecting,
659 and multiple senders' messages were queued, only one sender would get
660 notified on each configured delay_warning cycle.
662 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
664 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
665 cipher-suites, an error can be left on the stack even for a succeeding
666 accept; this results in impossible error messages when a later operation
669 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
670 return error codes indicating retry. Under TLS1.3 this becomes required.
672 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
673 it only wrote the new authenticators, resulting in a lack of tracking of
674 peer changes of ESMTP extensions until the next cache flush.
676 JH/41 Fix the loop reading a message header line to check for integer overflow,
677 and more-often against header_maxsize. Previously a crafted message could
678 induce a crash of the recive process; now the message is cleanly rejected.
680 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
681 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
687 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
688 When redis_servers is set to a list of > 1 element, and the Redis servers
689 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
690 case of MOVED into a DEFER case instead, thus moving the query onto the
691 next server in the list. For a cluster of N elements, all N servers must
692 be defined in redis_servers.
694 GF/02 Catch and remove uninitialized value warning in exiqsumm
695 Check for existence of @ARGV before looking at $ARGV[0]
697 JH/01 Replace the store_release() internal interface with store_newblock(),
698 which internalises the check required to safely use the old one, plus
699 the allocate and data copy operations duplicated in both (!) of the
700 extant use locations.
702 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
703 modifier. This matches the restriction on the commandline.
705 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
706 Previously only the last row was returned.
708 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
709 we assumed that tags in the header were well-formed, and parsed the
710 element content after inspecting only the first char of the tag.
711 Assumptions at that stage could crash the receive process on malformed
714 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
715 While running the DKIM ACL we operate on the Permanent memory pool so that
716 variables created with "set" persist to the DATA ACL. Also (at any time)
717 DNS lookups that fail create cache records using the Permanent pool. But
718 expansions release any allocations made on the current pool - so a dnsdb
719 lookup expansion done in the DKIM ACL releases the memory used for the
720 DNS negative-cache, and bad things result. Solution is to switch to the
721 Main pool for expansions.
722 While we're in that code, add checks on the DNS cache during store_reset,
723 active in the testsuite.
724 Problem spotted, and debugging aided, by Wolfgang Breyha.
726 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
727 When none of the hosts presented to a transport match an already-open
728 connection, close it and proceed with the list. Previously we would
729 queue the message. Spotted by Lena with Yahoo, probably involving
732 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
733 Previously a spurious "250 OK id=" response was appended to the proper
736 JH/08 The "support for" informational output now, which built with Content
737 Scanning support, has a line for the malware scanner interfaces compiled
738 in. Interface can be individually included or not at build time.
740 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
741 by the template makefile "src/EDITME". The "STREAM" support for an older
742 ClamAV interface method is removed.
744 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
745 rows affected is given instead).
747 JH/11 The runtime Berkeley DB library version is now additionally output by
748 "exim -d -bV". Previously only the compile-time version was shown.
750 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
751 SMTP connection. Previously, when one had more receipients than the
752 first, an abortive onward connection was made. Move to full support for
753 multiple onward connections in sequence, handling cutthrough connection
754 for all multi-message initiating connections.
756 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
757 routers. Previously, a multi-recipient message would fail to match the
758 onward-connection opened for the first recipient, and cause its closure.
760 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
761 a timeout on read on a GnuTLS initiating connection, resulting in the
762 initiating connection being dropped. This mattered most when the callout
763 was marked defer_ok. Fix to keep the two timeout-detection methods
766 JH/15 Relax results from ACL control request to enable cutthrough, in
767 unsupported situations, from error to silently (except under debug)
768 ignoring. This covers use with PRDR, frozen messages, queue-only and
771 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
773 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
774 metadata, resulting in a crash in free().
776 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
777 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
778 Broken also in d185889f4, with init system revamp.
780 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
781 Previously we abruptly closed the connection after reading a malware-
782 found indication; now we go on to read the "scan ok" response line,
785 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
786 ACL. Previously, a crash would result.
788 JH/19 Speed up macro lookups during configuration file read, by skipping non-
789 macro text after a replacement (previously it was only once per line) and
790 by skipping builtin macros when searching for an uppercase lead character.
792 JH/20 DANE support moved from Experimental to mainline. The Makefile control
793 for the build is renamed.
795 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
796 was allocated for every new TLS startup, meaning one per message. Fix
797 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
799 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
800 reported the original. Fix to report (as far as possible) the ACL
801 result replacing the original.
803 JH/23 Fix memory leak during multi-message connections using STARTTLS under
804 OpenSSL. Certificate information is loaded for every new TLS startup,
805 and the resources needed to be freed.
807 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
809 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
812 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
813 DATA response info to the (existing) per-recipient response info for
814 the "C=" log element. It can have useful tracking info from the
815 destination system. Patch from Simon Arlott.
817 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
818 length value. Previously this would segfault.
820 HS/02 Support Avast multiline protoocol, this allows passing flags to
821 newer versions of the scanner.
823 JH/28 Ensure that variables possibly set during message acceptance are marked
824 dead before release of memory in the daemon loop. This stops complaints
825 about them when the debug_store option is enabled. Discovered specifically
826 for sender_rate_period, but applies to a whole set of variables.
827 Do the same for the queue-runner and queue-list loops, for variables set
828 from spool message files. Do the same for the SMTP per-message loop, for
829 certain variables indirectly set in ACL operations.
831 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
832 as a multi-recipient message from a mailinglist manager). The coding had
833 an arbitrary cutoff number of characters while checking for more input;
834 enforced by writing a NUL into the buffer. This corrupted long / fast
835 input. The problem was exposed more widely when more pipelineing of SMTP
836 responses was introduced, and one Exim system was feeding another.
837 The symptom is log complaints of SMTP syntax error (NUL chars) on the
838 receiving system, and refused recipients seen by the sending system
839 (propating to people being dropped from mailing lists).
840 Discovered and pinpointed by David Carter.
842 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
843 replaced by the ${authresults } expansion.
845 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
847 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
848 allows proper process termination in container environments.
850 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
851 Previously the "final dot" had a newline after it; ensure it is CR,LF.
853 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
854 and "err_perm", deprecated since 4.83 when the RFC-defined words
855 "temperror" and "permerror" were introduced.
857 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
858 transport-filters or DKIM-signing. The restriction was lost in the
859 consolidation of verify-callout and delivery SMTP handling.
860 Extend the restriction to also cover ARC-signing.
862 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
863 in defer=pass mode supply a 450 to the initiator. Previously the message
866 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
867 tls_require_ciphers is used as before.
869 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
870 "pass_unscanned". Only tmpfails from the scanner are written to
871 the paniclog, as they may require admin intervention (permission
872 denied, license issues). Other scanner errors (like decompression
873 bombs) do not cause a paniclog entry.
875 JH/36 Fix reinitialisation of DKIM logging variable between messages.
876 Previously it was possible to log spurious information in receive log
879 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
880 triggered odd behaviour from Outlook Express clients.
882 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
885 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
886 since the IETF WG has not yet settled on that versus the original
887 "bare" representation.
889 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
890 Previously the millisecond value corrupted the output.
891 Fix also for syslog_pid=no and log_selector +pid, for which the pid
892 corrupted the output.
898 JH/01 Rework error string handling in TLS interface so that the caller in
899 more cases is responsible for logging. This permits library-sourced
900 string to be attached to addresses during delivery, and collapses
901 pairs of long lines into single ones.
903 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
904 during configuration. Wildcards are allowed and expanded.
906 JH/02 Rework error string handling in DKIM to pass more info back to callers.
907 This permits better logging.
909 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
910 do not close it down and have the child transport start it up again on
911 the passed-on TCP connection. Instead, proxy the child (and any
912 subsequent ones) for TLS via a unix-domain socket channel. Logging is
913 affected: the continued delivery log lines do not have any DNSSEC, TLS
914 Certificate or OCSP information. TLS cipher information is still logged.
916 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
917 identical IP addresses on different listening ports. Will also affect
920 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
921 add noisy ifdef guards to special-case this sillyness.
922 Patch from Bernd Kuhls.
924 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
925 than 255 are no longer allowed.
927 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
928 Disable the session-cache too, which might reduce our load. Since we
929 currrectly use a new context for every connection, both as server and
930 client, there is no benefit for these.
931 GnuTLS appears to not support tickets server-side by default (we don't
932 call gnutls_session_ticket_enable_server()) but client side is enabled
933 by default on recent versions (3.1.3 +) unless the PFS priority string
936 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
937 <https://reproducible-builds.org/specs/source-date-epoch/>.
939 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
940 the check for any unsuccessful recipients did not notice the limit, and
941 erroneously found still-pending ones.
943 JH/08 Pipeline CHUNKING command and data together, on kernels that support
944 MSG_MORE. Only in-clear (not on TLS connections).
946 JH/09 Avoid using a temporary file during transport using dkim. Unless a
947 transport-filter is involved we can buffer the headers in memory for
948 creating the signature, and read the spool data file once for the
949 signature and again for transmission.
951 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
952 4.77 as the kernel support then wasn't solid, having issues in 64bit
953 mode. Now, it's been long enough. Add support for FreeBSD also.
955 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
956 case where the routing stage had gathered several addresses to send to
957 a host before calling the transport for the first, we previously failed
958 to close down TLS in the old transport process before passing the TCP
959 connection to the new process. The new one sent a STARTTLS command
960 which naturally failed, giving a failed delivery and bloating the retry
961 database. Investigation and fix prototype from Wolfgang Breyha.
963 JH/12 Fix check on SMTP command input synchronisation. Previously there were
964 false-negatives in the check that the sender had not preempted a response
965 or prompt from Exim (running as a server), due to that code's lack of
966 awareness of the SMTP input buffering.
968 PP/04 Add commandline_checks_require_admin option.
969 Exim drops privileges sanely, various checks such as -be aren't a
970 security problem, as long as you trust local users with access to their
971 own account. When invoked by services which pass untrusted data to
972 Exim, this might be an issue. Set this option in main configuration
973 AND make fixes to the calling application, such as using `--` to stop
976 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
977 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
978 responses to those, into a single TLS record each way (this usually means
979 a single packet). As a side issue, smtp_enforce_sync now works on TLS
982 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
983 affects you only if you're dancing at the edge of the param size limits.
984 If you are, and this message makes sense to you, then: raise the
985 configured limit or use OpenSSL 1.1. Nothing we can do for older
988 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
989 cmdline element to get the documented default one. Previously it was
992 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
993 are unsuable for non-message-specific reasons - eg. connection timeout,
994 banner-time rejection.
996 JH/16 Fix logging of delivery remote port, when specified by router, under
999 PP/06 Repair manualroute's ability to take options in any order, even if one
1000 is the name of a transport.
1003 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1005 JH/17 Change the list-building routines interface to use the expanding-string
1006 triplet model, for better allocation and copying behaviour.
1008 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1009 Previously it was constructed the first time a possibly-matching string
1010 was met in the configuration file input during startup; now it is done
1013 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1014 compatible one, to avoid the (poorly documented) possibility of a config
1015 file in the working directory redirecting the DB files, possibly correpting
1016 some existing file. CVE-2017-10140 assigned for BDB.
1018 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1019 cache-hot. Previously, although the result was properly cached, the
1020 initial verify call returned a defer.
1022 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1023 the main verify for receipient in uncached-mode.
1025 JH/22 Retire historical build files to an "unsupported" subdir. These are
1026 defined as "ones for which we have no current evidence of testing".
1028 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1029 if present. Previously it was ignored.
1031 JH/24 Start using specified-initialisers in C structure init coding. This is
1032 a C99 feature (it's 2017, so now considered safe).
1034 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1035 if was a fixed-sized field and bitmask ops via macros; it is now more
1038 PP/07 GitHub PR 56: Apply MariaDB build fix.
1039 Patch provided by Jaroslav Škarvada.
1041 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1042 during Coverity cleanups [4.87 JH/47]
1043 Diagnosis and fix provided by Michael Fischer v. Mollard.
1045 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1046 the right size to place the terminating semicolon on its own folded
1047 line, the header hash was calculated to an incorrect value thanks to
1048 the (relaxed) space the fold became.
1050 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1051 and confused the parent.
1053 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1054 which could crash as a result. This could lead to undeliverable messages.
1056 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1059 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1060 matters on fast-turnover and PID-randomising systems, which were getting
1061 out-of-order delivery.
1063 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1064 a possibly-overlapping copy. The symptom was that "Remote host closed
1065 connection in response to HELO" was logged instead of the actual 4xx
1068 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1069 Previously only that bufferd was discarded, resulting in SYMTP command
1072 JH/32 DKIM: when a message has multiple signatures matching an identity given
1073 in dkim_verify_signers, run the dkim acl once for each. Previously only
1074 one run was done. Bug 2189.
1076 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1077 "panic the current process" to "deliberately defer". The panic log is
1078 still written with the problem list name; the mail and reject logs now
1079 get a temp-reject line for the message that was being handled, saying
1080 something like "domains check lookup or other defer". The SMTP 451
1081 message is still "Temporary local problem".
1083 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1084 A crafted sequence of BDAT commands could result in in-use memory beeing
1085 freed. CVE-2017-16943.
1087 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1088 from SMTP input. Previously it was always done; now only done for DATA
1089 and not BDAT commands. CVE-2017-16944.
1091 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1092 to the message (such as an overlong header line). Previously this was
1093 not done and we did not exit BDAT mode. Followon from the previous item
1094 though a different problem.
1100 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1101 than -2003 did; needs libidn2 in addition to libidn.
1103 JH/02 The path option on a pipe transport is now expanded before use.
1105 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1106 Patch provided by "Björn", documentation fix added too.
1108 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1109 missing a wire-to-host endian conversion.
1111 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1112 close after a BDAT command line could be taken as a following command,
1113 giving a synch failure. Fix by only checking for synch immediately
1114 before acknowledging the chunk.
1116 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1117 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1118 macro. Patches provided by Josh Soref.
1120 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1121 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1122 by IANA as a service extension. However, John Klensin suggests that we
1125 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1126 the dkim code may be unix-mode line endings rather than smtp wire-format
1127 CRLF, so prepend a CR to any bare LF.
1129 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1130 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1132 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1133 the first were themselves being wrongly included in the feed into dkim
1134 processing; with most chunk sizes in use this resulted in an incorrect
1135 body hash calculated value.
1137 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1138 DKIM signature block, for verification. Although advised against by
1139 standards it is specifically not ruled illegal.
1141 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1143 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1144 missing a body hash (the bh= tag).
1146 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1147 It seems that HAProxy sends the Proxy Protocol information in clear and
1148 only then does a TLS startup, so do the same.
1150 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1151 TCP connections (such as for Spamd) unless the daemon successfully set
1152 Fast Open mode on its listening sockets. This fixes breakage seen on
1153 too-old kernels or those not configured for Fast Open, at the cost of
1154 requiring both directions being enabled for TFO, and TFO never being used
1155 by non-daemon-related Exim processes.
1157 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1158 endings, at least on the first header line. Try to canonify any that get
1159 past that check, despite the cost.
1161 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1162 now limited to an arbitrary five deep, while parsing addresses with the
1163 strip_excess_angle_brackets option enabled.
1165 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1166 instead leave the unprompted TLS handshake in socket buffer for the
1167 TLS library to consume.
1169 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1171 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1173 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1174 region in one operation, for speed, and this leaves assigned pointers
1175 dangling. Add checks run only under the testsuite which checks all
1176 variables at a store-reset and panics on a dangling pointer; add code
1177 explicitly nulling out all the variables discovered. Fixes one known
1178 bug: a transport crash, where a dangling pointer for $sending_ip_address
1179 originally assigned in a verify callout, is re-used.
1181 PP/06 Drop '.' from @INC in various Perl scripts.
1183 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1185 PP/08 Reduce a number of compilation warnings under clang; building with
1186 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1187 should be warning-free.
1189 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1191 HS/01 Fix portability problems introduced by PP/08 for platforms where
1192 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1194 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1195 chunk. This allows us to accept broken chunked messages. We need a more
1196 general solution here.
1198 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1199 already-broken messages in the queue.
1201 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1203 JH/19 Fix reference counting bug in routing-generated-address tracking.
1209 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1210 supports it and a size is available (ie. the sending peer gave us one).
1212 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1213 years of being deprecated). The replacements are the ACLs
1214 acl_smtp_mime and acl_not_smtp_mime.
1216 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1217 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1218 or even an in-clear connection were permitted. Now, if the host lookup
1219 was dnssec and dane was requested then the host is only used if the
1220 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1221 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1222 if one fails this test.
1223 This means that a poorly-configured remote DNS will make it incommunicado;
1224 but it protects against a DNS-interception attack on it.
1226 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1227 non-noisy when a race steals the message being considered.
1229 JH/05 If main configuration option tls_certificate is unset, generate a
1230 self-signed certificate for inbound TLS connections.
1232 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1233 in rewrites and routers.
1235 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1236 and logged a warning sing 4.83; now they are a configuration file error.
1238 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1239 (lacking @domain). Apply the same qualification processing as RCPT.
1241 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1243 JH/10 Support ${sha256:} applied to a string (as well as the previous
1246 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1247 a cutthrough deliver is pending, as we always want to make a connection.
1248 This also avoids re-routing the message when later placing the cutthrough
1249 connection after a verify cache hit.
1250 Do not update it with the verify result either.
1252 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1253 when routing results in more than one destination address.
1255 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1256 signing (which inhibits the cutthrough capability). Previously only
1257 the presence of an option was tested; now an expansion evaluating as
1258 empty is permissible (obviously it should depend only on data available
1259 when the cutthrough connection is made).
1261 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1262 the relevant preceding SMTP command did not note the pipelining mode.
1264 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1265 Previously they were not counted.
1267 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1268 as one having no matching records. Previously we deferred the message
1269 that needed the lookup.
1271 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1272 distinguished as "(=".
1274 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1275 for missing MX records. Previously it only worked for missing A records.
1277 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1279 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1280 after the data-go-ahead and data-ack. Patch from Jason Betts.
1282 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1283 even for a "none" policy. Patch from Tony Meyer.
1285 JH/22 Fix continued use of a connection for further deliveries. If a port was
1286 specified by a router, it must also match for the delivery to be
1289 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1290 When one of the recipients of a message was unsuitable for the connection
1291 (has no matching addresses), we lost track of needing to mark it
1292 deferred. As a result mail would be lost.
1294 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1296 JH/25 Decoding ACL controls is now done using a binary search; the source code
1297 takes up less space and should be simpler to maintain. Merge the ACL
1298 condition decode tables also, with similar effect.
1300 JH/26 Fix problem with one_time used on a redirect router which returned the
1301 parent address unchanged. A retry would see the parent address marked as
1302 delivered, so not attempt the (identical) child. As a result mail would
1305 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1306 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1307 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1310 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1311 Main config options check_{log,spool}_{inodes,space} are now
1312 100 inodes, 10MB unless set otherwise in the configuration.
1314 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1315 Previously it only applied to the main-section connection policy
1318 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1320 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1321 by me. Added RFC7919 DH primes as an alternative.
1323 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1324 are not in the system include path.
1326 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1327 GnuTLS, when a session startup failed (eg because the client disconnected)
1328 Exim did stdio operations after fclose. This was exposed by a recent
1329 change which nulled out the file handle after the fclose.
1331 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1332 signed directly by the cert-signing cert, rather than an intermediate
1333 OCSP-signing cert. This is the model used by LetsEncrypt.
1335 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1337 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1338 an incoming connection.
1340 HS/02 Bug 1802: Do not half-close the connection after sending a request
1343 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1344 fallback to "prime256v1".
1346 JH/34 SECURITY: Use proper copy of DATA command in error message.
1347 Could leak key material. Remotely exploitable. CVE-2016-9963.
1353 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1354 and 3.4.4 - once the server is enabled to respond to an OCSP request
1355 it does even when not requested, resulting in a stapling non-aware
1356 client dropping the TLS connection.
1358 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1359 support variable-length bit vectors. No functional change.
1361 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1362 The I= interface field on outgoing lines is now after the H= remote
1363 host field, same as incoming lines. There is a separate
1364 outgoing_interface log selector which allows you to disable the
1367 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1368 If not running log_selector +smtp_connection the mainlog would be held
1369 open indefinitely after a "too many connections" event, including to a
1370 deleted file after a log rotate. Leave the per net connection logging
1371 leaving it open for efficiency as that will be quickly detected by the
1372 check on the next write.
1374 HS/01 Bug 1671: Fix post transport crash.
1375 Processing the wait-<transport> messages could crash the delivery
1376 process if the message IDs didn't exist for some reason. When
1377 using 'split_spool_directory=yes' the construction of the spool
1378 file name failed already, exposing the same netto behaviour.
1380 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1381 mime_regex ACL conditions.
1383 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1384 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1385 to HELO, local diagnostic string.
1387 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1388 log line to debug. Even when configured with a tls authenticator many
1389 client connections are expected to not authenticate in this way, so
1390 an authenticate fail is not an error.
1392 HS/02 Add the Exim version string to the process info. This way exiwhat
1393 gives some more detail about the running daemon.
1395 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1396 matter for fast-change records such as DNSBLs.
1398 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1399 retry record, even if constant. There may be multiple transports with
1400 different interface settings and the retry behaviour needs to be kept
1403 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1405 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1407 JH/10 Bug 840: fix log_defer_output option of pipe transport
1409 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1410 hosts_randomize. This matters a lot when combined with mua_wrapper.
1412 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1413 ${quote_pgsql:<string>} operator.
1415 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1417 JH/14 Change header file name for freeradius-client. Relevant if compiling
1418 with Radius support; from the Gentoo tree and checked under Fedora.
1420 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1422 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1423 option transport_filter, meaning no filtering.
1425 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1427 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1428 defaults to "*" (all hosts). The variable is now available when not built
1429 with TLS, default unset, mainly to enable keeping the testsuite sane.
1430 If a server certificate is not supplied (via tls_certificate) an error is
1431 logged, and clients will find TLS connections fail on startup. Presumably
1432 they will retry in-clear.
1433 Packagers of Exim are strongly encouraged to create a server certificate
1434 at installation time.
1436 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1437 with the $config_file variable.
1439 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1440 in transport context, after the attempt, and per-recipient. The latter type
1441 is per host attempted. The event data is the error message, and the errno
1442 information encodes the lookup type (A vs. MX) used for the (first) host,
1443 and the trailing two digits of the smtp 4xx response.
1445 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1446 to write to mainlog (or rejectlog, paniclog) in the window between file
1447 creation and permissions/ownership being changed. Particularly affects
1448 installations where exicyclog is run as root, rather than exim user;
1449 result is that the running daemon panics and dies.
1451 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1453 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1454 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1455 "pri" and "weight". Note that the previous implicit priority given by the
1456 list order is no longer honoured.
1458 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1459 for DKIM processing.
1461 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1462 by defining SUPPORT_SOCKS.
1464 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1465 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1466 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1467 variables are renamed to proxy_{local,external}_{address,port}.
1469 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1470 for a build by defining SUPPORT_I18N
1472 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1473 of the query string, and make ${quote_redis:} do that quoting.
1475 JH/29 Move Events support from Experimental to mainline, enabled by default
1476 and removable for a build by defining DISABLE_EVENT.
1478 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1480 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1481 cached by the daemon.
1483 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1484 by defining LOOKUP_REDIS. The libhiredis library is required.
1486 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1487 keys are given for lookup.
1489 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1490 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1491 only supported when built with TLS support. The PolarSSL SHA routines
1492 are still used when the TLS library is too old for convenient support.
1494 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1495 openssl_options), for security. OpenSSL forces this from version 1.1.0
1496 server-side so match that on older versions.
1498 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1499 allocation for $value could be released as the expansion processing
1500 concluded, but leaving the global pointer active for it.
1502 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1503 and to use the domains and local_parts ACL conditions.
1505 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1506 incorrectly not doubled on cutthrough transmission, hence seen as a
1507 body-termination at the receiving system - resulting in truncated mails.
1508 Commonly the sender saw a TCP-level error, and retransmitted the message
1509 via the normal store-and-forward channel. This could result in duplicates
1510 received - but deduplicating mailstores were liable to retain only the
1511 initial truncated version.
1513 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1515 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1517 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1518 we're in there, support oversigning also; bug 1309.
1520 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1522 HS/04 Add support for keep_environment and add_environment options.
1524 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1525 either intentional arithmetic overflow during PRNG, or testing config-
1528 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1529 delivery resulted in actual delivery. Cancel cutthrough before DATA
1532 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1533 rejects a recipient: pass the reject to the originator.
1535 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1536 Many were false-positives and ignorable, but it's worth fixing the
1539 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1540 for the new environment-manipulation done at startup. Move the routines
1541 from being local to tls.c to being global via the os.c file.
1543 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1544 an extract embedded as result-arg for a map, the first arg for extract
1545 is unavailable so we cannot tell if this is a numbered or keyed
1546 extraction. Accept either.
1552 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1555 JH/02 The smtp transport option "multi_domain" is now expanded.
1557 JH/03 The smtp transport now requests PRDR by default, if the server offers
1560 JH/04 Certificate name checking on server certificates, when exim is a client,
1561 is now done by default. The transport option tls_verify_cert_hostnames
1562 can be used to disable this per-host. The build option
1563 EXPERIMENTAL_CERTNAMES is withdrawn.
1565 JH/05 The value of the tls_verify_certificates smtp transport and main options
1566 default to the word "system" to access the system default CA bundle.
1567 For GnuTLS, only version 3.0.20 or later.
1569 JH/06 Verification of the server certificate for a TLS connection is now tried
1570 (but not required) by default. The verification status is now logged by
1571 default, for both outbound TLS and client-certificate supplying inbound
1574 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1577 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1578 Status Notification (bounce) messages are now MIME format per RFC 3464.
1579 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1580 under the control of the dsn_advertise_hosts option, and routers may
1581 have a dsn_lasthop option.
1583 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1584 default, modifiable by a malware= option. The list separator for
1585 the options can now be changed in the usual way. Bug 68.
1587 JH/10 The smtp_receive_timeout main option is now expanded before use.
1589 JH/11 The incoming_interface log option now also enables logging of the
1590 local interface on delivery outgoing connections.
1592 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1593 if the interface and destination host and port all match.
1595 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1598 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1599 Patch from Andrew Lewis.
1601 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1602 now supports optional time-restrictions, weighting, and priority
1603 modifiers per server. Patch originally by <rommer@active.by>.
1605 JH/16 The spamd_address main option now supports a mixed list of local
1606 and remote servers. Remote servers can be IPv6 addresses, and
1607 specify a port-range.
1609 JH/17 Bug 68: The spamd_address main option now supports an optional
1610 timeout value per server.
1612 JH/18 Bug 1581: Router and transport options headers_add/remove can
1613 now have the list separator specified.
1615 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1618 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1621 JH/21 Support for the A6 type of dns record is withdrawn.
1623 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1624 rather than the verbs used.
1626 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1627 from 255 to 1024 chars.
1629 JH/24 Verification callouts now attempt to use TLS by default.
1631 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1632 are generic router options now. The defaults didn't change.
1634 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1635 Original patch from Alexander Shikoff, worked over by JH.
1637 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1638 files if ZCAT_COMMAND is not executable.
1640 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1642 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1644 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1645 Normally benign, it bites when the pair was led to by a CNAME;
1646 modern usage is to not canonicalize the domain to a CNAME target
1647 (and we were inconsistent anyway for A-only vs AAAA+A).
1649 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1651 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1652 when evaluating $sender_host_dnssec.
1654 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1655 $sender_helo_dnssec variable.
1657 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1659 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1661 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1663 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1664 documented as working, but never had. Support all but $spam_report.
1666 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1667 added for tls authenticator.
1669 HS/03 Add perl_taintmode main config option
1675 TL/01 When running the test suite, the README says that variables such as
1676 no_msglog_check are global and can be placed anywhere in a specific
1677 test's script, however it was observed that placement needed to be near
1678 the beginning for it to behave that way. Changed the runtest perl
1679 script to read through the entire script once to detect and set these
1680 variables, reset to the beginning of the script, and then run through
1681 the script parsing/test process like normal.
1683 TL/02 The BSD's have an arc4random API. One of the functions to induce
1684 adding randomness was arc4random_stir(), but it has been removed in
1685 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1686 function when detected.
1688 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1689 cause callback expansion.
1691 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1692 syntax errors in an expansion can be treated as a string instead of
1693 logging or causing an error, due to the internal use of bool_lax
1694 instead of bool when processing it.
1696 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1697 server certificates when making smtp deliveries.
1699 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1701 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1703 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1705 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1706 Merged patch from Sebastian Wiedenroth.
1708 JH/05 Fix results-pipe from transport process. Several recipients, combined
1709 with certificate use, exposed issues where response data items split
1710 over buffer boundaries were not parsed properly. This eventually
1711 resulted in duplicates being sent. This issue only became common enough
1712 to notice due to the introduction of connection certificate information,
1713 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1715 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1716 size buffer was used, resulting in syntax errors when an expansion
1719 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1720 version 3.3.6 or later.
1722 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1723 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1724 both become "event_action", the variables become $event_name, $event_data
1725 and $event_defer_errno. There is a new variable $verify_mode, usable in
1726 routers, transports and related events. The tls:cert event is now also
1727 raised for inbound connections, if the main configuration event_action
1730 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1731 early OCSP support, but no stapling (appears to be less than 1.0.0).
1733 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1734 server certificate names available under the smtp transport option
1735 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1738 JH/10 Time-related extraction expansions from certificates now use the main
1739 option "timezone" setting for output formatting, and are consistent
1740 between OpenSSL and GnuTLS compilations. Bug 1541.
1742 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1743 encoded parameter in the incoming message. Bug 1558.
1745 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1746 include certificate info, eximon was claiming there were spoolfile
1749 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1751 JH/14 Log delivery-related information more consistently, using the sequence
1752 "H=<name> [<ip>]" wherever possible.
1754 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1755 are problematic for Debian distribution, omit them from the release
1758 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1760 JH/16 Fix string representation of time values on 64bit time_t architectures.
1763 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1764 output list separator was used.
1769 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1770 checkers that were complaining about end of non-void function with no
1773 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1774 This was a regression introduced in 4.83 by another bugfix.
1776 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1778 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1779 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1785 TF/01 Correctly close the server side of TLS when forking for delivery.
1787 When a message was received over SMTP with TLS, Exim failed to clear up
1788 the incoming connection properly after forking off the child process to
1789 deliver the message. In some situations the subsequent outgoing
1790 delivery connection happened to have the same fd number as the incoming
1791 connection previously had. Exim would try to use TLS and fail, logging
1792 a "Bad file descriptor" error.
1794 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1795 utilities have not been installed.
1797 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1798 temporary space as the ACL may create new global variables.
1800 TL/01 LDAP support uses per connection or global context settings, depending
1801 upon the detected version of the libraries at build time.
1803 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1804 to extract and use the src ip:port in logging and expansions as if it
1805 were a direct connection from the outside internet. PPv2 support was
1806 updated based on HAProxy spec change in May 2014.
1808 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1810 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1811 Properly escape header and check for NULL return.
1813 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1816 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1818 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1819 characters in header names, implemented as a verify condition.
1820 Contributed by Michael Fischer v. Mollard.
1822 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1823 results permerror and temperror. Previous values are deprecated but
1824 still accepted. In a future release, err_perm and err_temp will be
1825 completely removed, which will be a backward incompatibility if the
1826 ACL tests for either of these two old results. Patch contributed by
1827 user bes-internal on the mailing list.
1829 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1831 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1832 selectors, in both main and reject logs.
1834 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1837 JH/07 Add malware type "sock" for talking to simple daemon.
1839 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1841 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1842 routers/transports under cutthrough routing.
1844 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1845 numbers. Touch up "bool" conditional to keep the same definition.
1847 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1849 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1851 JH/12 Expand items in router/transport headers_add or headers_remove lists
1852 individually rather than the list as a whole. Bug 1452.
1854 Required for reasonable handling of multiple headers_ options when
1855 they may be empty; requires that headers_remove items with embedded
1856 colons must have them doubled (or the list-separator changed).
1858 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1859 view the policy declared in the DMARC record. Currently, $dmarc_status
1860 is a combined value of both the record presence and the result of the
1863 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1865 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1866 dnslookup router and the smtp transport (applying to the forward
1869 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1870 of ldap servers used for a specific lookup. Patch provided by Heiko
1873 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1874 New variable $lookup_dnssec_authenticated for observability.
1876 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1877 Patch submitted by Lars Timman.
1879 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1881 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1882 Requires trusted mode and valid format message id, aborts otherwise.
1883 Patch contributed by Heiko Schlichting.
1885 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1886 certextract with support for various fields. Bug 1358.
1888 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1889 is requested by default, modifiable by smtp transport option
1892 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1893 operate on certificate variables to give certificate fingerprints
1894 Also new ${sha256:cert_variable}.
1896 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1898 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1899 Christian Aistleitner.
1901 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1903 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1904 file. Patch from Wolfgang Breyha.
1906 JH/25 Expand the coverage of the delivery $host and $host_address to
1907 client authenticators run in verify callout. Bug 1476.
1909 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1910 align with daemon_smtp_ports. Bug 72.
1912 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1913 support and error reporting did not work properly.
1915 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1916 and is readable. Patch from Andrew Colin Kissa.
1918 TL/14 Enhance documentation of ${run expansion and how it parses the
1919 commandline after expansion, particularly in the case when an
1920 unquoted variable expansion results in an empty value.
1922 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1924 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1925 by renaming away from it. Addresses GH issue 15, reported by
1928 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1930 TL/15 SECURITY: prevent double expansion in math comparison functions
1931 (can expand unsanitized data). Not remotely exploitable.
1938 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1940 PP/02 Make -n do something, by making it not do something.
1941 When combined with -bP, the name of an option is not output.
1943 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1946 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1947 $sender_host_name and config options to manage this, and basic check
1950 PP/05 DSCP support for outbound connections and control modifier for inbound.
1952 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1953 (Only plugin which currently uses this is kerberos4, which nobody should
1954 be using, but we should make it available and other future plugins might
1955 conceivably use it, even though it would break NAT; stuff *should* be
1956 using channel bindings instead).
1958 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1959 name; added for Sendmail compatibility; requires admin caller.
1960 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1961 just ignore it); requires trusted caller.
1962 Also parse but ignore: -Ac -Am -X<logfile>
1965 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1967 TL/02 Add +smtp_confirmation as a default logging option.
1969 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1970 Patch by Magnus Holmgren from 2007-02-20.
1972 TL/04 Bugzilla 1281 - Spec typo.
1973 Bugzilla 1283 - Spec typo.
1974 Bugzilla 1290 - Spec grammar fixes.
1976 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1978 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1980 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1981 mailing list by by Dmitry Isaikin.
1983 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1985 JH/02 Support "G" suffix to numbers in ${if comparisons.
1987 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1989 NM/01 Bugzilla 1197 - Spec typo
1990 Bugzilla 1196 - Spec examples corrections
1992 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1994 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1995 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1998 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1999 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2001 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2002 "acl {{name}{arg}...}", and optional args on acl condition
2005 JH/05 Permit multiple router/transport headers_add/remove lines.
2007 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2009 JH/07 Avoid using a waiting database for a single-message-only transport.
2010 Performance patch from Paul Fisher. Bugzilla 1262.
2012 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2015 JH/09 Add $headers_added variable, with content from use of ACL modifier
2016 add_header (but not yet added to the message). Bugzilla 199.
2018 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2019 Pulled from Bugzilla 817 by Wolfgang Breyha.
2021 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2023 (nb: this is the same fix as in Exim 4.80.1)
2025 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2028 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2031 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2033 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2034 advertise SMTP AUTH mechanism to us, instead of a generic
2035 protocol violation error. Also, make Exim more robust to bad
2036 data from the Dovecot auth socket.
2038 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2040 When a queue runner is handling a message, Exim first routes the
2041 recipient addresses, during which it prunes them based on the retry
2042 hints database. After that it attempts to deliver the message to
2043 any remaining recipients. It then updates the hints database using
2046 So if a recipient address works intermittently, it can get repeatedly
2047 deferred at routing time. The retry hints record remains fresh so the
2048 address never reaches the final cutoff time.
2050 This is a fairly common occurrence when a user is bumping up against
2051 their storage quota. Exim had some logic in its local delivery code
2052 to deal with this. However it did not apply to per-recipient defers
2053 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2055 This change adds a proper retry rule check during routing so that the
2056 final cutoff time is checked against the message's age. We only do
2057 this check if there is an address retry record and there is not a
2058 domain retry record; this implies that previous attempts to handle
2059 the address had the retry_use_local_parts option turned on. We use
2060 this as an approximation for the destination being like a local
2061 delivery, as in LMTP.
2063 I suspect this new check makes the old local delivery cutoff check
2064 redundant, but I have not verified this so I left the code in place.
2066 TF/02 Correct gecos expansion when From: is a prefix of the username.
2068 Test 0254 submits a message to Exim with the header
2072 When I ran the test suite under the user fanf2, Exim expanded
2073 the header to contain my full name, whereas it should have added
2074 a Resent-Sender: header. It erroneously treats any prefix of the
2075 username as equal to the username.
2077 This change corrects that bug.
2079 GF/01 DCC debug and logging tidyup
2080 Error conditions log to paniclog rather than rejectlog.
2081 Debug lines prefixed by "DCC: " to remove any ambiguity.
2083 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2085 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2086 Bug spotted by Jeremy Harris; was flawed since initial commit.
2087 Would have resulted in OCSP responses post-SNI triggering an Exim
2088 NULL dereference and crash.
2090 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2092 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2093 Bug detection, analysis and fix by Samuel Thibault.
2094 Bugzilla 1331, Debian bug #698092.
2096 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2098 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2099 Server implementation by Todd Lyons, client by JH.
2100 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2101 config variable "prdr_enable" controls whether the server
2102 advertises the facility. If the client requests PRDR a new
2103 acl_data_smtp_prdr ACL is called once for each recipient, after
2104 the body content is received and before the acl_smtp_data ACL.
2105 The client is controlled by both of: a hosts_try_prdr option
2106 on the smtp transport, and the server advertisement.
2107 Default client logging of deliveries and rejections involving
2108 PRDR are flagged with the string "PRDR".
2110 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2111 fclose(). Diagnosis by Todd Lyons.
2113 PP/17 Update configure.default to handle IPv6 localhost better.
2114 Patch by Alain Williams (plus minor tweaks).
2117 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2118 This is now consistent with GnuTLS, and is now documented: the
2119 previous undocumented portable approach to treating the option as
2120 unset was to force an expansion failure. That still works, and
2121 an empty string is now equivalent.
2123 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2124 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2125 not performing validation itself.
2127 PP/20 Added force_command boolean option to pipe transport.
2128 Patch from Nick Koston, of cPanel Inc.
2130 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2133 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2135 PP/21 Fix eximon continuous updating with timestamped log-files.
2136 Broken in a format-string cleanup in 4.80, missed when I repaired the
2137 other false fix of the same issue.
2138 Report and fix from Heiko Schlichting.
2141 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2142 Report from Prashanth Katuri.
2144 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2145 It's SecureTransport, so affects any MacOS clients which use the
2146 system-integrated TLS libraries, including email clients.
2148 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2149 using a MIME ACL for non-SMTP local injection.
2150 Report and assistance in diagnosis by Warren Baker.
2152 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2154 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2156 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2157 last id that failed so it may be referenced in subsequent ACL's.
2159 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2162 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2163 ldap library initialization, allowing self-signed CA's to be
2164 used. Also properly sets require_cert option later in code by
2165 using NULL (global ldap config) instead of ldap handle (per
2166 session). Bug diagnosis and testing by alxgomz.
2168 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2169 the src/util/ subdirectory.
2171 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2172 renamed to Transport Post Delivery Action by Jeremy Harris, as
2175 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2176 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2177 redis_servers = needs to be configured which will be used by the redis
2178 lookup. Patch from Warren Baker, of The Packet Hub.
2180 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2182 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2183 hostname or reverse DNS when processing a host list. Used suggestions
2184 from multiple comments on this bug.
2186 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2188 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2189 Missed a few lines, added it to make the runtest require no keyboard
2192 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2193 contains upper case chars. Make router use caseful_local_part.
2195 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2196 support when GnuTLS has been built with p11-kit.
2202 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2204 This, or similar/improved, will also be change PP/11 of 4.82.
2210 PP/01 Handle short writes when writing local log-files.
2211 In practice, only affects FreeBSD (8 onwards).
2212 Bugzilla 1053, with thanks to Dmitry Isaikin.
2214 NM/01 Bugzilla 949 - Documentation tweak
2216 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2219 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2221 PP/02 Implemented gsasl authenticator.
2223 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2225 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2226 `pkg-config foo` for cflags/libs.
2228 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2229 with rest of GSASL and with heimdal_gssapi.
2231 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2232 `pkg-config foo` for cflags/libs for the TLS implementation.
2234 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2235 properties get this fed in as external SSF. A number of robustness
2236 and debugging improvements to the cyrus_sasl authenticator.
2238 PP/08 cyrus_sasl server now expands the server_realm option.
2240 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2241 Patch by Jeremy Harris.
2243 PP/10 Added dbmjz lookup type.
2245 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2247 PP/12 MAIL args handles TAB as well as SP, for better interop with
2248 non-compliant senders.
2249 Analysis and variant patch by Todd Lyons.
2251 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2252 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2253 Patch from Dirk Mueller <dmueller@suse.com>
2255 PP/13 tls_peerdn now print-escaped for spool files.
2256 Observed some $tls_peerdn in wild which contained \n, which resulted
2257 in spool file corruption.
2259 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2260 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2261 or write after TLS renegotiation, which otherwise led to messages
2264 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2265 as a tracking header (ie: a signed header comes before the signature).
2266 Patch from Wolfgang Breyha.
2268 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2269 comma-sep list; embedded commas doubled.
2271 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2273 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2275 Report and patch from Dmitry Banschikov.
2277 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2278 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2279 failures. We appear to now support TLS1.1+ with Exim.
2281 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2282 lets Exim select keys and certificates based upon TLS SNI from client.
2283 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2284 before an outbound SMTP session. New log_selector, +tls_sni.
2286 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2287 NULL dereference. Report and patch from Alun Jones.
2289 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2290 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2291 Not seeing resolver debug output on NetBSD, but suspect this is a
2292 resolver implementation change.
2294 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2295 Left warnings. Added "eximon gdb" invocation mode.
2297 PP/21 Defaulting "accept_8bitmime" to true, not false.
2299 PP/22 Added -bw for inetd wait mode support.
2301 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2302 locate the relevant includes and libraries. Made this the default.
2304 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2305 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2307 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2308 This may cause build issues on older platforms.
2310 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2311 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2312 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2313 Added SNI support via GnuTLS too.
2314 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2316 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2318 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2319 Applied second patch from Janne, implementing suggestion to default
2320 multiple-strings-in-record handling to match SPF spec.
2322 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2324 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2325 read-only, out of scope).
2326 Patch from Wolfgang Breyha, report from Stuart Northfield.
2328 PP/29 Fix three issues highlighted by clang analyser static analysis.
2329 Only crash-plausible issue would require the Cambridge-specific
2330 iplookup router and a misconfiguration.
2331 Report from Marcin Mirosław.
2333 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2335 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2336 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2337 As part of this, removing so much warning spew let me fix some minor
2338 real issues in debug logging.
2340 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2341 assignment on my part. Fixed.
2343 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2344 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2345 Janne Snabb (who went above and beyond: thank you).
2347 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2348 string otherwise requires a connection and a bunch more work and it's
2349 relatively easy to get wrong. Should also expose TLS library linkage
2352 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2353 64-bit ${eval} (JH/03).
2355 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2356 GNU libc to support some of the 64-bit stuff, should not lead to
2357 conflicts. Defined before os.h is pulled in, so if a given platform
2358 needs to override this, it can.
2360 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2361 protection layer was required, which is not implemented.
2362 Bugzilla 1254, patch from Wolfgang Breyha.
2364 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2365 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2366 tls_dhparam take prime identifiers. Also unbreak combination of
2367 OpenSSL+DH_params+TLSSNI.
2369 PP/39 Disable SSLv2 by default in OpenSSL support.
2375 PP/01 Solaris build fix for Oracle's LDAP libraries.
2376 Bugzilla 1109, patch from Stephen Usher.
2378 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2380 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2383 TF/02 Fix a couple more cases where we did not log the error message
2384 when unlink() failed. See also change 4.74-TF/03.
2386 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2387 lock up or crash if it happened to be inside a call to libc when it
2388 got a SIGUSR1 from exiwhat.
2390 The SIGUSR1 handler appends the current process status to the process
2391 log which is later printed by exiwhat. It used to use the general
2392 purpose logging code to do this, but several functions it calls are
2393 not safe for signals.
2395 The new output code in the SIGUSR1 handler is specific to the process
2396 log, and simple enough that it's easy to inspect for signal safety.
2397 Removing some special cases also simplifies the general logging code.
2398 Removing the spurious timestamps from the process log simplifies
2401 TF/04 Improved ratelimit ACL condition.
2403 The /noupdate option has been deprecated in favour of /readonly which
2404 has clearer semantics. The /leaky, /strict, and /readonly update modes
2405 are mutually exclusive. The update mode is no longer included in the
2406 database key; it just determines when the database is updated. (This
2407 means that when you upgrade Exim will forget old rate measurements.)
2409 Exim now checks that the per_* options are used with an update mode that
2410 makes sense for the current ACL. For example, when Exim is processing a
2411 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2412 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2413 must specify per_mail/readonly. If you omit the update mode it defaults to
2414 /leaky where that makes sense (as before) or /readonly where required.
2416 The /noupdate option is now undocumented but still supported for
2417 backwards compatibility. It is equivalent to /readonly except that in
2418 ACLs where /readonly is required you may specify /leaky/noupdate or
2419 /strict/noupdate which are treated the same as /readonly.
2421 A useful new feature is the /count= option. This is a generalization
2422 of the per_byte option, so that you can measure the throughput of other
2423 aggregate values. For example, the per_byte option is now equivalent
2424 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2426 The per_rcpt option has been generalized using the /count= mechanism
2427 (though it's more complicated than the per_byte equivalence). When it is
2428 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2429 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2430 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2431 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2432 Note that using per_rcpt with a non-readonly update mode in more than
2433 one ACL will cause the recipients to be double-counted. (The per_mail
2434 and per_byte options don't have this problem.)
2436 The handling of very low rates has changed slightly. If the computed rate
2437 is less than the event's count (usually one) then this event is the first
2438 after a long gap. In this case the rate is set to the same as this event's
2439 count, so that the first message of a spam run is counted properly.
2441 The major new feature is a mechanism for counting the rate of unique
2442 events. The new per_addr option counts the number of different
2443 recipients that someone has sent messages to in the last time period. It
2444 behaves like per_rcpt if all the recipient addresses are different, but
2445 duplicate recipient addresses do not increase the measured rate. Like
2446 the /count= option this is a general mechanism, so the per_addr option
2447 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2448 example, measure the rate that a client uses different sender addresses
2449 with the options per_mail/unique=$sender_address. There are further
2450 details in the main documentation.
2452 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2454 TF/06 Removed a few PCRE remnants.
2456 TF/07 Automatically extract Exim's version number from tags in the git
2457 repository when doing development or release builds.
2459 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2460 Bugzilla 879. Patch from Paul Fisher.
2462 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2463 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2466 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2468 PP/05 Variable $av_failed, true if the AV scanner deferred.
2469 Bugzilla 1078. Patch from John Horne.
2471 PP/06 Stop make process more reliably on build failure.
2472 Bugzilla 1087. Patch from Heiko Schlittermann.
2474 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2475 Bugzilla 1089. Patch from Heiko Schlittermann.
2477 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2478 Bugzilla 1131. Patch from Holger Weiß.
2480 PP/09 Handle IPv6 addresses with SPF.
2481 Bugzilla 860. Patch from Wolfgang Breyha.
2483 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2485 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2488 PP/11 match_* no longer expand right-hand-side by default.
2489 New compile-time build option, EXPAND_LISTMATCH_RHS.
2490 New expansion conditions, "inlist", "inlisti".
2492 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2494 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2496 PP/14 fix log_write() format string regression from TF/03.
2497 Bugzilla 1152. Patch from Dmitry Isaikin.
2503 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2505 PP/02 Harmonised TLS library version reporting; only show if debugging.
2506 Layout now matches that introduced for other libraries in 4.74 PP/03.
2508 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2510 PP/04 New "dns_use_edns0" global option.
2512 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2515 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2516 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2518 TK/01 Updated PolarSSL code to 0.14.2.
2519 Bugzilla 1097. Patch from Andreas Metzler.
2521 PP/07 Catch divide-by-zero in ${eval:...}.
2522 Fixes bugzilla 1102.
2524 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2527 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2528 format-string attack -- SECURITY: remote arbitrary code execution.
2530 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2531 time unintentionally subject to list matching rules, letting the header
2532 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2533 arbitrary string expansion). This allowed for information disclosure.
2535 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2536 INT_MIN/-1 -- value coerced to INT_MAX.
2542 NM/01 Workaround for PCRE version dependency in version reporting
2545 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2546 This fixes portability to compilers other than gcc, notably
2547 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2549 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2550 makefiles for portability to HP-UX and POSIX correctness.
2552 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2553 Also via indented variable definition in the Makefile.
2554 (Debugging by Oliver Heesakkers).
2556 PP/02 Restore caching of spamd results with expanded spamd_address.
2557 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2559 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2560 Improves build reliability. Fix from: Frank Elsner
2562 NM/02 Fix wide character breakage in the rfc2047 coding
2563 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2565 NM/03 Allow underscore in dnslist lookups
2566 Fixes bug 1026. Patch from Graeme Fowler
2568 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2569 Code patches from Adam Ciarcinski of NetBSD.
2571 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2574 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2575 is logged, to avoid truncation. Patch from John Horne.
2577 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2578 Patch from Jakob Hirsch.
2580 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2581 SQL string expansion failure details.
2582 Patch from Andrey Oktyabrski.
2584 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2585 Patch from Simon Arlott.
2587 PP/09 New lookups functionality failed to compile on old gcc which rejects
2588 extern declarations in function scope.
2589 Patch from Oliver Fleischmann
2591 PP/10 Use sig_atomic_t for flags set from signal handlers.
2592 Check getgroups() return and improve debugging.
2593 Fixed developed for diagnosis in bug 927 (which turned out to be
2596 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2597 Patch from Mark Zealey.
2599 PP/12 Bugzilla 1056: Improved spamd server selection.
2600 Patch from Mark Zealey.
2602 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2603 Based on patch from Heiko Schlittermann.
2605 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2606 Patch from Uwe Doering, sign-off by Michael Haardt.
2608 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2609 variable declaration deep within a block. Bug and patch from
2612 PP/15 lookups-Makefile IRIX compatibility coercion.
2614 PP/16 Make DISABLE_DKIM build knob functional.
2616 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2617 Patch by Simon Arlott
2619 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2620 variable argument macros. Our copy now differs from upstream.
2626 TF/01 Failure to get a lock on a hints database can have serious
2627 consequences so log it to the panic log.
2629 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2630 controlled using the smtp_confirmation log selector.
2632 TF/03 Include the error message when we fail to unlink a spool file.
2634 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2635 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2636 for maintaining out-of-tree patches for some time.
2638 PP/01 Bugzilla 139: Documentation and portability issues.
2639 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2640 Handle per-OS dynamic-module compilation flags.
2642 PP/02 Let /dev/null have normal permissions.
2643 The 4.73 fixes were a little too stringent and complained about the
2644 permissions on /dev/null. Exempt it from some checks.
2645 Reported by Andreas M. Kirchwitz.
2647 PP/03 Report version information for many libraries, including
2648 Exim version information for dynamically loaded libraries. Created
2649 version.h, now support a version extension string for distributors
2650 who patch heavily. Dynamic module ABI change.
2652 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2653 privilege escalation vulnerability whereby the Exim run-time user
2654 can cause root to append content of the attacker's choosing to
2657 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2660 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2661 If dropping privileges for untrusted macros, we disabled normal logging
2662 on the basis that it would fail; for the Exim run-time user, this is not
2663 the case, and it resulted in successful deliveries going unlogged.
2664 Fixed. Reported by Andreas Metzler.
2670 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2671 only prepend for the Resent-* case. Fixes regression introduced in
2672 Exim 4.70 by NM/22 for Bugzilla 607.
2674 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2675 increasing numbers of administrators be bitten by this.
2677 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2679 PP/03 Bugzilla 994: added openssl_options main configuration option.
2681 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2683 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2685 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2687 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2688 without a peer certificate, leading to a segfault because of an
2689 assumption that peers always have certificates. Be a little more
2690 paranoid. Problem reported by Martin Tscholak.
2692 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2693 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2694 NB: ClamAV planning to remove STREAM in "middle of 2010".
2695 CL also introduces -bmalware, various -d+acl logging additions and
2696 more caution in buffer sizes.
2698 PP/09 Implemented reverse_ip expansion operator.
2700 PP/10 Bugzilla 937: provide a "debug" ACL control.
2702 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2704 PP/12 Bugzilla 973: Implement --version.
2706 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2708 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2710 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2712 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2713 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2714 ignore trailing whitespace.
2716 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2718 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2721 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2722 Notification from Dr Andrew Aitchison.
2724 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2725 ExtendedDetectionInfo response format.
2726 Notification from John Horne.
2728 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2731 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2732 XSL and documented dependency on system catalogs, with examples of how
2735 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2738 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2739 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2740 configuration file which is writeable by the Exim user or group.
2742 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2743 of configuration files to cover files specified with the -C option if
2744 they are going to be used with root privileges, not just the default
2747 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2748 option (effectively making it always true).
2750 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2751 files to be used while preserving root privileges.
2753 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2754 that rogue child processes cannot use them.
2756 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2757 run-time user, instead of root.
2759 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2760 Exim run-time user without dropping privileges.
2762 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2763 result string, instead of calling string_vformat() twice with the same
2766 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2767 for other users. Others should always drop root privileges if they use
2768 -C on the command line, even for a whitelisted configure file.
2770 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2772 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2778 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2779 $data_path, and $header_path variables; fixed documentation bugs and
2782 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2783 exipick to access non-standard spools, including the "frozen" queue
2786 NM/01 Bugzilla 965: Support mysql stored procedures.
2787 Patch from Alain Williams
2789 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2791 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2792 Patch from Andreas Metzler
2794 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2795 Patch from Kirill Miazine
2797 NM/05 Bugzilla 671: Added umask to procmail example.
2799 JJ/03 installed exipick 20100323.0, fixing doc bug
2801 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2802 directory. Notification and patch from Dan Rosenberg.
2804 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2806 TK/02 Improve log output when DKIM signing operation fails.
2808 MH/01 Treat the transport option dkim_domain as a colon separated
2809 list, not as a single string, and sign the message with each element,
2810 omitting multiple occurences of the same signer.
2812 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2813 Bugzilla 985, 986. Patch by Simon Arlott
2815 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2816 Patch by Simon Arlott
2818 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2819 MBX locking. Notification from Dan Rosenberg.
2825 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2827 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2829 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2831 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2833 NM/04 Bugzilla 915: Flag broken perl installation during build.
2839 TK/01 Added patch by Johannes Berg that expands the main option
2840 "spamd_address" if it starts with a dollar sign.
2842 TK/02 Write list of recipients to X-Envelope-Sender header when building
2843 the mbox-format spool file for content scanning (suggested by Jakob
2846 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2847 (http://www.dcc-servers.net/) support via dccifd. Activated by
2848 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2850 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2851 by Mark Daniel Reidel <mr@df.eu>.
2853 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2854 When building exim an external PCRE library is now needed -
2855 PCRE is a system library on the majority of modern systems.
2856 See entry on PCRE_LIBS in EDITME file.
2858 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2859 conversation. Added nologin parameter to request.
2860 Patch contributed by Kirill Miazine.
2862 TF/01 Do not log submission mode rewrites if they do not change the address.
2864 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2866 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2867 log files in place. Contributed by Roberto Lima.
2869 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2871 TF/03 Bugzilla 615: When checking the local_parts router precondition
2872 after a local_part_suffix or local_part_prefix option, Exim now
2873 does not use the address's named list lookup cache, since this
2874 contains cached lookups for the whole local part.
2876 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2877 Robert Millan. Documentation is in experimental-spec.txt.
2879 TF/04 Bugzilla 668: Fix parallel build (make -j).
2881 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2883 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2884 Patch provided by Jan Srzednicki.
2886 TF/05 Leading white space used to be stripped from $spam_report which
2887 wrecked the formatting. Now it is preserved.
2889 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2890 that they are available at delivery time.
2892 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2894 TF/08 TLS error reporting now respects the incoming_interface and
2895 incoming_port log selectors.
2897 TF/09 Produce a more useful error message if an SMTP transport's hosts
2898 setting expands to an empty string.
2900 NM/06 Bugzilla 744: EXPN did not work under TLS.
2901 Patch provided by Phil Pennock.
2903 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2904 Patch provided by Richard Godbee.
2906 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2907 acl_smtp_notquit, added index entry.
2909 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2910 Patch provided by Eugene Bujak.
2912 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2913 accept(). Patch provided by Maxim Dounin.
2915 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2916 Patch provided by Phil Pennock.
2918 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2920 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2921 Patch provided by Brad "anomie" Jorsch.
2923 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2924 Patch provided by Dean Brooks.
2926 TK/05 Add native DKIM support (does not depend on external libraries).
2928 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2929 Patch provided by Graeme Fowler.
2931 NM/16 Bugzilla 851: Documentation example syntax fix.
2933 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2935 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2938 NM/19 Bugzilla 745: TLS version reporting.
2939 Patch provided by Phil Pennock.
2941 NM/20 Bugzilla 167: bool: condition support.
2942 Patch provided by Phil Pennock.
2944 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2945 clients. Patch provided by Phil Pennock.
2947 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2948 Patch provided by Brad "anomie" Jorsch.
2950 NM/23 Bugzilla 687: Fix misparses in eximstats.
2951 Patch provided by Heiko Schlittermann.
2953 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2954 Patch provided by Heiko Schlittermann.
2956 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2957 plus update to original patch.
2959 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2961 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2962 Patch provided by David Brownlee.
2964 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2966 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2968 NM/30 Bugzilla 888: TLS documentation bugfixes.
2970 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2972 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2973 Unlike the original bugzilla I have changed all shell scripts in src tree.
2975 NM/33 Bugzilla 898: Transport filter timeout fix.
2976 Patch by Todd Rinaldo.
2978 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2979 Patch by Serge Demonchaux.
2981 NM/35 Bugzilla 39: Base64 decode bug fixes.
2982 Patch by Jakob Hirsch.
2984 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2986 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2988 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2990 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2996 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2997 ALT-N's libdkim that I have put here:
2998 http://duncanthrax.net/exim-experimental/
3000 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3001 were called 'true' and it seems that C99 defines that as a reserved
3002 keyword to be used with 'bool' variable types. That means you could
3003 not include C99-style headers which use bools without triggering
3004 build errors in sieve.c.
3006 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3007 as mailq or other aliases. Changed the --help handling significantly
3008 to do whats expected. exim_usage() emits usage/help information.
3010 SC/01 Added the -bylocaldomain option to eximstats.
3012 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3014 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3016 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3022 PH/01 Another patch from the Sieve maintainer.
3024 PH/02 When an IPv6 address is converted to a string for single-key lookup
3025 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3026 dots are used instead of colons so that keys in lsearch files need not
3027 contain colons. This was done some time before quoting was made available
3028 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3029 (notated using the quote facility) so as to distinguish them from IPv4
3030 keys. This meant that lookups for IP addresses in host lists did not work
3031 for iplsearch lookups.
3033 This has been fixed by arranging for IPv6 addresses to be expressed with
3034 colons if the lookup type is iplsearch. This is not incompatible, because
3035 previously such lookups could never work.
3037 The situation is now rather anomalous, since one *can* have colons in
3038 ordinary lsearch keys. However, making the change in all cases is
3039 incompatible and would probably break a number of configurations.
3041 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3044 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3045 conversion specification without a maximum field width, thereby enabling
3046 a rogue spamd server to cause a buffer overflow. While nobody in their
3047 right mind would setup Exim to query an untrusted spamd server, an
3048 attacker that gains access to a server running spamd could potentially
3049 exploit this vulnerability to run arbitrary code as the Exim user.
3051 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3052 $primary_hostname instead of what libspf2 thinks the hosts name is.
3054 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3055 a directory entry by the name of the lookup key. Previously, if a
3056 symlink pointed to a non-existing file or a file in a directory that
3057 Exim lacked permissions to read, a lookup for a key matching that
3058 symlink would fail. Now it is enough that a matching directory entry
3059 exists, symlink or not. (Bugzilla 503.)
3061 PH/03 The body_linecount and body_zerocount variables are now exported in the
3064 PH/04 Added the $dnslist_matched variable.
3066 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3067 This means they are set thereafter only if the connection becomes
3070 PH/06 Added the client_condition to authenticators so that some can be skipped
3071 by clients under certain conditions.
3073 PH/07 The error message for a badly-placed control=no_multiline_responses left
3074 "_responses" off the end of the name.
3076 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3078 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3079 (without spaces) instead of just copying the configuration text.
3081 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3083 PH/11 Added $max_received_linelength.
3085 PH/12 Added +ignore_defer and +include_defer to host lists.
3087 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3088 way in which PCRE > 7.0 is built.
3090 PH/14 Implemented queue_only_load_latch.
3092 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3093 MAIL command. The effect was to mangle the value on 64-bit systems.
3095 PH/16 Another patch from the Sieve maintainer.
3097 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3099 PH/18 If a system quota error occurred while trying to create the file for
3100 a maildir delivery, the message "Mailbox is full" was not appended to the
3101 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3102 only to a quota excession during the actual writing of the file.
3104 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3105 characters?) which causes problems in log lines. The DN values are now
3106 passed through string_printing() before being added to log lines.
3108 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3109 and InterBase are left for another time.)
3111 PH/21 Added message_body_newlines option.
3113 PH/22 Guard against possible overflow in moan_check_errorcopy().
3115 PH/23 POSIX allows open() to be a macro; guard against that.
3117 PH/24 If the recipient of an error message contained an @ in the local part
3118 (suitably quoted, of course), incorrect values were put in $domain and
3119 $local_part during the evaluation of errors_copy.
3125 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3126 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3129 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3130 issue a MAIL command.
3132 PH/02 In an ACL statement such as
3134 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3136 if a client was not listed at all, or was listed with a value other than
3137 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3138 the condition was not true (as it should be), so access was not denied.
3139 The bug was that the ! inversion was incorrectly passed on to the second
3140 item. This has been fixed.
3142 PH/03 Added additional dnslists conditions == and =& which are different from
3143 = and & when the dns lookup returns more than one IP address.
3145 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3146 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3148 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3149 FSYNC, which compiles an option called disable_fsync that allows for
3150 bypassing fsync(). The documentation is heavily laced with warnings.
3152 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3154 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3155 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3156 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3157 including adding "make clean"; (3) Added -fPIC when compiling the test
3158 dynamically loaded module, to get rid of a warning.
3160 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3161 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3162 = 0s. The bug is otherwise harmless.
3164 PH/07 There was a bug in the dovecot authenticator such that the value of
3165 $auth1 could be overwritten, and so not correctly preserved, after a
3166 successful authentication. This usually meant that the value preserved by
3167 the server_setid option was incorrect.
3169 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3171 PH/09 Installed PCRE release 7.0.
3173 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3174 run for batched SMTP input. It is now run at the start of every message
3175 in the batch. While fixing this I discovered that the process information
3176 (output by running exiwhat) was not always getting set for -bs and -bS
3177 input. This is fixed, and it now also says "batched" for BSMTP.
3179 PH/11 Added control=no_pipelining.
3181 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3182 patch, slightly modified), and move the expansion of helo_data till after
3183 the connection is made in the smtp transport (so it can use these
3186 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3188 PH/14 Added log_selector = +pid.
3190 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3192 PH/16 Add ${if forany and ${if forall.
3194 PH/17 Added dsn_from option to vary the From: line in DSNs.
3196 PH/18 Flush SMTP output before performing a callout, unless control =
3197 no_callout_flush is set.
3199 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3200 was true (the default) a successful delivery failed to delete the retry
3201 item, thus causing premature timeout of the address. The bug is now
3204 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3206 PH/21 Long custom messages for fakedefer and fakereject are now split up
3207 into multiline responses in the same way that messages for "deny" and
3208 other ACL rejections are.
3210 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3211 with slight modification.
3213 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3214 draft, changing the syntax and factoring some duplicate code".
3216 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3217 for deliveries of the second and subsequent messages over the same SMTP
3220 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3221 ${reduce, with only minor "tidies".
3223 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3225 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3226 expansion side effects.
3228 PH/27 When a message times out after an over-quota error from an Exim-imposed
3229 quota, the bounce message says "mailbox is full". This message was not
3230 being given when it was a system quota that was exceeded. It now should
3233 MH/03 Made $recipients available in local_scan(). local_scan() already has
3234 better access to the recipient list through recipients_list[], but
3235 $recipients can be useful in postmaster-provided expansion strings.
3237 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3238 in the case of a MAIL command with additional options following the
3239 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3240 were accidentally chopped off.
3242 PH/29 SMTP synchronization checks are implemented when a command is read -
3243 there is a check that no more input is waiting when there shouldn't be
3244 any. However, for some commands, a delay in an ACL can mean that it is
3245 some time before the response is written. In this time, more input might
3246 arrive, invalidly. So now there are extra checks after an ACL has run for
3247 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3248 pipelining has not been advertised.
3250 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3252 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3253 returned in the SMTP response when smtp_return_error_details was false.
3254 This has been fixed.
3256 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3257 instead of the C I/O that was originally supplied, because problems were
3258 reported on Solaris.
3260 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3261 Exim which did not show up earlier: it was assuming that a call to
3262 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3263 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3264 was a macro that became an assignment, so it seemed to work. This has
3265 changed to a proper function call with a void return, hence the compile
3266 error. Exim's code has been fixed.
3268 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3271 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3273 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3275 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3276 because it turns out to be too big or there is a timeout) there is no
3277 "Completed" line in the log. When some messages of this type were
3278 selected by exigrep, they were listed as "not completed". Others were
3279 picked up by some special patterns. I have improved the selection
3280 criteria to be more general.
3282 PH/38 The host_find_failed option in the manualroute router can now be set
3283 to "ignore", to completely ignore a host whose IP address cannot be
3284 found. If all hosts are ignored, the behaviour is controlled by the new
3285 host_all_ignored option.
3287 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3288 homing or because of multiple MX records with /mx) generated more than
3289 one IP address, and the following item turned out to be the local host,
3290 all the secondary addresses of the first item were incorrectly removed
3291 from the list, along with the local host and any following hosts (which
3292 is what is supposed to happen).
3294 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3295 whoever called Exim into the -H file. In the case of the daemon it was
3296 behaving confusingly. When first started, it used values for whoever
3297 started the daemon, but after a SIGHUP it used the Exim user (because it
3298 calls itself on a restart). I have changed the code so that it now always
3301 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3302 message are rejected with the same error (e.g. no authentication or bad
3303 sender address), and a DATA command is nevertheless sent (as can happen
3304 with PIPELINING or a stupid MUA), the error message that was given to the
3305 RCPT commands is included in the rejection of the DATA command. This is
3306 intended to be helpful for MUAs that show only the final error to their
3309 PH/42 Another patch from the Sieve maintainer.
3311 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3312 Eximstats - Fixed some broken HTML links and added missing column headers
3314 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3315 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3317 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3319 PH/43 Yet another patch from the Sieve maintainer.
3321 PH/44 I found a way to check for a TCP/IP connection going away before sending
3322 the response to the final '.' that terminates a message, but only in the
3323 case where the client has not sent further data following the '.'
3324 (unfortunately, this is allowed). However, in many cases there won't be
3325 any further data because there won't be any more messages to send. A call
3326 to select() can be used: if it shows that the input is "ready", there is
3327 either input waiting, or the socket has been closed. An attempt to read
3328 the next input character can distinguish the two cases. Previously, Exim
3329 would have sent an OK response which the client would never have see.
3330 This could lead to message repetition. This fix should cure that, at
3331 least in a lot of common cases.
3333 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3334 advertised in response to EHLO.
3340 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3341 fixed by 4.65/MH/01 (is this a record?) are fixed:
3343 (i) An empty string was always treated as zero by the numeric comparison
3344 operators. This behaviour has been restored.
3346 (ii) It is documented that the numeric comparison operators always treat
3347 their arguments as decimal numbers. This was broken in that numbers
3348 starting with 0 were being interpreted as octal.
3350 While fixing these problems I realized that there was another issue that
3351 hadn't been noticed. Values of message_size_limit (both the global option
3352 and the transport option) were treated as octal if they started with 0.
3353 The documentation was vague. These values are now always treated as
3354 decimal, and I will make that clear in the documentation.
3360 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3361 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3364 MH/01 Don't check that the operands of numeric comparison operators are
3365 integers when their expansion is in "skipping" mode (fixes bug
3366 introduced by 4.64-PH/07).
3368 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3369 child addresses, Exim now panics and dies. Previously, because the count
3370 is held in a short int, deliveries were likely to be lost. As such a
3371 large number of recipients for a single message is ridiculous
3372 (performance will be very, very poor), I have chosen to impose a limit
3373 rather than extend the field.
3379 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3380 leftover -K file (the existence of which was triggered by #402).
3381 While we were at it, introduced process PID as part of the -K
3382 filename. This should rule out race conditions when creating
3385 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3386 processing considerably. Previous code took too long for large mails,
3387 triggering a timeout which in turn triggers #401.
3389 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3390 in the DK code in transports.c. sendfile() is not really portable,
3391 hence the _LINUX specificness.
3393 TF/01 In the add_headers option to the mail command in an Exim filter,
3394 there was a bug that Exim would claim a syntax error in any
3395 header after the first one which had an odd number of characters
3398 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3399 callout verification, Exim cached a "reject" for the entire domain. This
3400 is correct for most verifications, but it is not correct for a recipient
3401 verification with use_sender or use_postmaster set, because in that case
3402 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3403 case of MAIL FROM:<> rejection from other early rejections (e.g.
3404 rejection of HELO). When verifying a recipient using a non-null MAIL
3405 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3406 Whatever the result of the callout, the value of the domain cache is
3407 left unchanged (for any other kind of callout, getting as far as trying
3408 RCPT means that the domain itself is ok).
3410 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3413 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3414 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3415 session, the daemon ignores ECONNECT errors and logs others; it now
3416 ignores EPIPE as well.
3418 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3419 (quoted-printable decoding).
3421 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3422 later the small subsequent patch to fix an introduced bug.
3424 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3426 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3428 PH/08 An error is now given if message_size_limit is specified negative.
3430 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3431 to be given (somewhat) arbitrary names.
3433 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3436 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3437 miscellaneous code fixes
3439 PH/10 Added the log_reject_target ACL modifier to specify where to log
3442 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3443 hostname. This is wrong, because it relates to the incoming message (and
3444 probably the interface on which it is arriving) and not to the outgoing
3445 callout (which could be using a different interface). This has been
3446 changed to use the value of the helo_data option from the smtp transport
3447 instead - this is what is used when a message is actually being sent. If
3448 there is no remote transport (possible with a router that sets up host
3449 addresses), $smtp_active_hostname is used.
3451 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3452 tweaks were necessary in order to get it to work (see also 21 below):
3453 (a) The code assumed that strncpy() returns a negative number on buffer
3454 overflow, which isn't the case. Replaced with Exim's string_format()
3456 (b) There were several signed/unsigned issues. I just did the minimum
3457 hacking in of casts. There is scope for a larger refactoring.
3458 (c) The code used strcasecmp() which is not a standard C function.
3459 Replaced with Exim's strcmpic() function.
3460 (d) The code set only $1; it now sets $auth1 as well.
3461 (e) A simple test gave the error "authentication client didn't specify
3462 service in request". It would seem that Dovecot has changed its
3463 interface. Fortunately there's a specification; I followed it and
3464 changed what the client sends and it appears to be working now.
3466 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3469 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3470 address A is aliased to B and C, where B exists and C does not. Without
3471 -v the output is "A verified" because verification stops after a
3472 successful redirection if more than one address is generated. However,
3473 with -v the child addresses are also verified. Exim was outputting "A
3474 failed to verify" and then showing the successful verification for C,
3475 with its parentage. It now outputs "B failed to verify", showing B's
3476 parentage before showing the successful verification of C.
3478 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3479 look up a TXT record in a specific list after matching in a combined
3482 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3483 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3484 they consult the DNS. I had assumed they would set it the way they
3485 wanted; and indeed my experiments on Linux seem to show that in some
3486 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3487 To be on the safe side, however, I have now made the interface to
3488 host_find_byname() similar to host_find_bydns(), with an argument
3489 containing the DNS resolver options. The host_find_byname() function now
3490 sets these options at its start, just as host_find_bydns() does. The smtp
3491 transport options dns_qualify_single and dns_search_parents are passed to
3492 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3493 of host_find_byname() use the default settings of RES_DEFNAMES
3494 (qualify_single) but not RES_DNSRCH (search_parents).
3496 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3497 spool_read_header() do less string testing, by means of a preliminary
3498 switch on the second character of optional "-foo" lines. (This is
3499 overdue, caused by the large number of possibilities that now exist.
3500 Originally there were few.) While I was there, I also converted the
3501 str(n)cmp tests so they don't re-test the leading "-" and the first
3502 character, in the hope this might squeeze out yet more improvement.
3504 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3505 flag allowing group syntax was set by the header_syntax check but not
3506 turned off, possible causing trouble later; (2) The flag was not being
3507 set at all for the header_verify test, causing "group"-style headers to
3508 be rejected. I have now set it in this case, and also caused header_
3509 verify to ignore an empty address taken from a group. While doing this, I
3510 came across some other cases where the code for allowing group syntax
3511 while scanning a header line wasn't quite right (mostly, not resetting
3512 the flag correctly in the right place). These bugs could have caused
3513 trouble for malformed header lines. I hope it is now all correct.
3515 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3516 with the "reply" argument non-NULL. The code, however (which originally
3517 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3518 but it didn't always do it. This confused somebody who was copying the
3519 code for some other use. I have removed all the tests.
3521 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3522 feature that was used to support insecure browsers during the U.S. crypto
3523 embargo. It requires special client support, and Exim is probably the
3524 only MTA that supported it -- and would never use it because real RSA is
3525 always available. This code has been removed, because it had the bad
3526 effect of slowing Exim down by computing (never used) parameters for the
3527 RSA_EXPORT functionality.
3529 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3530 authenticator to fail if there's a tab character in the incoming data
3531 (there should never be unless someone is messing about, as it's supposed
3532 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3533 if the connection is using TLS or if the remote IP is the same as the
3534 local IP, and the "valid-client-cert option" if a client certificate has
3537 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3538 authenticators. This can be used for authorization after authentication
3539 succeeds. (In the case of plaintext, it servers for both authentication
3542 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3543 if any retry times were supplied.
3545 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3546 connection, where there is no client IP address to check. In this
3547 situation, the verify now always succeeds.
3549 PH/25 Applied John Jetmore's -Mset patch.
3551 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3553 PH/27 In a string expansion for a processed (not raw) header when multiple
3554 headers of the same name were present, leading whitespace was being
3555 removed from all of them, but trailing whitespace was being removed only
3556 from the last one. Now trailing whitespace is removed from each header
3557 before concatenation. Completely empty headers in a concatenation (as
3558 before) are ignored.
3560 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3561 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3563 PH/29 [Removed. This was a change that I later backed out, and forgot to
3564 correct the ChangeLog entry (that I had efficiently created) before
3565 committing the later change.]
3567 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3568 address errors (4xx response to RCPT) over the same connection as other
3569 messages routed to the same hosts. Such deliveries are always "forced",
3570 so retry times are not inspected. This resulted in far too many retries
3571 for the affected addresses. The effect occurred only when there were more
3572 hosts than the hosts_max_try setting in the smtp transport when it had
3573 the 4xx errors. Those hosts that it had tried were not added to the list
3574 of hosts for which the message was waiting, so if all were tried, there
3575 was no problem. Two fixes have been applied:
3577 (i) If there are any address or message errors in an SMTP delivery, none
3578 of the hosts (tried or untried) are now added to the list of hosts
3579 for which the message is waiting, so the message should not be a
3580 candidate for sending over the same connection that was used for a
3581 successful delivery of some other message. This seems entirely
3582 reasonable: after all the message is NOT "waiting for some host".
3583 This is so "obvious" that I'm not sure why it wasn't done
3584 previously. Hope I haven't missed anything, but it can't do any
3585 harm, as the worst effect is to miss an optimization.
3587 (ii) If, despite (i), such a delivery is accidentally attempted, the
3588 routing retry time is respected, so at least it doesn't keep
3589 hammering the server.
3591 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3592 in ${readsocket because some servers need this prod.
3594 PH/32 Added some extra debug output when updating a wait-xxx database.
3596 PH/33 The hint "could be header name not terminated by colon", which has been
3597 given for certain expansion errors for a long time, was not being given
3598 for the ${if def:h_colon_omitted{... case.
3600 PH/34 The spec says: "With one important exception, whenever a domain list is
3601 being scanned, $domain contains the subject domain." There was at least
3602 one case where this was not true.
3604 PH/35 The error "getsockname() failed: connection reset by peer" was being
3605 written to the panic log as well as the main log, but it isn't really
3606 panic-worthy as it just means the connection died rather early on. I have
3607 removed the panic log writing for the ECONNRESET error when getsockname()
3610 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3611 runs only) independently of the message's sender address. This meant
3612 that, if the 4xx error was in fact related to the sender, a different
3613 message to the same recipient with a different sender could confuse
3614 things. In particular, this can happen when sending to a greylisting
3615 server, but other circumstances could also provoke similar problems.
3616 I have changed the default so that the retry time for these errors is now
3617 based a combination of the sender and recipient addresses. This change
3618 can be overridden by setting address_retry_include_sender=false in the
3621 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3622 remote server are returned as part of bounce messages. This was not
3623 happening for LMTP over a pipe (the lmtp transport), but now it is the
3624 same for both kinds of LMTP.
3626 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3627 in header lines that were in fact CNAMEs. This is no longer the case.
3629 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3630 and queue runs started by the daemon processed all messages. This has
3631 been fixed so that -R and -S can now usefully be given with -q<time>.
3633 PH/40 Import PCRE release 6.7 (fixes some bugs).
3635 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3637 PH/42 Give an error if -q is specified more than once.
3639 PH/43 Renamed the variables $interface_address and $interface_port as
3640 $received_ip_address and $received_port, to make it clear that these
3641 values apply to message reception, and not to the outgoing interface when
3642 a message is delivered. (The old names remain recognized, of course.)
3644 PH/44 There was no timeout on the connect() call when using a Unix domain
3645 socket in the ${readsocket expansion. There now is.
3647 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3648 be meaningful with "accept".
3650 SC/01 Eximstats V1.43
3651 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3653 SC/02 Eximstats V1.44
3654 Use a glob alias rather than an array ref in the generated
3655 parser. This improves both readability and performance.
3657 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3658 Collect SpamAssassin and rejection statistics.
3659 Don't display local sender or destination tables unless
3660 there is data to show.
3661 Added average volumes into the top table text output.
3663 SC/04 Eximstats V1.46
3664 Collect data on the number of addresses (recipients)
3665 as well as the number of messages.
3667 SC/05 Eximstats V1.47
3668 Added 'Message too big' to the list of mail rejection
3669 reasons (thanks to Marco Gaiarin).
3671 SC/06 Eximstats V1.48
3672 Mainlog lines which have GMT offsets and are too short to
3673 have a flag are now skipped.
3675 SC/07 Eximstats V1.49 (Alain Williams)
3676 Added the -emptyok flag.
3678 SC/08 Eximstats V1.50
3679 Fixes for obtaining the IP address from reject messages.
3681 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3682 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3683 whitespace changes from 4.64-PH/27
3685 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3688 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3689 are found, allow negative numbers in numeric criteria)
3691 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3693 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3694 to match changes made in 4.64-PH/43
3696 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3698 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3699 "domain" is actually an IP address, and force a failure. This locks out
3700 those revolvers/nameservers that support "A-for-A" lookups, in
3701 contravention of the specifications.
3703 PH/48 When a host name was looked up from an IP address, and the subsequent
3704 forward lookup of the name timed out, the host name was left in
3705 $sender_host_name, contrary to the specification.
3707 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3708 restricted to single-key lookups, Exim was not diagnosing an error if
3709 * or *@ was used with a query-style lookup.
3711 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3713 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3714 long ago, but noone interested enough thought of it. Let's just say that
3715 the "1.1" means that there are some new functions that weren't there at
3716 some point in the past.
3718 PH/51 Error processing for expansion failure of helo_data from an smtp
3719 transport during callout processing was broken.
3721 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3722 tested/used via the -bh/-bhc/-bs options.
3724 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3725 bug, fixed in subsequent PCRE releases).
3727 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3728 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3730 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3736 SC/01 Use a glob alias rather than an array ref in eximstats generated
3737 parser. This improves both readability and performance.
3739 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3740 Don't display local sender or destination tables in eximstats unless
3741 there is data to show.
3742 Added average volumes into the eximstats top table text output.
3744 SC/03 Collect data on the number of addresses (recipients) as well
3745 as the number of messages in eximstats.
3747 TF/01 Correct an error in the documentation for the redirect router. Exim
3748 does (usually) call initgroups() when daemonizing.
3750 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3751 with consistent privilege compared to when running as a daemon.
3753 TF/03 Note in the spec that $authenticated_id is not set for local
3754 submissions from trusted users.
3756 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3757 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3759 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3760 by adding some example configuration directives to the default
3761 configuration file. A little bit of work is required to uncomment the
3762 directives and define how usernames and passwords are checked, but
3763 there is now a framework to start from.
3765 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3766 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3767 without this. I don't know how relevant this is to other LDAP libraries.
3769 PH/02 Add the verb name to the "unknown ACL verb" error.
3771 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3773 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3775 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3776 directory not expanded when it should be if an expanded home directory
3777 was set for the address (which is overridden by the transport).
3779 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3782 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3783 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3784 because it is too late at that time, and has no effect.
3786 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3787 security issue with \' (bugzilla #107). I could not use the
3788 PQescapeStringConn() function, because it needs a PGconn value as one of
3791 PH/08 When testing addresses using -bt, indicate those final addresses that
3792 are duplicates that would not cause an additional delivery. At least one
3793 person was confused, thinking that -bt output corresponded to deliveries.
3794 (Suppressing duplicates isn't a good idea as you lose the information
3795 about possibly different redirections that led to the duplicates.)
3797 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3798 systems where poll() doesn't work, in particular OS X.
3800 PH/10 Added more information to debugging output for retry time not reached.
3802 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3803 operations in malware.c.
3805 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3808 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3809 syslog_duplication set false, log lines that would normally be written
3810 both the the main log and to the reject log were not written to syslog at
3813 PH/14 In the default configuration, change the use of "message" in ACL warn
3814 statements to "add_header".
3816 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3817 not followed by a command (e.g. "seen endif").
3819 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3820 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3823 PH/17 Added extra conditions to the default value of delay_warning_condition
3827 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3828 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3829 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3832 The Auto-Submitted: and various List- headers are standardised, whereas I
3833 don't think Precedence: ever was.
3835 PH/18 Refactored debugging code in route_finduser() to show more information,
3836 in particular, the error code if getpwnam() issues one.
3838 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3839 This is apparently needed in addition to the PH/07 change above to avoid
3840 any possible encoding problems.
3842 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3843 but not after initializing Perl.
3845 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3846 output them only if debugging. By default they are written stderr,
3847 apparently, which is not desirable.
3849 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3852 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3855 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3857 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3858 authenticated or an ident call has been made. Suppress the default
3859 values for $authenticated_id and $authenticated_sender (but permit -oMai
3860 and -oMas) when testing with -bh.
3862 PH/24 Re-jigged the order of the tests in the default configuration so that the
3863 tests for valid domains and recipients precede the DNS black list and CSA
3864 tests, on the grounds that those ones are more expensive.
3866 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3867 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3868 This bug exists in every version of Exim that I still have, right back to
3871 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3872 However, an attempt to turn on case-sensitivity in a regex key by
3873 including (?-i) didn't work because the subject string was already
3874 lowercased, and the effects were non-intuitive. It turns out that a
3875 one-line patch can be used to allow (?-i) to work as expected.
3881 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3882 other effects) broke the use of negated acl sub-conditions.
3884 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3887 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3888 "Deny" causes Exim to reject the incoming connection with a 554 error.
3889 Unfortunately, if there is a major crisis, such as a disk failure,
3890 tcp-wrappers gives "deny", whereas what one would like would be some
3891 kind of temporary error. A kludge has been added to help with this.
3892 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3893 554 error is used if errno is still zero or contains ENOENT (which occurs
3894 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3897 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3899 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3900 errors. Otherwise a message that provokes a temporary error (when other
3901 messages do not) can cause a whole host to time out.
3903 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3904 addresses were routed directly to files or pipes from a redirect router.
3905 File deliveries just didn't batch; pipe deliveries might have suffered
3908 PH/06 A failure to get a lock for a hints database would erroneously always say
3909 "Failed to get write lock", even when it was really a read lock.
3911 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3912 of 0600. This has been changed to use the value of the lockfile_mode
3913 option (which defaults to 0600).
3915 PH/08 Applied small patch from the Sieve maintainer.
3917 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3918 folder from quota calculations, a direct delivery into this folder messed
3919 up the contents of the maildirsize file. This was because the regex was
3920 used only to exclude .Trash (or whatever) when the size of the mailbox
3921 was calculated. There was no check that a delivery was happening into an
3922 excluded directory. This bug has been fixed by ignoring all quota
3923 processing for deliveries into excluded directories.
3925 PH/10 Added the maildirfolder_create_regex option to appendfile.
3931 PH/01 The code for finding all the local interface addresses on a FreeBSD
3932 system running IPv6 was broken. This may well have applied to all BSD
3933 systems, as well as to others that have similar system calls. The broken
3934 code found IPv4 interfaces correctly, but gave incorrect values for the
3935 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3936 that it would not match correctly against @[] and not recognize the IPv6
3939 PH/02 The ipliteral router was not recognizing addresses of the form user@
3940 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3942 PH/03 Added disable_ipv6.
3944 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3945 decoded form, because it is most often used to construct To: headers
3946 lines in autoreplies, and the decoded form may well be syntactically
3947 invalid. However, $reply_address has leading white space removed, and all
3948 newlines turned into spaces so that the autoreply transport does not
3951 PH/05 If group was specified without a user on a router, and no group or user
3952 was specified on a transport, the group from the router was ignored.
3954 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3955 for visible compile-time settings that can be used to change these
3956 numbers, for those that want even more. Backwards compatibility with old
3957 spool files has been maintained. However, going back to a previous Exim
3958 release will lost any variables that are in spool files.
3960 PH/07 Two small changes when running in the test harness: increase delay when
3961 passing a TCP/IP connection to a new process, in case the original
3962 process has to generate a bounce, and remove special handling of
3963 127.0.0.2 (sic), which is no longer necessary.
3965 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3966 be the same on different OS.
3968 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3971 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3972 whether --show-vars was specified or not
3974 JJ/02 exipick: Added support for new ACL variable spool format introduced
3977 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3978 syntactically invalid From: or Reply-to: line, and a filter used this to
3979 generate an autoreply, and therefore failed to obtain an address for the
3980 autoreply, Exim could try to deliver to a non-existent relative file
3981 name, causing unrelated and misleading errors. What now happens is that
3982 it logs this as a hard delivery error, but does not attempt to create a
3985 PH/11 The exinext utility has a -C option for testing purposes, but although
3986 the given file was scanned by exinext itself; it wasn't being passed on
3987 when Exim was called.
3989 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3990 an end-of-file indication when reading a command response.
3992 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3993 compiled. In many other places in Exim, IPv6 addresses are always
3994 recognized, so I have changed this. It also means that IPv4 domain
3995 literals of the form [IPV4:n.n.n.n] are now always recognized.
3997 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3998 used if the router is not running as root, for example, when verifying at
3999 ACL time, or when using -bh. The debugging output from this situation was
4000 non-existent - all you got was a failure to exec. I have made two
4003 (a) Failures to set uid/gid, the current directory, or a process leader
4004 in a subprocess such as that created by queryprogram now generate
4005 suitable debugging output when -d is set.
4007 (b) The queryprogram router detects when it is not running as root,
4008 outputs suitable debugging information if -d is set, and then runs
4009 the subprocess without attempting to change uid/gid.
4011 PH/15 Minor change to Makefile for building test_host (undocumented testing
4014 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4015 additional section of a DNS packet that returns MX or SRV records.
4016 Instead, it always explicitly searches for A/AAAA records. This avoids
4017 major problems that occur when a DNS server includes only records of one
4018 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4019 fixed another bug: if SRV records were looked up and the corresponding
4020 address records were *not* found in the additional section, the port
4021 values from the SRV records were lost.
4023 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4024 using the correct key (the original address) when searching the retry
4025 rules in order to find which one to use for generating the retry hint.
4027 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4028 adding the default one. Similarly, if it contains a Reply-To: header, the
4029 errors_reply_to option, if set, is not used.
4031 PH/19 When calculating a retry time, Exim used to measure the "time since
4032 failure" by looking at the "first failed" field in the retry record. Now
4033 it does not use this if it is later than than the arrival time of the
4034 message. Instead it uses the arrival time. This makes for better
4035 behaviour in cases where some deliveries succeed, thus re-setting the
4036 "first failed" field. An example is a quota failure for a huge message
4037 when small messages continue to be delivered. Without this change, the
4038 "time since failure" will always be short, possible causing more frequent
4039 delivery attempts for the huge message than are intended.
4040 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4042 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4043 $1, $2, $3) because the numerical variables can be reset during some
4044 expansion items (e.g. "match"), thereby losing the authentication data.
4046 PH/21 Make -bV show the size of off_t variables so that the test suite can
4047 decide whether to run tests for quotas > 2G.
4049 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4050 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4051 filecount value is greater than 2G or if a quota value is greater than 2G
4052 on a system where the size of off_t is not greater than 4, a panic error
4055 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4056 never match. The debug and -bh output now contains an explicit error
4057 message indicating a malformed IPv4 address or mask.
4059 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4060 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4061 PH/23 above applies.
4063 PH/25 Do not write to syslog when running in the test harness. The only
4064 occasion when this arises is a failure to open the main or panic logs
4065 (for which there is an explicit test).
4067 PH/26 Added the /no_tell option to "control=freeze".
4069 PH/27 If a host name lookup failed very early in a connection, for example, if
4070 the IP address matched host_lookup and the reverse lookup yielded a name
4071 that did not have a forward lookup, an error message of the form "no IP
4072 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4073 could be logged. Now it outputs the IP address instead of "NULL".
4075 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4076 allows the sender and the authenticated sender to be set when
4077 submitting a message from within Exim. Since child_open_exim() is
4078 documented for local_scan(), the new function should be too.
4080 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4081 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4082 results in an empty string is now treated as unset.
4084 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4086 PH/31 Added sender_verify_fail logging option.
4088 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4089 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4090 tidied the source and removed it altogether.
4092 PH/33 When a queue run was abandoned because the load average was too high, a
4093 log line was always written; now it is written only if the queue_run log
4094 selector is set. In addition, the log line for abandonment now contains
4095 information about the queue run such as the pid. This is always present
4096 in "start" and "stop" lines but was omitted from the "abandon" line.
4098 PH/34 Omit spaces between a header name and the colon in the error message that
4099 is given when verify = headers_syntax fails (if there are lots of them,
4100 the message gets confusing).
4102 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4103 names, as there are now some PTR records that contain slashes. This check
4104 is only to protect against broken name servers that fall over on strange
4105 characters, so the fact that it applies to all lookups doesn't matter.
4107 PH/36 Now that the new test suite is complete, we can remove some of the
4108 special code in Exim that was needed for the old test suite. For example,
4109 sorting DNS records because real resolvers return them in an arbitrary
4110 order. The new test suite's fake resolver always returns records in the
4113 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4114 bounces) except when queue_only is set, to avoid logging races between
4115 the different processes.
4117 PH/38 Panic-die if .include specifies a non-absolute path.
4119 PH/39 A tweak to the "H" retry rule from its user.
4121 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4122 a label. They prevented compilation on older perls.
4124 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4125 a warning to be raised on newish perls.
4127 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4128 on queue. Changes to match documented behaviour of showing count of
4129 messages matching specified criteria.
4131 PH/40 Changed the default ident timeout from 30s to 5s.
4133 PH/41 Added support for the use of login_cap features, on those BSD systems
4134 that have them, for controlling the resources used by pipe deliveries.
4136 PH/42 The content-scanning code uses fopen() to create files in which to put
4137 message data. Previously it was not paying any attention to the mode of
4138 the files. Exim runs with umask(0) because the rest of the code creates
4139 files with open(), and sets the required mode explicitly. Thus, these
4140 files were ending up world-writeable. This was not a big issue, because,
4141 being within the spool directory, they were not world-accessible. I have
4142 created a function called modefopen, which takes an additional mode
4143 argument. It sets umask(777), creates the file, chmods it to the required
4144 mode, then resets the umask. All the relevant calls to fopen() in the
4145 content scanning code have been changed to use this function.
4147 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4148 to 24 hours. This avoids potential overflow problems when processing G
4149 and H retry rules. I suspect nobody ever tinkers with this value.
4151 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4153 PH/45 When the plaintext authenticator is running as a client, the server's
4154 challenges are checked to ensure they are valid base64 strings. By
4155 default, the authentication attempt is cancelled if an invalid string is
4156 received. Setting client_ignore_invalid_base64 true ignores these errors.
4157 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4158 they are received. Thus, the responses can be made to depend on the
4159 challenges. If an invalid string is ignored, an empty string is placed in
4162 PH/46 Messages that are created by the autoreply transport now contains a
4163 References: header, in accordance with RFCs 2822 and 3834.
4165 PH/47 Added authenticated_sender_force to the smtp transport.
4167 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4169 PH/49 Installed latest patch from the Sieve maintainer.
4171 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4172 was also set, the appendfile transport was unnecessarily scanning a
4173 directory of message files (e.g. for maildir delivery) to find the count
4174 of files (along with the size), even though it did not need this
4175 information. It now does the scan only if it needs to find either the
4176 size of the count of files.
4178 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4180 PH/52 Two bugs concerned with error handling when the smtp transport is
4183 (i) Exim was not creating retry information for temporary errors given
4184 for individual recipients after the DATA command when the smtp transport
4185 was used in LMTP mode. This meant that they could be retried too
4186 frequently, and not timed out correctly.
4188 (ii) Exim was setting the flag that allows error details to be returned
4189 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4190 recipients that were returned after the DATA command.
4192 PH/53 This is related to PH/52, but is more general: for any failing address,
4193 when detailed error information was permitted to be returned to the
4194 sender, but the error was temporary, then after the final timeout, only
4195 "retry timeout exceeded" was returned. Now it returns the full error as
4196 well as "retry timeout exceeded".
4198 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4199 do this, and (what is worse) MTAs that accept it.
4201 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4202 will now be deprecated.
4204 PH/56 New os.c-cygwin from the Cygwin maintainer.
4206 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4207 formats (previously only available in exim formats via -bpr, -bpru,
4208 and -bpra. Now also available in native and exiqgrep formats)
4210 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4211 with very large, slow to parse queues
4213 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4215 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4217 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4218 responses to authentication challenges, though it was showing the
4219 challenges; (ii) I've removed the CR characters from the debug output for
4222 PH/58 Allow for the insertion of a newline as well as a space when a string
4223 is turned into more than one encoded-word during RFC 2047 encoding. The
4224 Sieve code now uses this.
4226 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4227 data_4xx, lost_connection, tls_required.
4229 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4230 message was being sent as an SMTP response.
4232 PH/61 Add -l and -k options to exicyclog.
4234 PH/62 When verifying, if an address was redirected to one new address, so that
4235 verification continued, and the new address failed or deferred after
4236 having set something in $address_data, the value of $address_data was not
4237 passed back to the ACL. This was different to the case when no
4238 redirection occurred. The value is now passed back in both cases.
4240 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4241 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4242 use login_cap.h, so on its own it isn't the distinguishing feature. The
4243 new name refers directly to the setclassresources() function.
4245 PH/65 Added configuration files for NetBSD3.
4247 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4249 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4250 is preferred over IPv4.
4252 PH/68 The bounce_return_message and bounce_return_body options were not being
4253 honoured for bounces generated during the reception of non-SMTP messages.
4254 In particular, this applied to messages rejected by the ACL. This bug has
4255 been fixed. However, if bounce_return_message is true and bounce_return_
4256 body is false, the headers that are returned for a non-SMTP message
4257 include only those that have been read before the error was detected.
4258 (In the case of an ACL rejection, they have all been read.)
4260 PH/69 The HTML version of the specification is now built in a directory called
4261 spec_html instead of spec.html, because the latter looks like a path with
4262 a MIME-type, and this confuses some software.
4264 PH/70 Catch two compiler warnings in sieve.c.
4266 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4267 function verify_get_ident() calls ip_connect() to connect a socket, but
4268 if the "connect()" function timed out, ip_connect() used to close the
4269 socket. However, verify_get_ident() also closes the socket later, and in
4270 between Exim writes to the log, which may get opened at this point. When
4271 the socket was closed in ip_connect(), the log could get the same file
4272 descriptor number as the socket. This naturally causes chaos. The fix is
4273 not to close the socket in ip_connect(); the socket should be closed by
4274 the function that creates it. There was only one place in the code where
4275 this was missing, in the iplookup router, which I don't think anybody now
4276 uses, but I've fixed it anyway.
4278 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4279 well as to direct DNS lookups. Otherwise the handling of names in host
4280 lists is inconsistent and therefore confusing.
4286 PH/01 Two changes to the default runtime configuration:
4288 (1) Move the checks for relay_from_hosts and authenticated clients from
4289 after to before the (commented out) DNS black list checks.
4291 (2) Add control=submission to the relay_from_hosts and authenticated
4292 clients checks, on the grounds that messages accepted by these
4293 statements are most likely to be submissions.
4295 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4297 (1) Generate an error if the third argument for the ${prvs expansion is
4300 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4303 (3) Reset the variables that are obtained from the first argument of
4304 ${prvscheck and used in the second argument before leaving the code,
4305 because their memory is reclaimed, so using them afterwards may do
4308 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4309 one (it's much easier than Tom thought :-).
4311 (5) Because of (4), we can now allow for the use of $prvscheck_result
4312 inside the third argument.
4314 PH/03 For some reason, the default setting of PATH when running a command from
4315 a pipe transport was just "/usr/bin". I have changed it to
4318 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4319 anything to be listed in the output from -bV.
4321 PH/05 When a filter generated an autoreply, the entire To: header line was
4322 quoted in the delivery log line, like this:
4324 => >A.N.Other <ano@some.domain> <original@ddress> ...
4326 This has been changed so that it extracts the operative address. There
4327 may be more than one such address. If so, they are comma-separated, like
4330 => >ano@some.domain,ona@other.domain <original@ddress> ...
4332 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4333 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4334 not being looked up in the rDNS to get a host name, Exim was showing the
4335 IP address twice in Received: lines, even though the IP addresses were
4336 identical. For example:
4338 Received: from [1.2.3.4] (helo=[1.2.3.4])
4340 However, if the real host name was known, it was omitting the HELO data
4341 if it matched the actual IP address. This has been tidied up so that it
4342 doesn't show the same IP address twice.
4344 PH/07 When both +timestamp and +memory debugging was on, the value given by
4345 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4346 called by the debug printing, thereby overwriting the timestamp buffer.
4347 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4349 PH/08 When the original message was included in an autoreply transport, it
4350 always said "this is a copy of the message, including all the headers",
4351 even if body_only or headers_only was set. It now gives an appropriate
4354 PH/09 Applied a patch from the Sieve maintainer which:
4356 o fixes some comments
4357 o adds the (disabled) notify extension core
4358 o adds some debug output for the result of if/elsif tests
4359 o points to the current vacation draft in the documentation
4360 and documents the missing references header update
4364 o fixes a bug in processing the envelope test (when testing
4365 multiple envelope elements, the last element determined the
4368 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4369 Electronic Mail") by including:
4371 Auto-submitted: auto-generated
4373 in the messages that it generates (bounce messages and others, such as
4374 warnings). In the case of bounce messages for non-SMTP messages, there was
4375 also a typo: it was using "Auto_submitted" (underscore instead of
4376 hyphen). Since every message generated by Exim is necessarily in response
4377 to another message, thes have all been changed to:
4379 Auto-Submitted: auto-replied
4381 in accordance with these statements in the RFC:
4383 The auto-replied keyword:
4385 - SHOULD be used on messages sent in direct response to another
4386 message by an automatic process,
4388 - MUST NOT be used on manually-generated messages,
4390 - MAY be used on Delivery Status Notifications (DSNs) and Message
4391 Disposition Notifications (MDNs),
4393 - MUST NOT be used on messages generated by automatic or periodic
4394 processes, except for messages which are automatic responses to
4397 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4398 to the default Received: header definition.
4400 PH/12 Added log selector acl_warn_skipped (default on).
4402 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4403 variables because (a) they are in the wrong storage pool and (b) even if
4404 they were copied, it wouldn't work properly because of the caching.
4406 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4407 checking when decoding. Apparently there are clients that generate
4408 overlong encoded strings. Why am I not surprised?
4410 PH/15 If the first argument of "${if match_address" was not empty, but did not
4411 contain an "@" character, Exim crashed. Now it writes a panic log message
4412 and treats the condition as false.
4414 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4416 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4417 extension "envelope-auth". The code is finished and in agreement with
4418 other implementations, but there is no documentation so far and in fact,
4419 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4420 not changing the active code.
4422 Print executed "if" and "elsif" statements when debugging is used. This
4423 helps a great deal to understand what a filter does.
4425 Document more things not specified clearly in RFC3028. I had all this
4426 sorted out, when out of a sudden new issues came to my mind. Oops."
4428 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4431 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4432 canonical form (as documented). However, after a host name lookup from
4433 the IP address, check_host() was doing a simple string comparison with
4434 addresses acquired from the DNS when checking that the found name did
4435 have the original IP as one of its addresses. Since any found IPv6
4436 addresses are likely to be in abbreviated form, the comparison could
4437 fail. Luckily, there already exists a function for doing the comparison
4438 by converting both addresses to binary, so now that is used instead of
4439 the text comparison.
4441 PH/20 There was another similar case to PH/19, when a complete host name was
4442 given in a host list; looking up its IP address could give an abbreviated
4443 form, whereas the current host's name might or might not be abbreviated.
4444 The same fix has been applied.
4450 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4451 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4454 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4455 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4457 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4459 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4460 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4461 Also, for "Auto-Submitted", treat anything other than "no" as
4462 non-personal, in accordance with RFC 3834. (Previously it treated
4463 anything starting "auto-" as non-personal.)
4465 TF/01 The control=submission/name=... option had a problem with syntax
4466 errors if the name included a slash character. The /name= option
4467 now slurps the rest of the string, so it can include any characters
4468 but it must come last in the list of options (after /sender_retain
4471 PH/05 Some modifications to the interface to the fake nameserver for the new
4479 TK/01 Added the "success_on_redirect" address verification option. See
4480 NewStuff for rationale and an example.
4482 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4484 PH/02 Patch to exigrep to allow it to work on syslog lines.
4486 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4487 fread() to skip over the body file's header line, because in Cygwin the
4488 header line is locked and is inaccessible.
4490 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4491 co-exist for some time) to make it clear that it is the Exim ID that is
4492 referenced, not the Message-ID: header line.
4494 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4495 string_format() function, because snprintf() does not exist on all
4498 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4499 ${stat: expansion item.
4501 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4502 protocol synchronization error", to keep the pedants happy.
4504 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4505 well as for IRIX systems, when gcc is being used. See the host.c source
4508 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4510 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4513 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4514 transport and to the smtp transport in LMTP mode.
4516 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4518 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4519 run a filter in a subprocess. This could lead to confusion in subsequent
4520 lookups in the parent process. There should also be a search_tidyup() at
4521 the end of the subprocess.
4523 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4524 only if the host matched helo_try_verify_hosts, which caused the
4525 verification to occur when the EHLO/HELO command was issued. The ACL just
4526 tested the remembered result. Now, if a previous verification attempt has
4527 not happened, "verify = helo" does it there and then.
4529 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4531 TK/03 Fix log output including CR from clamd.
4533 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4534 exist provoked a memory error which could cause a segfault.
4536 PH/15 Installed PCRE 6.2
4538 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4540 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4541 of the problem. Specifically, suggested +O2 rather than +O1 for the
4544 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4546 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4547 record, the port was not correctly logged when the outgoing_port log
4548 selector was set (it logged the transort's default port).
4550 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4551 fallback_hosts, and "hosts" in the smtp transport.
4553 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4554 host errors such as "Connection refused".
4556 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4557 authentication with radiusclient 0.4.9:
4559 - Error returned from rc_read_config was caught wrongly
4560 - Username/password not passed on to radius server due to wrong length.
4562 The presumption is that some radiusclient API changes for 4.51/PH/17
4563 were not taken care of correctly. The code is still untested by me (my
4564 Linux distribution still has 0.3.2 of radiusclient), but it was
4565 contributed by a Radius user.
4567 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4568 expanding the "port" option of the smtp transport.
4570 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4571 while reading a MIME header. Thanks to Tom Hughes for a patch.
4573 PH/24 Include config.h inside local_scan.h so that configuration settings are
4576 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4577 that in an ACL for RCPT (for example), you can examine exactly what was
4580 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4581 commands, but it was not correctly comparing the address with the actual
4582 client host address. Thus, it would show the EHLO address in Received:
4583 header lines when this was not necessary.
4585 PH/27 Added the % operator to ${eval:}.
4587 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4588 it should be ignoring failures (because with -C, for example, it has lost
4589 privilege). It wasn't ignoring creation failures other than "already
4592 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4595 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4596 because an input line was too long, either on its own, or by virtue of
4597 too many continuations, the temporary file was not being removed, and the
4598 return code was incorrect.
4600 PH/31 Missing "BOOL" in function definition in filtertest.c.
4602 PH/32 Applied Sieve patches from the maintainer.
4604 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4606 PH/33 Added "verify = not_blind".
4608 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4609 Local/Makefile (with some defaults set). These are used in built scripts
4610 such as exicyclog, but they have never been used in the exim_install
4611 script (though there are many overriding facilities there). I have
4612 arranged that the exim_install script now takes note of these two
4615 PH/35 Installed configuration files for Dragonfly.
4617 PH/36 When a locally submitted message by a trusted user did not contain a
4618 From: header, and the sender address was obtained from -f or from an SMTP
4619 MAIL command, and the trusted user did not use -F to supply a sender
4620 name, $originator_name was incorrectly used when constructing a From:
4621 header. Furthermore, $originator_name was used for submission mode
4622 messages from external hosts without From: headers in a similar way,
4623 which is clearly wrong.
4625 PH/37 Added control=suppress_local_fixups.
4627 PH/38 When log_selector = +received_sender was set, and the addition of the
4628 sender made the log line's construction buffer exactly full, or one byte
4629 less than full, an overflow happened when the terminating "\n" was
4632 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4633 when the result of a list match is failure because a DNS lookup failed.
4635 PH/40 RM_COMMAND is now used in the building process.
4637 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4638 the "build-* directories that it finds.
4640 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4641 address in a domain literal was a prefix of an interface address.
4643 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4644 when verifying a sender address, unless rewrite_headers is false.
4646 PH/44 Wrote a long comment about why errors_to addresses are verified as
4647 recipients, not senders.
4649 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4650 the ratelimit ACL was added.
4652 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4654 PH/46 Added extra information about PostgreSQL errors to the error string.
4656 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4657 avoiding the need to install special zones in a real server. This is
4658 backwards compatible; if it can't find the fake resolver, it drops back.
4659 Thus, both old and new test suites can be run.
4661 TF/02 Added util/ratelimit.pl
4663 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4664 clock is set back in time.
4666 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4667 Candler <B.Candler@pobox.com>.
4669 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4670 OK for addresses that are the result of redirections.
4672 PH/48 A number of further additions for the benefit of the new test suite,
4673 including a fake gethostbyname() that interfaces to the fake DNS resolver
4676 TF/06 The fix for widen_domains has also been applied to qualify_single and
4677 search_parents which are the other dnslookup options that can cause
4680 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4683 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4685 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4686 service" if it encountered an MX record with an empty target hostname.
4687 The message is now "an MX or SRV record indicated no SMTP service".
4689 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4690 if the DNS of the sending site is misconfigured. This is quite a
4691 common situation. This change restores the behaviour of treating a
4692 helo verification defer as a failure.
4694 PH/51 If self=fail was set on a router, the bounce message did not include the
4695 actual error message.
4701 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4703 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4704 command itself ended in error, the underlying message about the transport
4705 filter timeout was being overwritten with the pipe command error. Now the
4706 underlying error message should be appended to the second error message.
4708 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4710 PH/02 Reduce the amount of output that "make" produces by default. Full output
4711 can still be requested.
4713 PH/03 The warning log line about a condition test deferring for a "warn" verb
4714 was being output only once per connection, rather than after each
4715 occurrence (because it was using the same function as for successful
4716 "warn" verbs). This seems wrong, so I have changed it.
4718 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4719 it should not have, which might have caused a crash in the right
4720 circumstances, but probably never did.
4722 PH/04 Installed a modified version of Tony Finch's patch to make submission
4723 mode fix the return path as well as the Sender: header line, and to
4724 add a /name= option so that you can make the user's friendly name appear
4727 TF/03 Added the control = fakedefer ACL modifier.
4729 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4730 Mark Lowes for thorough testing.
4732 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4734 TK/03 Merged latest SRS patch from Miles Wilton.
4736 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4737 with the definition in sysexits.h (which is #included earlier).
4738 Fortunately, Exim does not actually use EX_OK. The code used to try to
4739 preserve the sysexits.h value, by assuming that macro definitions were
4740 scanned for macro replacements. I have been disabused of this notion,
4741 so now the code just undefines EX_OK before #including unistd.h.
4743 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4744 in the smtp transport. When a block could not be written in a single
4745 write() function, the timeout was being re-applied to each part-write.
4746 This seems wrong - if the receiver was accepting one byte at a time it
4747 would take for ever. The timeout is now adjusted when this happens. It
4748 doesn't have to be particularly precise.
4750 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4751 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4753 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4754 without a domain if the check to <postmaster@domain> fails.
4756 SC/01 Eximstats: added -xls and the ability to specify output files
4757 (patch written by Frank Heydlauf).
4759 SC/02 Eximstats: use FileHandles for outputting results.
4761 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4763 SC/04 Eximstats: fixed display of large numbers with -nvr option
4765 SC/05 Eximstats: fixed merging of reports with empty tables.
4767 SC/06 Eximstats: added the -include_original_destination flag
4769 SC/07 Eximstats: removed tabs and trailing whitespace.
4771 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4773 TK/06 MBOX spool code: Add real "From " MBOX separator line
4774 so the .eml file is really in mbox format (even though
4775 most programs do not really care). Patch from Alex Miller.
4777 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4778 The latter is generated from $received_to and is only set if the
4779 message has one envelope recipient. SA can use these headers,
4780 obviously out-of-the-box. Patch from Alex Miller.
4782 PH/08 The ${def test on a variable was returning false if the variable's
4783 value was "0", contrary to what the specification has always said!
4784 The result should be true unless the variable is empty.
4786 PH/09 The syntax error of a character other than { following "${if
4787 def:variable_name" (after optional whitespace) was not being diagnosed.
4788 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4789 accidental colon was present, for example, could give incorrect results.
4791 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4792 result is used (not including appendfile, where other changes are about
4795 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4796 This involved changing a lot of size variables from int to off_t. It
4797 should work with maildirs and everything.
4799 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4800 spamd dying while we are connected to it.
4802 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4805 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4806 function for BDB 4.3.
4808 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4810 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4811 the sources for PCRE have been re-arranged and more files are now
4814 PH/15 The code I had for printing potentially long long variables in PH/11
4815 above was not the best (it lost precision). The length of off_t variables
4816 is now inspected at build time, and an appropriate printing format (%ld
4817 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4818 to be "long long int" or "long int". This is needed for the internal
4819 formatting function string_vformat().
4821 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4822 the configuration file to be ":syslog", then the script "guesses" where
4823 the logs files are, rather than using the compiled in default. In our
4824 case the guess is not the same as the compiled default, so the script
4825 suddenly stopped working when I started to use syslog. The patch checks
4826 to see if log_file_path is "". If so, it attempts to read it from exim
4827 with no configuration file to get the compiled in version, before it
4828 falls back to the previous guessing code."
4830 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4831 implementing BATV in an Exim configuration. See NewStuff for the gory
4834 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4835 Makefile that are specific to HP-UX.
4837 PH/18 If the "use_postmaster" option was set for a recipient callout together
4838 with the "random" option, the postmaster address was used as the MAIL
4839 FROM address for the random test, but not for the subsequent recipient
4840 test. It is now used for both.
4842 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4843 patch removes a few documentation additions to RFC 3028, because the
4844 latest draft now contains them. It adds the new en;ascii-case comparator
4845 and a new error check for 8bit text in MIME parts. Comparator and
4846 require names are now matched exactly. I enabled the subaddress
4847 extension, but it is not well tested yet (read: it works for me)."
4849 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4850 rework some of the code of TK/09 above to avoid the hardwired use of
4851 "%lld" and "long long". Replaced the call to snprintf() with a call to
4854 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4855 records point to non-existent hosts", "retry timeout exceeded", and
4856 "retry time not reached for any host after a long failure period".
4858 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4859 experimental DomainKeys support:
4861 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4862 (2) On an error such as an illegally used "control", the wrong name for
4863 the control was given.
4865 These problems did NOT occur unless DomainKeys support was compiled.
4867 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4869 PH/24 Added ${if match_ip condition.
4871 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4872 databases so that it will be absolutely obvious if a crash occurs in the
4873 DB library. This is a regular occurrence (often caused by mis-matched
4876 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4877 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4878 picked up on a user's system that detects such things. There doesn't seem
4879 to be a gcc warning option for this - only an attribute that has to be
4880 put on the function's prototype. It seems that in Fedora Core 4 they have
4881 set this on a number of new functions. No doubt there will be more in due
4884 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4885 specify a transport. However, if an address that was verified by such a
4886 router was the subject of a callout, Exim crashed because it tried to
4887 read the rcpt_include_affixes from the non-existent transport. Now it
4888 just assumes that the setting of that option is false. This bug was
4889 introduced by 4.51/PH/31.
4891 PH/28 Changed -d+all to exclude +memory, because that information is very
4892 rarely of interest, but it makes the output a lot bigger. People tend to
4893 do -d+all out of habit.
4895 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4896 code in os-type was giving problems when libc.so lives in lib64, like on
4899 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4900 aren't the modern standard, and it seems that some systems' include files
4901 don't always have them. Exim was already checking for some of the newer
4902 ones like T_AAAA, and defining it itself. I've added checks for all the
4903 record types that Exim uses.
4905 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4906 not automatically generating a new one, as it is supposed to. This
4907 prevented TLS from working. If the file did exist, but contained invalid
4908 data, a new version was generated, as expected. It was only the case of a
4909 non-existent file that was broken.
4911 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4912 with a change in libdomainkeys > 0.64.
4914 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4915 from DNS. If the selector record carries the flag, it now has
4916 precedence over the domain-wide flag.
4918 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4920 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4921 the use of a port name that isn't defined in /etc/services) occurred, the
4922 message was deferred as in a normal delivery, and thus remained on the
4923 spool, instead of being failed because of the mua_wrapper setting. This
4924 is now fixed, and I tidied up some of the mua_wrapper messages at the
4927 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4928 the messages in a hash of arrays rather than using individual hashes.
4929 This is a bit cleaner and results in dramatic memory savings, albeit
4930 at a slight CPU cost.
4932 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4933 as requested by Marc Sherman.
4935 SC/10 Eximstats: added histograms for user specified patterns as requested
4938 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4940 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4941 fopen() in the content-scanning modules that did not already have it.
4947 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4948 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4950 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4952 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4954 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4955 to test Sieve filters that use "vacation".
4957 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4958 that changes the way the GnuTLS parameters are stored in the cache file.
4959 The new format can be generated externally. For backward compatibility,
4960 if the data in the cache doesn't make sense, Exim assumes it has read an
4961 old-format file, and it generates new data and writes a new file. This
4962 means that you can't go back to an older release without removing the
4965 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4966 work if there are any delivery delays because "one_time" forces the
4967 parent to be marked "delivered", so its unseen clone is never tried
4968 again. For this reason, Exim now forbids the simultaneous setting of
4971 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4972 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4973 entry if you want to know the details. The fix, however, neglected to
4974 consider the case when local delivery batching is involved. The test for
4975 "previously delivered" was not happening when checking to see if an
4976 address could be batched with a previous (undelivered) one; under
4977 certain circumstances this could lead to multiple deliveries to the same
4980 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4981 in its include files, and this causes problems building Exim.
4983 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4984 header_syntax) but Exim was just ignoring anything given after a slash.
4985 In particular, this caused confusion with an attempt to use "verify =
4986 reverse_host_lookup/defer_ok". An error is now given when options are
4987 supplied for verify items that do not have them. (Maybe reverse_host_
4988 lookup should have a defer_ok option, but that's a different point.)
4990 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4991 defined by RFC 821) to 2048, because there were problems with some AUTH
4992 commands, and RFC 1869 says the size should be increased for extended
4993 SMTP commands that take arguments.
4995 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4998 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4999 "unknown" error; now it says that the functionality isn't in the binary.
5001 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5002 an address' error message when a string expansion fails (syntax or
5003 whatever). Otherwise the password may appear in the log. Following change
5004 PH/42 below, there is no longer a chance of it appearing in a bounce
5007 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5009 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5010 of its addresses was ever tried. (Bugzilla bug #2.)
5012 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5013 the result incorrectly in the debug output. (It correctly added a newline
5014 to what was transported.)
5016 TF/01 Added $received_time.
5018 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5019 commented out examples of how to interface to a virus scanner and to
5020 SpamAssassin. Also added commented examples of av_scanner and
5021 spamd_address settings.
5023 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5024 and controls are allowed in which ACLs. There were a couple of minor
5025 errors. Some of the entries in the conditions table (which is a table of
5026 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5027 negation of where the condition IS allowed.
5029 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5031 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5032 header file does not have a version number, so I've had to invent a new
5033 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5034 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5035 radiusclient), but it was contributed by a Radius user.
5037 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5038 files or directories, for OpenSSL.
5040 PH/19 When an Exim process that is running as root has to create an Exim log
5041 file, it does so in a subprocess that runs as exim:exim so as to get the
5042 ownership right at creation (otherwise, other Exim processes might see
5043 the file with the wrong ownership). There was no test for failure of this
5044 fork() call, which would lead to the process getting stuck as it waited
5045 for a non-existent subprocess. Forks do occasionally fail when resources
5046 run out. I reviewed all the other calls to fork(); they all seem to check
5049 PH/20 When checking for unexpected SMTP input at connect time (before writing
5050 the banner), Exim was not dealing correctly with a non-positive return
5051 from the read() function. If the client had disconnected by this time,
5052 the result was a log entry for a synchronization error with an empty
5053 string after "input=" when read() returned zero. If read() returned -1
5054 (an event I could not check), uninitialized data bytes were printed.
5055 There were reports of junk text (parts of files, etc) appearing after
5058 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5060 PH/22 Added support for macro redefinition, and (re)definition in between
5061 driver and ACL definitions.
5063 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5064 forgetting to use the resulting value; it was using the unexpanded value.
5066 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5067 hadn't been configured. The fix is from Juergen Kreileder, who
5068 understands it better than I do:
5070 "Here's what I see happening with three configured cyrus_sasl
5071 authenticators configured (plain, login, cram-md5):
5073 On startup auth_cyrus_sasl_init() gets called for each of these.
5074 This means three calls to sasl_listmech() without a specified mech_list.
5075 => SASL tests which mechs of all available mechs actually work
5076 => three warnings about OTP not working
5077 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5079 With the patch, sasl_listmech() also gets called three times. But now
5080 SASL's mech_list option is set to the server_mech specified in the the
5081 authenticator. Or in other words, the answer from sasl_listmech()
5082 gets limited to just the mech you're testing for (which is different
5084 => the return list contains just 'plain' or 'login', 'cram-md5' or
5085 nothing depending on the value of ob->server_mech.
5087 I've just tested the patch: Authentication still works fine,
5088 unavailable mechs specified in the exim configuration are still
5089 caught, and the auth.log warnings about OTP are gone."
5091 PH/25 When debugging is enabled, the contents of the command line are added
5092 to the debugging output, even when log_selector=+arguments is not
5095 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5096 answer is "GNU", and only if the return is "GNU/something" is the answer
5099 PH/27 $acl_verify_message is now set immediately after the failure of a
5100 verification in an ACL, and so is available in subsequent modifiers. In
5101 particular, the message can be preserved by coding like this:
5103 warn !verify = sender
5104 set acl_m0 = $acl_verify_message
5106 Previously, $acl_verify_message was set only while expanding "message"
5107 and "log_message" when a very denied access.
5109 PH/28 Modified OS/os.c-Linux with
5111 -#ifndef OS_LOAD_AVERAGE
5112 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5114 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5115 nomenclature these days.)
5117 PH/29 Installed patch from the Sieve maintainer that adds the options
5118 sieve_useraddress and sieve_subaddress to the redirect router.
5120 PH/30 In these circumstances:
5121 . Two addresses routed to the same list of hosts;
5122 . First host does not offer TLS;
5123 . First host accepts first address;
5124 . First host gives temporary error to second address;
5125 . Second host offers TLS and a TLS session is established;
5126 . Second host accepts second address.
5127 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5128 and peerdn, if requested) that were in fact used only for the second
5131 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5132 attention to any local part prefix or suffix that was matched by the
5133 router that accepted the address. It now behaves in the same way as it
5134 does for delivery: the affixes are removed from the local part unless
5135 rcpt_include_affixes is set on the transport.
5137 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5138 timeout during the DATA phase of an incoming message.
5140 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5141 applied a patch sanctioned by the Sieve maintainer.
5143 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5144 the uid or gid is negative. A case of a negative gid caused this to be
5145 noticed. The fix allows for either to be negative.
5147 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5148 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5151 PH/36 The change PH/12 above was broken. Fixed it.
5153 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5154 the grounds that routing the same address twice would always produce the
5155 same answer. This might have been true once, but it is certainly no
5156 longer true now. Routing a child address may depend on the previous
5157 routing that produced that child. Some complicated redirection strategies
5158 went wrong when messages had multiple recipients, and made Exim's
5159 behaviour dependent on the order in which the addresses were given.
5161 I have moved the duplicate checking until after the routing is complete.
5162 Exim scans the addresses that are assigned to local and remote
5163 transports, and removes any duplicates. This means that more work will be
5164 done, as duplicates will always all be routed, but duplicates are
5165 presumably rare, so I don't expect this is of any significance.
5167 For deliveries to pipes, files, and autoreplies, the duplicate checking
5168 still happens during the routing process, since they are not going to be
5171 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5172 It corrects a timeout issue with spamd. This is Ian's comment: "The
5173 background is that sometimes spamd either never reads data from a
5174 connection it has accepted, or it never writes response data. The exiscan
5175 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5176 blindly assumes that writes won't block so it may never time out."
5178 PH/39 Allow G after quota size as well as K and M.
5180 PH/40 The value set for $authenticated_id in an authenticator may not contain
5181 binary zeroes or newlines because the value is written to log lines and
5182 to spool files. There was no check on this. Now the value is run through
5183 the string_printing() function so that such characters are converted to
5184 printable escape sequences.
5186 PH/41 $message_linecount is a new variable that contains the total number of
5187 lines in the message. Compare $body_linecount, which is the count for the
5190 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5191 bounce and delay warning messages, except in certain special cases, which
5194 (a) An SMTP error message from a remote host;
5195 (b) A message specified in a :fail: redirection;
5196 (c) A message specified in a "fail" command in a system filter;
5197 (d) A message specified in a FAIL return from the queryprogram router;
5198 (e) A message specified by the cannot_route_message router option.
5200 In these cases only, Exim does include the error details in bounce and
5201 warning messages. There are also a few cases where bland messages such
5202 as "unrouteable address" or "local delivery error" are given.
5204 PH/43 $value is now also set for the "else" part of a ${run expansion.
5206 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5207 being worked on, but at least Exim now implements the latest version to
5210 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5211 process to complete was treated as a delivery failure, a timeout while
5212 writing the message to the pipe was logged, but erroneously treated as a
5213 successful delivery. Such timeouts include transport filter timeouts. For
5214 consistency with the overall process timeout, these timeouts are now
5215 treated as errors, giving rise to delivery failures by default. However,
5216 there is now a new Boolean option for the pipe transport called
5217 timeout_defer, which, if set TRUE, converts the failures into defers for
5218 both kinds of timeout. A transport filter timeout is now identified in
5221 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5222 systems where "make" and "gmake" are different, calling "gmake" at top
5223 level broke things. I've arranged for the value of $(MAKE) to be passed
5224 from the Makefile to this script so that it can call the same version of
5228 A note about Exim versions 4.44 and 4.50
5229 ----------------------------------------
5231 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5232 changes of various kinds. As a consequence, a big documentation update was
5233 needed. This delayed the release for rather longer than seemed good, especially
5234 in the light of a couple of (minor) security issues. Therefore, the changes
5235 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5236 release. So 4.44 and 4.50 are in effect two different branches that both start
5239 I have left the 4.50 change log unchanged; it contains all the changes since
5240 4.43. The change log for 4.44 is below; many of its items are identical to
5241 those for 4.50. This seems to be the most sensible way to preserve the
5242 historical information.
5248 1. Minor wording change to the doc/README.SIEVE file.
5250 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5251 computation of the current number of files was incorrect.
5253 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5254 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5257 4. Give more explanation in the error message when the command for a transport
5258 filter fails to execute.
5260 5. There are several places where Exim runs a non-Exim command in a
5261 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5262 was being done only for the command run by the queryprogram router. It is
5263 now done for all such subprocesses. The other cases are: ${run, transport
5264 filters, and the commands run by the lmtp and pipe transports.
5266 6. Added CONFIGURE_GROUP build-time option.
5268 7. Some older OS have a limit of 256 on the maximum number of file
5269 descriptors. Exim was using setrlimit() to set 1000 as a large value
5270 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5271 systems. I've change it so that if it can't get 1000, it tries for 256.
5273 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5274 was an oversight, and furthermore, ever since the addition of extra
5275 controls (e.g. 4.43/32), the checks on when to allow different forms of
5276 "control" were broken. There should now be diagnostics for all cases when a
5277 control that does not make sense is encountered.
5279 9. Added the /retain_sender option to "control=submission".
5281 10. $recipients is now available in the predata ACL (oversight).
5283 11. Tidy the search cache before the fork to do a delivery from a message
5284 received from the command line. Otherwise the child will trigger a lookup
5285 failure and thereby defer the delivery if it tries to use (for example) a
5286 cached ldap connection that the parent has called unbind on.
5288 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5289 of $address_data from the recipient verification was clobbered by the
5290 sender verification.
5292 13. The value of address_data from a sender verification is now available in
5293 $sender_address_data in subsequent conditions in the ACL statement.
5295 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5297 15. Added a new option "connect=<time>" to callout options, to set a different
5300 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5301 was its contents. (It was OK if the option was not defined at all.)
5303 17. A "Completed" log line is now written for messages that are removed from
5304 the spool by the -Mrm option.
5306 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5307 information about exactly what failed.
5309 19. Added -dd to debug only the daemon process.
5311 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5312 handles timeouts, both on the server side and network timeouts. Renamed the
5313 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5315 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5316 It is now set to "smtps".
5318 22. $host_address is now set to the target address during the checking of
5319 ignore_target_hosts.
5321 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5322 being passed; this would have caused $sender_host_name to have been used if
5323 matching the list had actually called for a host name (not very likely,
5324 since this list is usually IP addresses). A host name is now passed as
5327 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5328 code that specifies a non-restarting handler (typically sigaction() in
5329 modern systems) in an attempt to fix a rare and obscure crash bug.
5331 25. Narrowed the window for a race in the daemon that could cause it to ignore
5332 SIGCHLD signals. This is not a major problem, because they are used only to
5333 wake it up if nothing else does.
5335 26. A malformed maildirsize file could cause Exim to calculate negative values
5336 for the mailbox size or file count. Odd effects could occur as a result.
5337 The maildirsize information is now recalculated if the size or filecount
5340 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5341 support for a long time. Removed HAVE_SYS_VFS_H.
5343 28. Installed the latest version of exipick from John Jetmore.
5345 29. In an address list, if the pattern was not a regular expression, an empty
5346 subject address (from a bounce message) matched only if the pattern was an
5347 empty string. Non-empty patterns were not even tested. This was the wrong
5348 because it is perfectly reasonable to use an empty address as part of a
5349 database query. An empty address is now tested by patterns that are
5350 lookups. However, all the other forms of pattern expect the subject to
5351 contain a local part and a domain, and therefore, for them, an empty
5352 address still always fails if the pattern is not itself empty.
5354 30. Exim went into a mad DNS loop when attempting to do a callout where the
5355 host was specified on an smtp transport, and looking it up yielded more
5356 than one IP address.
5358 31. Re-factored the code for checking spool and log partition space into a
5359 function that finds that data and another that does the check. The former
5360 is then used to implement four new variables: $spool_space, $log_space,
5361 $spool_inodes, and $log_inodes.
5363 32. The RFC2047 encoding function was originally intended for short strings
5364 such as real names; it was not keeping to the 75-character limit for
5365 encoded words that the RFC imposes. It now respects the limit, and
5366 generates multiple encoded words if necessary. To be on the safe side, I
5367 have increased the buffer size for the ${rfc2047: expansion operator from
5370 33. It is now permitted to omit both strings after an "if" condition; if the
5371 condition is true, the result is "true". As before, when the second string
5372 is omitted, a false condition yields an empty string. This makes it less
5373 cumbersome to write custom ACL and router conditions.
5375 34. Failure to deliver a bounce message always caused it to be frozen, even if
5376 there was an errors_to setting on the router. The errors_to setting is now
5379 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5380 canonical form (fully expanded) before being placed in
5381 $sender_host_address.
5383 36. The table in the code that translates DNS record types into text (T_A to
5384 "A" for instance) was missing entries for NS and CNAME. It is just possible
5385 that this could have caused confusion if both these types were looked up
5386 for the same domain, because the text type is used as part of Exim's
5387 per-process caching. But the chance of anyone hitting this buglet seems
5390 37. The dnsdb lookup has been extended in a number of ways.
5392 (1) There is a new type, "zns", which walks up the domain tree until it
5393 finds some nameserver records. It should be used with care.
5395 (2) There is a new type, "mxh", which is like "mx" except that it returns
5396 just the host names, not the priorities.
5398 (3) It is now possible to give a list of domains (or IP addresses) to be
5399 looked up. The behaviour when one of the lookups defers can be
5400 controlled by a keyword.
5402 (4) It is now possible to specify the separator character for use when
5403 multiple records are returned.
5405 38. The dnslists ACL condition has been extended: it is now possible to supply
5406 a list of IP addresses and/or domains to be looked up in a particular DNS
5409 39. Added log_selector=+queue_time_overall.
5411 40. When running the queue in the test harness, wait just a tad after forking a
5412 delivery process, to get repeatability of debugging output.
5414 41. Include certificate and key file names in error message when GnuTLS fails
5415 to set them up, because the GnuTLS error message doesn't include the name
5416 of the failing file when there is a problem reading it.
5418 42. Allow both -bf and -bF in the same test run.
5420 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5422 44. The "Exiscan patch" is now merged into the mainline Exim source.
5424 45. Sometimes the final signoff response after QUIT could fail to get
5425 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5426 < 0 before doing a fflush(). This bug looks as though it goes back to the
5427 introduction of TLS in release 3.20, but "sometimes" must have been rare
5428 because the tests only now provoked it.
5430 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5431 (this can affect the format of dates).
5433 47. exim_tidydb, when checking for the continued existence of a message for
5434 which it has found a message-specific retry record, was not finding
5435 messages that were in split spool directories. Consequently, it was
5436 deleting retry records that should have stayed in existence.
5438 48. Steve fixed some bugs in eximstats.
5440 49. The SPA authentication driver was not abandoning authentication and moving
5441 on to the next authenticator when an expansion was forced to fail,
5442 contradicting the general specification for all authenticators. Instead it
5443 was generating a temporary error. It now behaves as specified.
5445 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5446 (the order specifies the preference for clients). The order is now AES256,
5447 AES128, 3DES, ARCFOUR128.
5449 51. Small patch to Sieve code - explicitly set From: when generating an
5452 52. Exim crashed if a remote delivery caused a very long error message to be
5453 recorded - for instance if somebody sent an entire SpamAssassin report back
5454 as a large number of 550 error lines. This bug was coincidentally fixed by
5455 increasing the size of one of Exim's internal buffers (big_buffer) that
5456 happened as part of the Exiscan merge. However, to be on the safe side, I
5457 have made the code more robust (and fixed the comments that describe what
5460 53. Now that there can be additional text after "Completed" in log lines (if
5461 the queue_time_overall log selector is set), a one-byte patch to exigrep
5462 was needed to allow it to recognize "Completed" as not the last thing in
5465 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5466 patch that reportedly fixes this has been added. I am not expert enough to
5467 create a test for it. This is what the patch creator wrote:
5469 "I found a little strange behaviour of ldap code when working with
5470 Windows 2003 AD Domain, where users was placed in more than one
5471 Organization Units. When I tried to give exim partial DN, the exit code
5472 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5473 But simultaneously result of request was absolutely normal ldap result,
5474 so I produce this patch..."
5476 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5477 REFERENCE, so I have modified the code to exclude the patch when that macro
5480 55. Some experimental protocols are using DNS PTR records for new purposes. The
5481 keys for these records are domain names, not reversed IP addresses. The
5482 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5483 leaves it alone. Component reversal etc. now happens only for IP addresses.
5486 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5488 57. Double the size of the debug message buffer (to 2048) so that more of very
5489 long debug lines gets shown.
5491 58. The exicyclog utility now does better if the number of log files to keep
5492 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5494 59. Two changes related to the smtp_active_hostname option:
5496 (1) $smtp_active_hostname is now available as a variable.
5497 (2) The default for smtp_banner uses $smtp_active_hostname instead
5498 of $primary_hostname.
5500 60. The host_aton() function is supposed to be passed a string that is known
5501 to be a valid IP address. However, in the case of IPv6 addresses, it was
5502 not checking this. This is a hostage to fortune. Exim now panics and dies
5503 if the condition is not met. A case was found where this could be provoked
5504 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5505 components; fortuitously, this particular loophole had already been fixed
5506 by change 4.50/55 above.
5508 If there are any other similar loopholes, the new check in host_aton()
5509 itself should stop them being exploited. The report I received stated that
5510 data on the command line could provoke the exploit when Exim was running as
5511 exim, but did not say which command line option was involved. All I could
5512 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5513 running as the user.
5516 61. There was a buffer overflow vulnerability in the SPA authentication code
5517 (which came originally from the Samba project). I have added a test to the
5518 spa_base64_to_bits() function which I hope fixes it.
5521 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5522 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5524 63. The daemon start-up calls getloadavg() while still root for those OS that
5525 need the first call to be done as root, but it missed one case: when
5526 deliver_queue_load_max is set with deliver_drop_privilege. This is
5527 necessary for the benefit of the queue runner, because there is no re-exec
5528 when deliver_drop_privilege is set.
5530 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5531 This has been fixed.
5533 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5534 was in use, was not putting the data itself into the right store pool;
5535 consequently, it could be overwritten for a subsequent message in the same
5536 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5539 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5541 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5542 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5543 the function were treating the return as a boolean value, which happened to
5544 work because 0=false and not-0=true, but is not correct code.
5546 68. The host_aton() function was not handling scoped IPv6 addresses (those
5547 with, for example, "%eth0" on the end) correctly.
5549 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5550 negated items (that is, ~something) in unsigned ints. Some compilers
5551 apparently mutter when there is no cast.
5553 70. If an address verification called from an ACL failed, and did not produce a
5554 user-specific message (i.e. there was only a "system" message), nothing was
5555 put in $acl_verify_message. In this situation, it now puts the system
5558 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5559 session; change 4.31/43 added the unwanted input to the log line - except
5560 that it did not do this in the start of session case. It now does.
5562 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5563 This is wrong and can cause the other end to generate a synchronization
5564 error if it is another Exim or anything else that does the synchronization
5565 check. A QUIT command is no longer sent after a timeout.
5567 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5568 during host lookups.
5570 74. The defer_ok option of callout verification was not working if it was used
5571 when verifying addresses in header lines, that is, for this case:
5573 verify = header_sender/callout=defer_ok
5575 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5576 those file descriptors could be used for SMTP connections. If anything
5577 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5578 could be sent to the SMTP client, causing chaos. The daemon now opens
5579 stdin, stdout, and stderr to /dev/null when it puts itself into the
5582 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5583 log by default. The user can override this with suitable Perl magic.
5585 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5586 the log message when discard triggers, was not working for the DATA ACL or
5587 for the non-SMTP ACL.
5589 78. Error message wording change in sieve.c.
5591 79. If smtp_accept_max_per_host was set, the number of connections could be
5592 restricted to fewer than expected, because the daemon was trying to set up
5593 a new connection before checking whether the processes handling previous
5594 connections had finished. The check for completed processes is now done
5595 earlier. On busy systems, this bug wouldn't be noticed because something
5596 else would have woken the daemon, and it would have reaped the completed
5599 80. If a message was submitted locally by a user whose login name contained one
5600 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5601 It caused a spool format error. I have fixed the spool reading code. A
5602 related problem was that the "from" clause in the Received: line became
5603 illegal because of the space(s). It is now covered by ${quote_local_part.
5605 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5608 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5610 83. Patch from Sieve maintainer to fix unterminated string problem in
5611 "vacation" handling.
5613 84. Some minor changes to the Linux configuration files to help with other
5614 OS variants using glibc.
5616 85. One more patch for Sieve to update vacation handling to latest spec.
5619 ----------------------------------------------------
5620 See the note above about the 4.44 and 4.50 releases.
5621 ----------------------------------------------------
5627 1. Change 4.43/35 introduced a bug that caused file counts to be
5628 incorrectly computed when quota_filecount was set in an appendfile
5631 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5632 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5635 3. Give more explanation in the error message when the command for a transport
5636 filter fails to execute.
5638 4. There are several places where Exim runs a non-Exim command in a
5639 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5640 was being done only for the command run by the queryprogram router. It is
5641 now done for all such subprocesses. The other cases are: ${run, transport
5642 filters, and the commands run by the lmtp and pipe transports.
5644 5. Some older OS have a limit of 256 on the maximum number of file
5645 descriptors. Exim was using setrlimit() to set 1000 as a large value
5646 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5647 systems. I've change it so that if it can't get 1000, it tries for 256.
5649 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5650 was an oversight, and furthermore, ever since the addition of extra
5651 controls (e.g. 4.43/32), the checks on when to allow different forms of
5652 "control" were broken. There should now be diagnostics for all cases when a
5653 control that does not make sense is encountered.
5655 7. $recipients is now available in the predata ACL (oversight).
5657 8. Tidy the search cache before the fork to do a delivery from a message
5658 received from the command line. Otherwise the child will trigger a lookup
5659 failure and thereby defer the delivery if it tries to use (for example) a
5660 cached ldap connection that the parent has called unbind on.
5662 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5663 of $address_data from the recipient verification was clobbered by the
5664 sender verification.
5666 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5667 was its contents. (It was OK if the option was not defined at all.)
5669 11. A "Completed" log line is now written for messages that are removed from
5670 the spool by the -Mrm option.
5672 12. $host_address is now set to the target address during the checking of
5673 ignore_target_hosts.
5675 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5676 being passed; this would have caused $sender_host_name to have been used if
5677 matching the list had actually called for a host name (not very likely,
5678 since this list is usually IP addresses). A host name is now passed as
5681 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5682 code that specifies a non-restarting handler (typically sigaction() in
5683 modern systems) in an attempt to fix a rare and obscure crash bug.
5685 15. Narrowed the window for a race in the daemon that could cause it to ignore
5686 SIGCHLD signals. This is not a major problem, because they are used only to
5687 wake it up if nothing else does.
5689 16. A malformed maildirsize file could cause Exim to calculate negative values
5690 for the mailbox size or file count. Odd effects could occur as a result.
5691 The maildirsize information is now recalculated if the size or filecount
5694 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5695 support for a long time. Removed HAVE_SYS_VFS_H.
5697 18. Updated exipick to current release from John Jetmore.
5699 19. Allow an empty sender to be matched against a lookup in an address list.
5700 Previously the only cases considered were a regular expression, or an
5703 20. Exim went into a mad DNS lookup loop when doing a callout where the
5704 host was specified on the transport, if the DNS lookup yielded more than
5707 21. The RFC2047 encoding function was originally intended for short strings
5708 such as real names; it was not keeping to the 75-character limit for
5709 encoded words that the RFC imposes. It now respects the limit, and
5710 generates multiple encoded words if necessary. To be on the safe side, I
5711 have increased the buffer size for the ${rfc2047: expansion operator from
5714 22. Failure to deliver a bounce message always caused it to be frozen, even if
5715 there was an errors_to setting on the router. The errors_to setting is now
5718 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5719 canonical form (fully expanded) before being placed in
5720 $sender_host_address.
5722 24. Updated eximstats to version 1.33
5724 25. Include certificate and key file names in error message when GnuTLS fails
5725 to set them up, because the GnuTLS error message doesn't include the name
5726 of the failing file when there is a problem reading it.
5728 26. Expand error message when OpenSSL has problems setting up cert/key files.
5731 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5732 (this can affect the format of dates).
5734 28. exim_tidydb, when checking for the continued existence of a message for
5735 which it has found a message-specific retry record, was not finding
5736 messages that were in split spool directories. Consequently, it was
5737 deleting retry records that should have stayed in existence.
5739 29. eximstats updated to version 1.35
5740 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5741 1.35 - bugfix such that pie charts by volume are generated correctly
5743 30. The SPA authentication driver was not abandoning authentication and moving
5744 on to the next authenticator when an expansion was forced to fail,
5745 contradicting the general specification for all authenticators. Instead it
5746 was generating a temporary error. It now behaves as specified.
5748 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5749 (the order specifies the preference for clients). The order is now AES256,
5750 AES128, 3DES, ARCFOUR128.
5752 31. Small patch to Sieve code - explicitly set From: when generating an
5755 32. Exim crashed if a remote delivery caused a very long error message to be
5756 recorded - for instance if somebody sent an entire SpamAssassin report back
5757 as a large number of 550 error lines. This bug was coincidentally fixed by
5758 increasing the size of one of Exim's internal buffers (big_buffer) that
5759 happened as part of the Exiscan merge. However, to be on the safe side, I
5760 have made the code more robust (and fixed the comments that describe what
5763 33. Some experimental protocols are using DNS PTR records for new purposes. The
5764 keys for these records are domain names, not reversed IP addresses. The
5765 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5766 leaves it alone. Component reversal etc. now happens only for IP addresses.
5769 34. The host_aton() function is supposed to be passed a string that is known
5770 to be a valid IP address. However, in the case of IPv6 addresses, it was
5771 not checking this. This is a hostage to fortune. Exim now panics and dies
5772 if the condition is not met. A case was found where this could be provoked
5773 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5774 components; fortuitously, this particular loophole had already been fixed
5775 by change 4.50/55 or 4.44/33 above.
5777 If there are any other similar loopholes, the new check in host_aton()
5778 itself should stop them being exploited. The report I received stated that
5779 data on the command line could provoke the exploit when Exim was running as
5780 exim, but did not say which command line option was involved. All I could
5781 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5782 running as the user.
5785 35. There was a buffer overflow vulnerability in the SPA authentication code
5786 (which came originally from the Samba project). I have added a test to the
5787 spa_base64_to_bits() function which I hope fixes it.
5790 36. The daemon start-up calls getloadavg() while still root for those OS that
5791 need the first call to be done as root, but it missed one case: when
5792 deliver_queue_load_max is set with deliver_drop_privilege. This is
5793 necessary for the benefit of the queue runner, because there is no re-exec
5794 when deliver_drop_privilege is set.
5796 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5797 was in use, was not putting the data itself into the right store pool;
5798 consequently, it could be overwritten for a subsequent message in the same
5799 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5802 38. Sometimes the final signoff response after QUIT could fail to get
5803 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5804 < 0 before doing a fflush(). This bug looks as though it goes back to the
5805 introduction of TLS in release 3.20, but "sometimes" must have been rare
5806 because the tests only now provoked it.
5812 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5813 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5814 Now it returns an integer. A number of calls were still expecting a T/F
5815 return. Fortuitously, in all cases, the tests worked in OK situations,
5816 which is the norm. However, things would have gone wrong on any write
5817 failures on the smtp file descriptor. This function is used when sending
5818 messages over SMTP and also when doing verify callouts.
5820 2. When Exim is called to do synchronous delivery of a locally submitted
5821 message (the -odf or -odi options), it no longer closes stderr before doing
5824 3. Implemented the mua_wrapper option.
5826 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5828 5. Implemented the functions header_remove(), header_testname(),
5829 header_add_at_position(), and receive_remove_recipient(), and exported them
5832 6. If an ACL "warn" statement specified the addition of headers, Exim already
5833 inserted X-ACL-Warn: at the start if there was no header name. However, it
5834 was not making this test for the second and subsequent header lines if
5835 there were newlines in the string. This meant that an invalid header could
5836 be inserted if Exim was badly configured.
5838 7. Allow an ACL "warn" statement to add header lines at the start or after all
5839 the Received: headers, as well as at the end.
5841 8. Added the rcpt_4xx retry error code.
5843 9. Added postmaster_mailfrom=xxx to callout verification option.
5845 10. Added mailfrom=xxxx to the callout verification option, for verify=
5848 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5849 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5851 12. Inserted some casts to stop certain compilers warning when using pointer
5852 differences as field lengths or precisions in printf-type calls (mostly
5853 affecting debugging statements).
5855 13. Added optional readline() support for -be (dynamically loaded).
5857 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5858 same clock tick as a message's arrival, so that its received time was the
5859 same as the "first fail" time on the retry record, and that message
5860 remained on the queue past the ultimate address timeout, every queue runner
5861 would try a delivery (because it was past the ultimate address timeout) but
5862 after another failure, the ultimate address timeout, which should have then
5863 bounced the address, did not kick in. This was a "< instead of <=" error;
5864 in most cases the first failure would have been in the next clock tick
5865 after the received time, and all would be well.
5867 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5868 being recognized when the domain list was tested by the match_domain
5869 condition in an expansion string.
5871 16. Added the ${str2b64: operator.
5873 17. Exim was always calling setrlimit() to set a large limit for the number of
5874 processes, without checking whether the existing limit was already
5875 adequate. (It did check for the limit on file descriptors.) Furthermore,
5876 errors from getrlimit() and setrlimit() were being ignored. Now they are
5877 logged to the main and panic logs, but Exim does carry on, to try to do its
5878 job under whatever limits there are.
5880 18. Imported PCRE 5.0.
5882 19. Trivial typo in log message " temporarily refused connection" (the leading
5885 20. If the log selector return_path_on_delivery was set and an address was
5886 redirected to /dev/null, the delivery process crashed because it assumed
5887 that a return path would always be set for a "successful" delivery. In this
5888 case, the whole delivery is bypassed as an optimization, and therefore no
5891 21. Internal re-arrangement: the function for sending a challenge and reading
5892 a response while authentication was assuming a zero-terminated challenge
5893 string. It's now changed to take a pointer and a length, to allow for
5894 binary data in such strings.
5896 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5898 23. Exim was not respecting finduser_retries when seeking the login of the
5899 uid under which it was called; it was always trying 10 times. (The default
5900 setting of finduser_retries is zero.) Also, it was sleeping after the final
5901 failure, which is pointless.
5903 24. Implemented tls_on_connect_ports.
5905 25. Implemented acl_smtp_predata.
5907 26. If the domain in control=submission is set empty, Exim assumes that the
5908 authenticated id is a complete email address when it generates From: or
5909 Sender: header lines.
5911 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5912 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5913 chown and chgrp in /bin and hostname in /usr/bin.
5915 28. Exim was keeping the "process log" file open after each use, just as it
5916 does for the main log. This opens the possibility of it remaining open for
5917 long periods when the USR1 signal hits a daemon. Occasional processlog
5918 errors were reported, that could have been caused by this. Anyway, it seems
5919 much more sensible not to leave this file open at all, so that is what now
5922 29. The long-running daemon process does not normally write to the log once it
5923 has entered its main loop, and it closes the log before doing so. This is
5924 so that log files can straightforwardly be renamed and moved. However,
5925 there are a couple of unusual error situations where the daemon does write
5926 log entries, and I had neglected to close the log afterwards.
5928 30. The text of an SMTP error response that was received during a remote
5929 delivery was being truncated at 512 bytes. This is too short for some of
5930 the long messages that one sometimes sees. I've increased the limit to
5933 31. It is now possible to make retry rules that apply only when a message has a
5934 specific sender, in particular, an empty sender.
5936 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5937 it possible to be selective about when SMTP synchronization is enforced.
5939 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5941 32. Implemented hosts_connection_nolog.
5943 33. Added an ACL for QUIT.
5945 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5948 35. Added mailbox_size and mailbox_filecount to appendfile.
5950 36. Added control = no_multiline_responses to ACLs.
5952 37. There was a bug in the logic of the code that waits for the clock to tick
5953 in the case where the clock went backwards by a substantial amount such
5954 that the microsecond fraction of "now" was more than the microsecond
5955 fraction of "then" (but the whole seconds number was less).
5957 38. Added support for the libradius Radius client library this is found on
5958 FreeBSD (previously only the radiusclient library was supported).
5964 1. When certain lookups returned multiple values in the form name=value, the
5965 quoting of the values was not always being done properly. Specifically:
5966 (a) If the value started with a double quote, but contained no whitespace,
5968 (b) If the value contained whitespace other than a space character (i.e.
5969 tabs or newlines or carriage returns) it was not quoted.
5970 This fix has been applied to the mysql and pgsql lookups by writing a
5971 separate quoting function and calling it from the lookup code. The fix
5972 should probably also be applied to nisplus, ibase and oracle lookups, but
5973 since I cannot test any of those, I have not disturbed their existing code.
5975 2. A hit in the callout cache for a specific address caused a log line with no
5976 reason for rejecting RCPT. Now it says "Previous (cached) callout
5977 verification failure".
5979 3. There was an off-by-one bug in the queryprogram router. An over-long
5980 return line was truncated at 256 instead of 255 characters, thereby
5981 overflowing its buffer with the terminating zero. As well as fixing this, I
5982 have increased the buffer size to 1024 (and made a note to document this).
5984 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5985 when Exim is waiting for an SMTP response from a remote server, Exim
5986 restarts its select() call on the socket, thereby resetting its timeout.
5987 This is not a problem when such interrupts are rare. Somebody set up a cron
5988 job to run exiwhat every 2 minutes, which is less than the normal select()
5989 timeout (5 or 10 minutes). This meant that the select() timeout never
5990 kicked in because it was always reset. I have fixed this by comparing the
5991 time when an interrupt arrives with the time at the start of the first call
5992 to select(). If more time than the timeout has elapsed, the interrupt is
5993 treated as a timeout.
5995 5. Some internal re-factoring in preparation for the addition of Sieve
5996 extensions (by MH). In particular, the "personal" test is moved to a
5997 separate function, and given an option for scanning Cc: and Bcc: (which is
5998 not set for Exim filters).
6000 6. When Exim created an email address using the login of the caller as the
6001 local part (e.g. when creating a From: or Sender: header line), it was not
6002 quoting the local part when it contained special characters such as @.
6004 7. Installed new OpenBSD configuration files.
6006 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6007 try to make them clearer.
6009 9. Callout options, other than the timeout value, were being ignored when
6010 verifying sender addresses in header lines. For example, when using
6012 verify = header_sender/callout=no_cache
6014 the cache was (incorrectly) being used.
6016 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6017 only the Cygwin environment.
6019 11. When return_path_on_delivery was set as a log selector, if different remote
6020 addresses in the same message used different return paths and parallel
6021 remote delivery occurred, the wrong values would sometimes be logged.
6022 (Whenever a remote delivery process finished, the return path value from
6023 the most recently started remote delivery process was logged.)
6025 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6026 lines when AUTH and/or TLS are in use. This is the "received protocol"
6027 field. Exim used to use "asmtp" for authenticated SMTP, without any
6028 indication (in the protocol name) for TLS use. Now it follows the RFC and
6029 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6030 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6031 names appear in log lines as well as in Received: header lines.
6033 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6034 extensions, and comparison tests, and to fix some bugs.
6036 14. Changes to the "personal" filter test:
6038 (1) The test was buggy in that it was just doing the equivalent of
6039 "contains" tests on header lines. For example, if a user's address was
6040 anne@some.where, the "personal" test would incorrectly be true for
6042 To: susanne@some.where
6044 This test is now done by extracting each address from the header in turn,
6045 and checking the entire address. Other tests that are part of "personal"
6046 are now done using regular expressions (for example, to check local parts
6047 of addresses in From: header lines).
6049 (2) The list of non-personal local parts in From: addresses has been
6050 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6051 taken from the Sieve specification recommendations.
6053 (3) If the message contains any header line starting with "List-" it is
6054 treated as non-personal.
6056 (4) The test for "circular" in the Subject: header line has been removed
6057 because it now seems ill-conceived.
6059 15. Minor typos in src/EDITME comments corrected.
6061 16. Installed latest exipick from John Jetmore.
6063 17. If headers_add on a router specified a text string that was too long for
6064 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6065 of string_sprintf() is now avoided.
6067 18. $message_body_size was not set (it was always zero) when running the DATA
6068 ACL and the local_scan() function.
6070 19. For the "mail" command in an Exim filter, no default was being set for
6071 the once_repeat time, causing a random time value to be used if "once" was
6072 specified. (If the value happened to be <= 0, no repeat happened.) The
6073 default is now 0s, meaning "never repeat". The "vacation" command was OK
6074 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6075 (I found it when inspecting the code).
6077 20. There is now an overall timeout for performing a callout verification. It
6078 defaults to 4 times the callout timeout, which applies to individual SMTP
6079 commands during the callout. The overall timeout applies when there is more
6080 than one host that can be tried. The timeout is checked before trying the
6081 next host. This prevents very long delays if there are a large number of
6082 hosts and all are timing out (e.g. when the network connections are timing
6083 out). The value of the overall timeout can be changed by specifying an
6084 additional sub-option for "callout", called "maxwait". For example:
6086 verify = sender/callout=5s,maxwait=20s
6088 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6089 to the end before writing, but this should make it even safer).
6091 22. Exim was forgetting that it had advertised PIPELINING for the second and
6092 subsequent messages on an SMTP connection. It was also not resetting its
6093 memory on STARTTLS and an internal HELO.
6095 23. When Exim logs an SMTP synchronization error within a session, it now
6096 records whether PIPELINING has been advertised or not.
6098 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6099 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6100 rather than long int.
6102 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6104 26. Added the never_mail option to autoreply.
6110 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6111 crash if the getsockname() call failed; this can happen if a connection is
6112 closed very soon after it is established. The problem was simply in the
6113 order in which certain operations were done, causing Exim to try to write
6114 to the SMTP stream before it had set up the file descriptor. The bug has
6115 been fixed by making things happen in the correct order.
6121 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6122 before the connection was closed, thus losing the rejection response.
6124 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6125 some early Solaris releases, but causes trouble in current releases where
6126 socklen_t is defined.
6128 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6131 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6134 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6135 that releasing the top of it at the end releases what was used for sub-
6136 expansions (unless the block got too big). However, discard this block if
6137 the first thing is a variable or header, so that we can use its block when
6138 it is dynamic (useful for very large $message_headers, for example).
6140 6. Lookups now cache *every* query, not just the most recent. A new, separate
6141 store pool is used for this. It can be recovered when all lookup caches are
6142 flushed. Lookups now release memory at the end of their result strings.
6143 This has involved some general refactoring of the lookup sources.
6145 7. Some code has been added to the store_xxx() functions to reduce the amount
6146 of flapping under certain conditions.
6148 8. log_incoming_interface used to affect only the <= reception log lines. Now
6149 it causes the local interface and port to be added to several more SMTP log
6150 lines, for example "SMTP connection from", and rejection lines.
6152 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6154 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6156 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6157 could be overwritten at the end of the current message (or the start of a
6158 new message if it was set in a HELO ACL). The value is now preserved for
6159 the duration of the SMTP connection.
6161 12. If a transport had a headers_rewrite setting, and a matching header line
6162 contained an unqualified address, that address was qualified, even if it
6163 did not match any rewriting rules. The underlying bug was that the values
6164 of the flags that permit the existence of unqualified sender and recipient
6165 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6166 non-local messages, and by -bnq for local messages) were not being
6167 preserved with the message after it was received.
6169 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6170 "next input=" as part of the text comprising the host identity instead of
6171 the correct text. The code was using the same buffer for two different
6172 strings. However, depending on which order the printing function evaluated
6173 its arguments, the bug did not always show up. Under Linux, for example, my
6174 test suite worked just fine.
6176 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6177 doesn't work with some older versions of Perl. It has been changed to "my",
6178 which in any case is probably the better facility to use.
6180 15. A really picky compiler found some instances of statements for creating
6181 error messages that either had too many or two few arguments for the format
6184 16. The size of the buffer for calls to the DNS resolver has been increased
6185 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6186 for addresses that have a lot of PTR records. This alleviates a problem; it
6187 does not fully solve it.
6189 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6190 buffer now truncates the list and logs the incident, which is the same
6191 action as happens when Exim is looking up a host name and its aliases.
6192 Previously in this situation something unpredictable would happen;
6193 sometimes it was "internal error: store_reset failed".
6195 18. If a server dropped the connection unexpectedly when an Exim client was
6196 using GnuTLS and trying to read a response, the client delivery process
6197 crashed while trying to generate an error log message.
6199 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6200 string, for example:
6202 warn message = H1: something\nH2: something
6204 the text was added as a single header line from Exim's point of view
6205 though it ended up OK in the delivered message. However, searching for the
6206 second and subsequent header lines using $h_h2: did not work. This has been
6207 fixed. Similarly, if a system filter added multiple headers in this way,
6208 the routers could not see them.
6210 20. Expanded the error message when iplsearch is called with an invalid key to
6211 suggest using net-iplsearch in a host list.
6213 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6214 ACLs are no longer actually imposed (and a message to that effect is
6217 22. If a "gecos" field in a passwd entry contained escaped characters, in
6218 particular, if it contained a \" sequence, Exim got it wrong when building
6219 a From: or a Sender: header from that name. A second bug also caused
6220 incorrect handling when an unquoted " was present following a character
6221 that needed quoting.
6223 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6224 was not being matched caselessly.
6226 24. Arranged for all hyphens in the exim.8 source to be escaped with
6229 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6230 properly. Recipient callout cache records were still being keyed to include
6231 the sender, even when use_sender was set false. This led to far more
6232 callouts that were necessary. The sender is no longer included in the key
6233 when use_sender is false.
6235 26. Added "control = submission" modifier to ACLs.
6237 27. Added the ${base62d: operator to decode base 62 numbers.
6239 28. dnsdb lookups can now access SRV records.
6241 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6242 the configuration file.
6244 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6245 (-v) mode. This makes the output for a verbose queue run more intelligible.
6247 31. Added a use_postmaster feature to recipient callouts.
6249 32. Added the $body_zerocount variable, containing the number of binary zero
6250 bytes in the message body.
6252 33. The time of last modification of the "new" subdirectory is now used as the
6253 "mailbox time last read" when there is a quota error for a maildir
6256 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6258 35. Added +ignore_unknown as a special item in host lists.
6260 36. Code for decoding IPv6 addresses in host lists is now included, even if
6261 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6262 address was recognized as an IP address, but was then not correctly decoded
6263 into binary, causing unexpected and incorrect effects when compared with
6270 1. Very minor rewording of debugging text in manualroute to say "list of
6271 hosts" instead of "hostlist".
6273 2. If verify=header_syntax was set, and a header line with an unqualified
6274 address (no domain) and a large number of spaces between the end of the
6275 name and the colon was received, the reception process suffered a buffer
6276 overflow, and (when I tested it) crashed. This was caused by some obsolete
6277 code that should have been removed. The fix is to remove it!
6279 3. When running in the test harness, delay a bit after writing a bounce
6280 message to get a bit more predictability in the log output.
6282 4. Added a call to search_tidyup() just before forking a reception process. In
6283 theory, someone could use a lookup in the expansion of smtp_accept_max_
6284 per_host which, without the tidyup, could leave open a database connection.
6286 5. Added the variables $recipient_data and $sender_data which get set from a
6287 lookup success in an ACL "recipients" or "senders" condition, or a router
6288 "senders" option, similar to $domain_data and $local_part_data.
6290 6. Moved the writing of debug_print from before to after the "senders" test
6293 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6294 problems for message scanning, either using a data ACL, or using
6295 local_scan() because the Received: header was not generated till after they
6296 were called (in order to set the time as the time of reception completion).
6297 I have revised the way this works. The header is now generated after the
6298 body is received, but before the ACL or local_scan() are called. After they
6299 are run, the timestamp in the header is updated.
6305 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6306 before starting a queue runner without re-exec. This happened only when
6307 deliver_drop_privilege was set or when the Exim user was set to root. The
6308 effect of the bug was that timeouts during subsequent deliveries caused
6309 crashes instead of being properly handled. The handler is now left at its
6310 default (and expected) setting.
6312 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6313 message, again when deliver_drop_privilege is set or Exim is run as root.
6314 The bug described in (1) was not present in this case, but the tidying up
6315 of the other signals was missing. I have made the two cases consistent.
6317 3. The ignore_target_hosts setting on a manualroute router was being ignored
6318 for hosts that were looked up using the /MX notation.
6320 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6323 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6324 operated on the sender address. After changing the $sender_address to <>
6325 for the sender address verify, Exim was re-instated it as the original
6326 (before rewriting) address, but remembering that it had rewritten it, so it
6327 wasn't rewriting it again. This bug also had the effect of breaking the
6328 sender address verification caching when the sender address was rewritten.
6330 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6331 This has been changed so that if the ip literal address matches
6332 ignore_target_hosts, the router declines.
6334 7. Added expansion conditions match_domain, match_address, and match_local_
6335 part (NOT match_host).
6337 8. The placeholder for the Received: header didn't have a length field set.
6339 9. Added code to Exim itself and to exim_lock to test for a specific race
6340 condition that could lead to file corruption when using MBX delivery. The
6341 issue is with the lockfile that is created in /tmp. If this file is removed
6342 after a process has opened it but before that process has acquired a lock,
6343 there is the potential for a second process to recreate the file and also
6344 acquire a lock. This could lead to two Exim processes writing to the file
6345 at the same time. The added code performs the same test as UW imapd; it
6346 checks after acquiring the lock that its file descriptor still refers to
6347 the same named file.
6349 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6350 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6353 11. Added the smtp_active_hostname option. If used, this will typically be made
6354 to depend on the incoming interface address. Because $interface_address is
6355 not set up until the daemon has forked a reception process, error responses
6356 that can happen earlier (such as "too many connections") no longer contain
6359 12. If an expansion in a condition on a "warn" statement fails because a lookup
6360 defers, the "warn" statement is abandoned, and the next ACL statement is
6361 processed. Previously this caused the whole ACL to be aborted.
6363 13. Added the iplsearch lookup type.
6365 14. Added ident_timeout as a log selector.
6367 15. Added tls_certificate_verified as a log selector.
6369 16. Added a global option tls_require_ciphers (compare the smtp transport
6370 option of the same name). This controls incoming TLS connections.
6372 17. I finally figured out how to make tls_require_ciphers do a similar thing
6373 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6374 before starting the TLS session.
6376 18. Tabs are now shown as \t in -bP output.
6378 19. If the log selector return_path_on_delivery was set, Exim crashed when
6379 bouncing a message because it had too many Received: header lines.
6381 20. If two routers both had headers_remove settings, and the first one included
6382 a superfluous trailing colon, the final name in the first list and the
6383 first name in the second list were incorrectly joined into one item (with a
6384 colon in the middle).
6390 1. Added -C and -D options to the exinext utility, mainly to make it easier
6391 to include in the automated testing, but these could be helpful when
6392 multiple configurations are in use.
6394 2. The exinext utility was not formatting the output nicely when there was
6395 an alternate port involved in the retry record key, nor when there was a
6396 message id as well (for retries that were specific to a specific message
6397 and a specific host). It was also confused by IPv6 addresses, because of
6398 the additional colons they contain. I have fixed the IPv4 problem, and
6399 patched it up to do a reasonable job for IPv6.
6401 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6402 delivery, the log line now contains "pipelined" if PIPELINING was used.
6404 4. An SMTP transport process used to panic and die if the bind() call to set
6405 an explicit outgoing interface failed. This has been changed; it is now
6406 treated in the same way as a connect() failure.
6408 5. A reference to $sender_host_name in the part of a conditional expansion
6409 that was being skipped was still causing a DNS lookup. This no longer
6412 6. The def: expansion condition was not recognizing references to header lines
6413 that used bh_ and bheader_.
6415 7. Added the _cache feature to named lists.
6417 8. The code for checking quota_filecount in the appendfile transport was
6418 allowing one more file than it should have been.
6420 9. For compatibility with Sendmail, the command line option
6428 and sets the incoming protocol and host name (for trusted callers). The
6429 host name and its colon can be omitted when only the protocol is to be set.
6430 Note the Exim already has two private options, -pd and -ps, that refer to
6431 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6432 "s", but I don't think that's a major issue.
6434 10. A number of refactoring changes to the code, none of which should affect
6437 (a) The number of logging options was getting close to filling up the
6438 32-bit word that was used as a bit map. I have split them into two classes:
6439 those that are passed in the argument to log_write(), and those that are
6440 only ever tested independently outside of that function. These are now in
6441 separate 32-bit words, so there is plenty of room for expansion again.
6442 There is no change in the user interface or the logging behaviour.
6444 (b) When building, for example, log lines, the code previously used a
6445 macro that called string_cat() twice, in order to add two strings. This is
6446 not really sufficiently general. Furthermore, there was one instance where
6447 it was actually wrong because one of the argument was used twice, and in
6448 one call a function was used. (As it happened, calling the function twice
6449 did not affect the overall behaviour.) The macro has been replaced by a
6450 function that can join an arbitrary number of extra strings onto a growing
6453 (c) The code for expansion conditions now uses a table and a binary chop
6454 instead of a serial search (which was left over from when there were very
6455 few conditions). Also, it now recognizes conditions like "pam" even when
6456 the relevant support is not compiled in: a suitably worded error message is
6457 given if an attempt is made to use such a condition.
6459 11. Added ${time_interval:xxxxx}.
6461 12. A bug was causing one of the ddress fields not to be passed back correctly
6462 from remote delivery subprocesses. The field in question was not being
6463 subsequently used, so this caused to problems in practice.
6465 13. Added new log selectors queue_time and deliver_time.
6467 14. Might have fixed a bug in maildirsizefile handling that threw up
6468 "unexpected character" debug warnings, and recalculated the data
6469 unnecessarily. In any case, I expanded the warning message to give more
6472 15. Added the message "Restricted characters in address" to the statements in
6473 the default ACL that block characters like @ and % in local parts.
6475 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6476 Three changes have been made:
6478 (a) There was a serious bug; a negative response to MAIL caused the whole
6479 recipient domain to be cached as invalid, thereby blocking all messages
6480 to all local parts at the same domain, from all senders. This bug has
6481 been fixed. The domain is no longer cached after a negative response to
6482 MAIL if the sender used is not empty.
6484 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6487 (c) A new callout option, "use_sender" has been added for people who want
6488 the modified behaviour.
6494 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6497 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6498 indeed breaks things for older releases.
6500 3. Added additional logging to the case where there is a problem reading data
6501 from a filter that is running in a subprocess using a pipe, in order to
6502 try to track down a specific problem.
6504 4. Testing facility fudge: when running in the test harness and attempting
6505 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6506 getting "No route to host". Convert this to a timeout.
6508 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6511 6. Some OS don't have socklen_t but use size_t instead. This affects the
6512 fifth argument of getsockopt() amongst other things. This is now
6513 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6514 can be set for individual OS. I have set it for SunOS5, OSF1, and
6515 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6516 some earlier ones do not.
6518 7. Change 4.30/15 was not doing the test caselessly.
6520 8. The standard form for an IPv6 address literal was being rejected by address
6521 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6522 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6523 this, as well as the form without the "IPv6" on the front (but only when
6524 address literals are enabled, of course).
6526 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6528 10. Exim crashed if a message with an empty sender address specified by -f
6529 encountered a router with an errors_to setting. This could be provoked only
6530 by a command such as
6534 where an empty string was supplied; "<>" did not hit this bug.
6536 11. Installed PCRE release 4.5.
6538 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6539 remained set. It is now erased.
6541 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6542 times from message ids (which are base 36 rather than the normal 62).
6544 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6545 were being counted as actual protocol errors, and logged if the log
6546 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6547 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6548 and DATA following a set of rejected RCPTs do not count as protocol errors.
6549 In other words, Exim assumes they were pipelined, though this may not
6550 actually be the case. Of course, in all cases the client gets an
6551 appropriate error code.
6553 15. If a lookup fails in an ACL condition, a message about the failure may
6554 be available; it is used if testing the ACL cannot continue, because most
6555 such messages specify what the cause of the deferral is. However, some
6556 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6557 that caused an old message to be retained and used if a later statement
6558 caused a defer, replacing the real cause of the deferral.
6560 16. If an IP address had so many PTR records that the DNS lookup buffer
6561 was not large enough to hold them, Exim could crash while trying to process
6562 the truncated data. It now detects and logs this case.
6564 17. Further to 4.21/58, another change has been made: if (and only if) the
6565 first line of a message (the first header line) ends with CRLF, a bare LF
6566 in a subsequent header line has a space inserted after it, so as not to
6567 terminate the header.
6569 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6570 unnecessarily, used atoi() instead of strtol(), and didn't check the
6571 termination when getting file sizes from file names by regex.
6573 19. Completely re-implemented the support for maildirsize files, in the light
6574 of a number of problems with the previous contributed implementation
6575 (4.30/29). In particular:
6577 . If the quota is zero, the maildirsize file is maintained, but no quota is
6580 . If the maildir directory does not exist, it is created before any attempt
6581 to write a maildirsize file.
6583 . The quota value in the file is just a cache; if the quota is changed in
6584 the transport, the new value overrides.
6586 . A regular expression is available for excluding directories from the
6589 20. The autoreply transport checks the characters in options that define the
6590 message's headers; it allows continued headers, but it was checking with
6591 isspace() after an embedded newline instead of explicitly looking for a
6594 21. If all the "regular" hosts to which an address was routed had passed their
6595 expiry times, and had not reached their retry times, the address was
6596 bounced, even if fallback hosts were defined. Now Exim should go on to try
6599 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6600 equivalent code in the SMTP transport. Some hosts send humungous responses
6601 to HELO/EHLO, more than 1024 it seems.
6603 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6604 gives compiler warnings in some environments. I've now done it "properly",
6607 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6608 (because of problems with the built-in one) was declared to return uschar *
6609 instead of char *, causing compiler failure.
6611 25. Fixed a file descriptor leak when processing alias/forward files.
6613 26. Fixed a minor format string issue in dbfn.c.
6615 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6617 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6618 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6621 29. When a sender address is verified, it is cached, to save repeating the test
6622 when there is more than one recipient in a message. However, when the
6623 verification involves a callout, it is possible for different callout
6624 options to be set for different recipients. It is too complicated to keep
6625 track of this in the cache, so now Exim always runs a verification when a
6626 callout is required, relying on the callout cache for the optimization.
6627 The overhead is duplication of the address routing, but this should not be
6630 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6631 to be verified with callout=postmaster, and the main callout worked but the
6632 postmaster check failed, the verification correctly failed. However, if a
6633 subsequent RCPT command asked for sender verification *without* the
6634 postmaster check, incorrect caching caused this verification also to fail,
6637 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6638 it was not caching the DNS options (qualify_single, search_parents) that
6639 were used when the lookup failed. A subsequent lookup with different
6640 options therefore always gave the same answer, though there were cases
6641 where it should not have. (Example: a "domains = !$mx_any" option on a
6642 dnslookup router: the "domains" option is always processed without any
6643 widening, but the router might have qualify_single set.) Now Exim uses the
6644 cached value only when the same options are set.
6646 32. Added John Jetmore's "exipick" utility to the distribution.
6648 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6649 than a timeout (e.g. a certificate is required, and is not provided), an
6650 Exim server now closes the connection immediately. Previously it waited for
6651 the client to close - but if the client is SSL, it seems that they each
6652 wait for each other, leading to a delay before one of them times out.
6654 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6655 maintained 0.8.x compatibility because I don't think many are using it, and
6656 it is clearly obsolete.
6658 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6661 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6662 client certificate was expired. A simple patch fixes this, though I don't
6663 understand the full logic of why the verify callback is called multiple
6666 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6667 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6668 which causes problems with some clients (such as the Certicom SSL Plus
6669 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6670 disables the coutermeasure allowing Eudora to connect."
6672 38. Exim was not checking that a write() to a log file succeeded. This could
6673 lead to Bad Things if a log got too big, in particular if it hit a file
6674 size limit. Exim now panics and dies if it cannot write to a log file, just
6675 as it does if it cannot open a log file.
6677 39. Modified OS/Makefile-Linux so that it now contains
6679 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6681 The two -D definitions ensure that Exim is compiled with large file
6682 support, which makes it possible to handle log files that are bigger than
6685 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6686 instance) a domain was checked against a named list that involved a lookup,
6687 causing $domain_data to be set, then another domain was checked against the
6688 same list, then the first domain was re-checked, the value of $domain_data
6689 after the final check could be wrong. In particular, if the second check
6690 failed, it could be set empty. This bug probably also applied to
6693 41. The strip_trailing_dot option was not being applied to the address given
6694 with the -f command-line option.
6696 42. The code for reading a message's header from the spool was incrementing
6697 $received_count, but never initializing it. This meant that the value was
6698 incorrect (doubled) while delivering a message in the same process in which
6699 it was received. In the most common configuration of Exim, this never
6700 happens - a fresh exec is done - but it can happen when
6701 deliver_drop_privilege is set.
6703 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6704 it now includes up to 150 characters of the unexpected data in the log
6707 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6708 and building data strings. The size of both of these buffers was 10 000
6709 bytes - far larger than anybody would *ever* want, thought I. Needless to
6710 say, somebody hit the limit. I have increased the maximum line length to
6711 20 000 and the maximum data length of concatenated lines to 100 000. I have
6712 also fixed two bugs, because there was no checking on these buffers. Tsk,
6713 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6714 buffer is too small.
6716 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6717 lsearch lookups. Now it does.
6719 46. When parsing a route_list item in a manualroute router, a fixed-length
6720 buffer was used for the list of hosts. I made this 1024 bytes long,
6721 thinking that nobody would ever have a list of hosts that long. Wrong.
6722 Somebody had a whole pile of complicated expansion conditions, and the
6723 string was silently truncated, leading to an expansion error. It turns out
6724 that it is easier to change to an unlimited length (owing to other changes
6725 that have happened since this code was originally written) than to build
6726 structure for giving a limitation error. The length of the item that
6727 expands into the list of hosts is now unlimited.
6729 47. The lsearch lookup could not handle data where the length of text line was
6730 more than 4095 characters. Such lines were truncated, leading to shortened
6731 data being returned. It should now handle lines of any length.
6733 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6734 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6737 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6738 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6739 no longer happens in comment lines. A list of replacements is now placed
6740 at the head of all of the source files, except those whose only change is
6741 to replace PERL_COMMAND in the very first #! line.
6743 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6744 messages on the queue, with a bottom-up merge sort, using code contributed
6745 by Michael Haardt. This should make operations like -bp somewhat faster on
6746 large queues. It won't affect queue runners, except when queue_run_in_order
6749 51. Installed eximstats 1.31 in the distribution.
6751 52. Added support for SRV lookups to the dnslookup router.
6753 53. If an ACL referred to $message_body or $message_body_end, the value was not
6754 reset for any messages that followed in the same SMTP session.
6756 54. The store-handling optimization for building very long strings was not
6757 differentiating between the different store pools. I don't think this
6758 actually made any difference in practice, but I've tidied it.
6760 55. While running the routers to verify a sender address, $sender_address
6761 was still set to the sender address. This is wrong, because when routing to
6762 send a bounce to the sender, it would be empty. Therefore, I have changed
6763 it so that, while verifying a sender address, $sender_address is set to <>.
6764 (There is no change to what happens when verifying a recipient address.)
6766 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6767 target A or AAAA records (if not already returned) without resetting the
6768 qualify_single or search_parents options of the DNS resolver. These are
6769 inappropriate in this case because the targets of MX and SRV records must
6770 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6771 target that, when qualified, matched something in the local domain. These
6772 two options are now turned off when doing these lookups.
6774 57. It seems that at least some releases of Reiserfs (which does not have the
6775 concept of a fixed number of inodes) returns zero and not -1 for the
6776 number of available inodes. This interacted badly with check_spool_inodes,
6777 which assumed that -1 was the "no such thing" setting. What I have done is
6778 to check that the total number of inodes is greater than zero before doing
6779 the test of how many are available.
6781 58. When a "warn" ACL statement has a log_message modifier, the message is
6782 remembered, and not repeated. This is to avoid a lot of repetition when a
6783 message has many recipients that cause the same warning to be written.
6784 However, Exim was preserving the list of already written lines for an
6785 entire SMTP session, which doesn't seem right. The memory is now reset if a
6786 new message is started.
6788 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6789 debugging output unless log_rewrite was also set.
6791 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6792 of (int)(handle) when we know that handle contains (void *)(-1).
6794 61. The Exim daemon panic-logs an error return when it closes the incoming
6795 connection. However "connection reset by peer" seems to be common, and
6796 isn't really an error worthy of noting specially, so that particular error
6799 62. When Exim is trying to find all the local interfaces, it used to panic and
6800 die if the ioctl to get the interface flags failed. However, it seems that
6801 on at least one OS (Solaris 9) it is possible to have an interface that is
6802 included in the list of interfaces, but for which you get a failure error
6803 for this call. This happens when the interface is not "plumbed" into a
6804 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6805 failure of the "get flags" call assumes that the interface is down.
6807 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6808 makes life easier for people who are doing arithmetic on fields extracted
6809 from dates, where you often get leading zeros that should not be
6810 interpreted as octal.
6812 64. Added qualify_domain to the redirect router, to override the global
6815 65. If a pathologically long header line contained very many addresses (the
6816 report of this problem mentioned 10 000) and each of them was rewritten,
6817 Exim could use up a very large amount of memory. (It kept on making new
6818 copies of the header line as it rewrote, and never released the old ones.)
6819 At the expense of a bit more processing, the header rewriting function has
6820 been changed so that it no longer eats memory in this way.
6822 66. The generation of the Received: header has been moved from the time that a
6823 message starts to be received, to the time that it finishes. The timestamp
6824 in the Received: header should now be very close to that of the <= log
6825 line. There are two side-effects of this change:
6827 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6828 logged header lines no longer include the local Received: line, because
6829 it has not yet been created. The same applies to a copy of the message
6830 that is returned to a non-SMTP sender when a message is rejected.
6832 (b) When a filter file is tested using -bf, no additional Received: header
6833 is added to the test message. After some thought, I decided that this
6836 This change does not affect the value of $received_for. It is still set
6837 after address rewriting, but before local_scan() is called.
6839 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6841 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6842 gave an unhelpful panic error message, and a defer error. I have managed to
6843 change this behaviour so that it now rejects any supplied certificate,
6844 which seems right, as the list of acceptable certificates is empty.
6846 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6847 gave an unhelpful defer error. I have not managed to make this reject any
6848 supplied certificates, but the error message it gives is "no certificate
6849 supplied", which is not helpful.
6851 70. exigrep's output now also includes lines that are not associated with any
6852 message, but which match the given pattern. Implemented by a patch from
6853 Martin Sluka, which also tidied up the Perl a bit.
6855 71. Recipient callout verification, like sender verification, was using <> in
6856 the MAIL FROM command. This isn't really the right thing, since the actual
6857 sender may affect whether the remote host accepts the recipient or not. I
6858 have changed it to use the actual sender in the callout; this means that
6859 the cache record is now keyed on a recipient/sender pair, not just the
6860 recipient address. There doesn't seem to be a real danger of callout loops,
6861 since a callout by the remote host to check the sender would use <>.
6862 [SEE ABOVE: changed after hitting problems.]
6864 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6865 temporary errors. However, in the case of such a code being given after
6866 the end of a data transmission (i.e. after ".") Exim was failing to write
6867 a retry record for the message. (Yes, there was some broken host that was
6868 actually sending 8xx at this point.)
6870 73. An unknown lookup type in a host list could cause Exim to panic-die when
6871 the list was checked. (An example that provoked this was putting <; in the
6872 middle of a list instead of at the start.) If this happened during a DATA
6873 ACL check, a -D file could be left lying around. This kind of configuration
6874 error no longer causes Exim to die; instead it causes a defer error. The
6875 incident is still logged to the main and panic logs.
6877 74. Buglet left over from Exim 3 conversion. The message "too many messages
6878 in one connection" was written to the rejectlog but not the mainlog, except
6879 when address rewriting (yes!) was being logged.
6881 75. Added write_rejectlog option.
6883 76. When a system filter was run not as root (that is, when system_filter_user
6884 was set), the values of the $n variables were not being returned to the
6885 main process; thus, they were not subsequently available in the $sn
6888 77. Added +return_path_on_delivery log selector.
6890 78. A connection timeout was being treated differently from recipients deferred
6891 when testing hosts_max_try with a message that was older than the host's
6892 retry timeout. (The host should not be counted, thus allowing all hosts to
6893 be tried at least once before bouncing.) This may have been the cause of an
6894 occasionally reported bug whereby a message would remain on the queue
6895 longer than the retry timeout, but would be bounced if a delivery was
6896 forced. I say "may" because I never totally pinned down the problem;
6897 setting up timeout/retry tests is difficult. See also the next item.
6899 79. The ultimate address timeout was not being applied to errors that involved
6900 a combination of host plus message (for example, a timeout on a MAIL
6901 command). When an address resolved to a number of possible hosts, and they
6902 were not all tried for each delivery (e.g. because of hosts_max_try), a
6903 message could remain on the queue longer than the retry timeout.
6905 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6908 81. Fixed an obscure SMTP outgoing bug which required at least the following
6909 conditions: (a) there was another message waiting for the same server;
6910 (b) the server returned 5xx to all RCPT commands in the first message so
6911 that the message was not completed; (c) the server dropped the connection
6912 or gave a negative response to the RSET that Exim sends to abort the
6913 transaction. The observed case was a dropped connection after DATA that had
6914 been sent in pipelining mode. That is, the server had advertised PIPELINING
6915 but was not implementing it correctly. The effect of the bug was incorrect
6916 behaviour, such as trying another host, and this could lead to a crash.
6922 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6923 and daemon.c were passed as pointers to ints; they should have been
6924 pointers to socklen_t variables (which are typically unsigned ints).
6926 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6929 3. Fixed a really odd bug that affected only the testing scheme; patching a
6930 certain fixed string in the binary changed the value of another string that
6931 happened to be identical to the end of the original first string.
6933 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6934 name", it returns that address as the IP address. On some operating
6935 systems (e.g. Solaris), it also passes back the IP address string as the
6936 "host name". However, on others (e.g. Linux), it passes back an empty
6937 string. Exim wasn't checking for this, and was changing the host name to an
6938 empty string, assuming it had been canonicalized.
6940 5. Although rare, it is permitted to have more than one PTR record for a given
6941 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6942 all the names associated with an address, because they do in Solaris.
6943 However, it seems that they do not in Linux for data that comes from the
6944 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6945 I found this out when I moved to a new Linux workstation and tried to run
6946 the Exim test suite.
6948 To get round this problem I have changed the code so that it now does its
6949 own call to the DNS to look up PTR records when searching for a host name.
6950 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6951 addresses that are only in /etc/hosts are still found.
6953 This behaviour is, however, controlled by an option called host_lookup_
6954 order, which defaults to "bydns:byaddr". If people want to use the other
6955 order, or indeed, just use one or the other means of lookup, they can
6956 specify it in this variable.
6958 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6959 some operating systems, this comes back from gethostbyaddr() as an empty
6960 string, and this is what Exim used to test for. However, it seems that in
6961 other systems, "." is yielded. Exim now tests for this case too.
6963 7. The values of check_spool_space and check_log_space are now held internally
6964 as a number of kilobytes instead of an absolute number of bytes. If a
6965 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6966 kilobyte. This means that much larger values can be stored.
6968 8. Exim monitor: an attempt to get the action menu when not actually pointing
6969 at a message produces an empty menu entitled "No message selected". This
6970 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6971 no entries in it ("Shell widget menu has zero width and/or height"). So I
6972 have added a single, blank menu entry in this case.
6974 9. Added ${quote_local_part.
6976 10. MIME decoding is now applied to the contents of Subject: header lines when
6979 11. Now that a reference to $sender_host_address automatically causes a reverse
6980 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6981 host lookup before query-style lookups in lists that might use this
6982 variable. This has therefore been abolished, and the "net-" prefix is no
6983 longer necessary for query-style lookups.
6985 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6986 have been a typo for LFLAGS, so it has been changed.
6988 13. The install script calls Exim with "-C /dev/null" in order to find the
6989 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6990 to be output. However, since Exim outputs its version number before the
6991 error, it didn't break the script. It just looked ugly. I fixed this by
6992 always allowing "-C /dev/null" if the caller is root.
6994 14. Ignore overlarge ACL variable number when reading spool file - insurance
6995 against a later release with more variables having written the file.
6997 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6998 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6999 this, as well as the form without the "IPv6" on the front.
7001 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7002 OS/Makefile-Darwin file.
7004 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7005 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7007 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7008 variable. However, some OS use ints for inodes. I've added cast to long int
7009 to get rid of the compiler warning.
7011 19. I had forgotten to lock out "/../" in configuration file names when
7012 ALT_CONFIG_PREFIX was set.
7014 20. Routers used for verification do not need to specify transports. However,
7015 if such a router generated a host list, and callout was configured, Exim
7016 crashed, because it could not find a port number from the (non-existent)
7017 transport. It now assumes port 25 in this circumstance.
7019 21. Added the -t option to exigrep.
7021 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7022 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7023 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7024 them is set, LOOKUP_LSEARCH is forced.
7026 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7028 24. Added sender and host information to the "rejected by local_scan()" log
7029 line; previously there was no indication of these.
7031 25. Added .include_if_exists.
7033 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7034 Linux. It turns out that not all file systems support this. Apparently some
7035 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7036 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7037 not supported on the file descriptor, is now ignored when Exim is trying to
7038 sync a directory. This applies only to Linux.
7040 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7042 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7043 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7046 29. Added support for maildirsize files from supplied patch (modified a bit).
7048 30. The use of :fail: followed by an empty string could lead Exim to respond to
7049 sender verification failures with (e.g.):
7051 550 Verification failed for <xxx>
7052 550 Sender verify failed
7054 where the first response line was missing the '-' that indicates it is not
7055 the final line of the response.
7057 31. The loop for finding the name of the user that called Exim had a hardwired
7058 limit of 10; it now uses the value of finduser_retries, which is used for
7059 all other user lookups.
7061 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7064 33. Exim was neglecting to zero errno before one call of strtol() when
7065 expanding a string and expecting an integer value. On some systems this
7066 resulted in spurious "integer overflow" errors. Also, it was casting the
7067 result into an int without checking.
7069 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7070 did not work. The code looks as if it has *never* worked, though it appears
7071 to have been documented since at least release 1.62. I have made it work.
7073 35. The "timeout_DNS" error in retry rules, also documented since at least
7074 1.62, also never worked. As it isn't clear exactly what this means, and
7075 clearly it isn't a major issue, I have abolished the feature by treating it
7076 as "timeout", and writing a warning to the main and panic logs.
7078 36. The display of retry rules for -brt wasn't always showing the error code
7081 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7082 timeout_connect_A, timeout_connect_MX.
7084 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7085 to the empty sender.
7087 39. The daemon was not analysing the content of -oX till after it had closed
7088 stderr and disconnected from the controlling terminal. This meant that any
7089 syntax errors were only noted on the panic log, and the return code from
7090 the command was 0. By re-arranging the code a little, I've made the
7091 decoding happen first, so such errors now appear on stderr, and the return
7092 code is 1. However, the actual setting up of the sockets still happens in
7093 the disconnected process, so errors there are still only recorded on the
7096 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7097 connections (as happens on some IP stacks) was logged at start up time as
7098 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7099 it from "IPv6 and IPv4", which means that two separate sockets are being
7102 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7103 says whether AF_INET or AF_INET6 was passed as an argument.
7105 42. Exiwhat output was messed up when time zones were included in log
7108 43. Exiwhat now gives more information about the daemon's listening ports,
7109 and whether -tls-on-connect was used.
7111 44. The "port" option of the smtp transport is now expanded.
7113 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7114 silently ignored. Now an error message is written to the main and panic
7117 46. There's a new ACL modifier called "logwrite" which writes to a log file
7118 as soon as it is encountered.
7120 47. Added $local_user_uid and $local_user_gid at routing time.
7122 48. Exim crashed when trying to verify a sender address that was being
7125 49. Exim was recognizing only a space character after ".include". It now also
7126 recognizes a tab character.
7128 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7129 extracting the relevant information from the specification. The man page no
7130 longer contains scrambled data for the -d option, and I've added a section
7131 at the front about calling Exim under different names.
7133 51. Added "extra_headers" argument to the "mail" command in filter files.
7135 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7138 53. Installed eximstats 1.29.
7140 54. Added transport_filter_timeout as a generic transport option.
7142 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7143 Cc: header lines. This was required by RFC 822, but it not required by RFC
7146 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7147 incoming messages that did not have them. Now it does so only if the
7148 message originates locally, that is, if there is no associated remote host
7149 address. When Resent- header lines are present, this applies to the Resent-
7150 lines rather than the non-Resent- lines.
7152 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7153 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7155 58. Messages for configuration errors now include the name of the main
7156 configuration file - useful now that there may be more than one file in a
7157 list (.included file names were always shown).
7159 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7160 for those rare installations that do not start the daemon as root or run it
7161 setuid root. I've cut out the call to initgroups() if the daemon is not
7164 60. The Exim user and group can now be bound into the binary as text strings
7165 that are looked up at the start of Exim's processing.
7167 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7169 62. Added $mailstore_basename variable.
7171 63. Installed patch to sieve.c from Michael Haardt.
7173 64. When Exim failed to open the panic log after failing to open the main log,
7174 the original message it was trying to log was written to stderr and debug
7175 output, but if they were not available (the usual case in production), it
7176 was lost. Now it is written to syslog before the two lines that record the
7177 failures to open the logs.
7179 65. Users' Exim filters run in subprocesses under the user's uid. It is
7180 possible for a "deliver" command or an alias in a "personal" command to
7181 provoke an address rewrite. If logging of address rewriting is configured,
7182 this fails because the process is not running as root or exim. There may be
7183 a better way of dealing with this, but for the moment (because 4.30 needs
7184 to be released), I have disabled address rewrite logging when running a
7185 filter in a non-root, non-exim process.
7191 1. The buildconfig auxiliary program wasn't quoting the value set for
7192 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7193 not defined. This bug was masked in 4.22 by the effect that was fixed in
7196 2. Some messages that were rejected after a message id was allocated were
7197 shown as "incomplete" by exigrep. It no longer does this for messages that
7198 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7200 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7201 have allow_domain_literals set, the ID did not get logged in the <= line.
7202 Domain literals are now always recognized in Message-ID: header lines.
7204 4. The first argument for a ${extract expansion item is the key name or field
7205 number. Leading and trailing spaces in this item were not being ignored,
7206 causing some misleading effects.
7208 5. When deliver_drop_privilege was set, single queue runner processes started
7209 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7210 same command in the process it spins off) were not dropping privilege.
7212 6. When the daemon running as "exim" started a queue runner, it always
7213 re-executed Exim in the spun-off process. This is a waste of effort when
7214 deliver_drop_privilege is set. The new process now just calls the
7215 queue-runner function directly.
7221 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7224 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7225 ignored. Though the use of -oP was forcing the writing of a pid file, it
7226 was always written to the default place.
7228 3. If the message "no IP address found for host xxxx" is generated during
7229 incoming verification, it is now followed by identification of the incoming
7230 connection (so you can more easily find what provoked it).
7232 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7234 5. Added some features to "harden" Exim a bit more against certain attacks:
7236 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7237 be put in Local/Makefile. This is like the never_users runtime option,
7238 but it cannot be overridden. The default setting is "root".
7240 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7241 prefix string with which any file named in a -C command line option
7244 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7245 is retained for -C and -D only if the caller of Exim is root. Without
7246 it, the exim user may also use -C and -D and retain privilege.
7248 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7249 command line option is disabled.
7251 6. Macro names set by the -D option must start with an upper case letter, just
7252 like macro names defined in the configuration file.
7254 7. Added "dereference=" facility to LDAP.
7256 8. Two instances of the typo "uknown" in the source files are fixed.
7258 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7259 the Configure-Makefile script screwed up while processing it.
7261 10. Incorporated PCRE 4.4.
7263 11. The SMTP synchronization check was not operating right at the start of an
7264 SMTP session. For example, it could not catch a HELO sent before the client
7265 waited for the greeting. There is now a check for outstanding input at the
7266 point when the greeting is written. Because of the duplex, asynchronous
7267 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7268 way, but not yet received, when the check is performed.
7270 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7271 on TCP/IP sockets, because this apparently causes some broken clients to
7274 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7275 unchanged) from the Cygwin maintainer.
7277 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7278 was supported instead of testing for mbx. Effectively a typo.
7280 15. The spa authenticator server code was not checking that the input it
7281 received was valid base64.
7283 16. The debug output line for the "set" modifier in ACLs was not showing the
7284 name of the variable that was being set.
7286 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7288 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7289 lookup on its own. Something else was needed to trigger the lookup. For
7290 example, a match in host_lookup or the need for a host name in a host list.
7291 Now, if $sender_host_name is referenced and the host name has not yet been
7292 looked up, a lookup is performed. If the lookup fails, the variable remains
7293 empty, and $host_lookup_failed is set to "1".
7295 19. Added "eqi" as a case-independent comparison operator.
7297 20. The saslauthd authentication condition could segfault if neither service
7298 nor realm was specified.
7300 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7301 error message that was logged was misleading, and incoming SMTP
7302 connections were dropped. The message is now more accurate, and temporary
7303 errors are given to SMTP connections.
7305 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7306 (except RSET) with a 421 error, until QUIT is received. However, it was
7307 failing to send a response to QUIT.
7309 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7310 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7311 ACL is now run after the helo_try_verify_hosts code.
7313 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7314 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7315 case-independent, so other case variants are also recognized). Apparently
7316 some systems use these upper case variants.
7318 25. If more than two messages were waiting for the same host, and a transport
7319 filter was specified for the transport, Exim sent two messages over the
7320 same TCP/IP connection, and then failed with "socket operation on non-
7321 socket" when it tried to send the third.
7323 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7325 27. The extern definition of crypt16() in expand.c was not being excluded when
7326 the OS had its own crypt16() function.
7328 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7329 as a preferred synonym for return_size_limit, both as an option and as an
7332 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7334 30. Changed the default configuration ACL to relax the local part checking rule
7335 for addresses that are not in any local domains. For these addresses,
7336 slashes and pipe symbols are allowed within local parts, but the sequence
7337 /../ is explicitly forbidden.
7339 31. SPA server authentication was not clearing the challenge buffer before
7342 32. log_message in a "warn" ACL statement was writing to the reject log as
7343 well as to the main log, which contradicts the documentation and doesn't
7344 seem right (because no rejection is happening). So I have stopped it.
7346 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7347 However, I am unable to do any testing of this.
7349 34. Fixed an infelicity in the appendfile transport. When checking directories
7350 for a mailbox, to see if any needed to be created, it was accidentally
7351 using path names with one or more superfluous leading slashes; tracing
7352 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7354 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7355 contents are added to the log line that is written for every discarded
7356 recipient. (Previously a log_message setting was ignored.)
7358 36. The ${quote: operator now quotes the string if it is empty.
7360 37. The install script runs exim in order to find its version number. If for
7361 some reason other than non-existence or emptiness, which it checks, it
7362 could not run './exim', it was installing it with an empty version number,
7363 i.e. as "exim-". This error state is now caught, and the installation is
7366 38. An argument was missing from the function that creates an error message
7367 when Exim fails to connect to the socket for saslauthd authentication.
7368 This could cause Exim to crash, or give a corrupted message.
7370 39. Added isip, isip4, and isip6 to ${if conditions.
7372 40. The ACL variables $acl_xx are now saved with the message, and can be
7373 accessed later in routers, transports, and filters.
7375 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7376 strings in the file are not string-expanded.
7378 42. If a MAIL command specified a SIZE value that was too large to fit into an
7379 int variable, the check against message_size_limit failed. Such values are
7380 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7381 day this will have to be increased, but I don't think I want to be around
7382 when emails are that large.
7389 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7390 iconv() is not standard in FreeBSD.
7392 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7393 IPv6 enabled. The observed symptom was a segmentation fault on return from
7394 the function os_common_find_running_interfaces() in src/os.c.
7396 3. In the check_special_case() function in daemon.c I had used "errno" as an
7397 argument name, which causes warnings on some systems. This was basically a
7398 typo, since it was named "eno" in the comments!
7400 4. The code that waits for the clock to tick (at a resolution of some fraction
7401 of a second) so as to ensure message-id uniqueness was always waiting for
7402 at least one whole tick, when it could have waited for less. [This is
7403 almost certainly not relevant at current processor speeds, where it is
7404 unlikely to ever wait at all. But we try to future-proof.]
7406 5. The function that sleeps for a time interval that includes fractions of a
7407 second contained a race. It did not block SIGALRM between setting the
7408 timer, and suspending (a couple of lines later). If the interval was short
7409 and the sigsuspend() was delayed until after it had expired, the suspension
7410 never ended. On busy systems this could lead to processes getting stuck for
7413 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7414 process, before it forks any delivery processes. The open lookup caching
7415 mechanism meant that the open file or database connection was passed into
7416 the delivery process. The problem was that delivery processes always tidy
7417 up cached lookup data. This could cause a problem for the next delivery
7418 process started by the queue runner, because the external queue runner
7419 process does not know about the closure. So the next delivery process
7420 still has data in the lookup cache. In the case of a file lookup, there was
7421 no problem because closing a file descriptor in a subprocess doesn't affect
7422 the parent. However, if the lookup was caching a connection to a database,
7423 the connection was closed, and the second delivery process was likely to
7424 see errors such as "PGSQL: query failed: server closed the connection
7425 unexpectedly". The problem has been fixed by closing all cached lookups
7426 in a queue runner before running a delivery process.
7428 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7429 seem to have the "const" qualifier which it has on other OS. I've
7432 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7433 *of the same type* (client or server) with the same public name that an
7434 error should be diagnosed.
7436 9. When Exim looked up a host name for an IP address, but failed to find the
7437 original IP address when looking up the host name (a safety check), it
7438 output the message "<ip address> does not match any IP for NULL", which was
7439 confusing, to say the least. The bug was that the host name should have
7440 appeared instead of "NULL".
7442 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7443 user that is built into the binary, and the -C or -D options is used, root
7444 privilege is dropped before the configuration file is read. In addition,
7445 logging is switched to stderr instead of the normal log files. If the
7446 configuration then re-defines the Exim user, the unprivileged environment
7447 is probably not what is expected, so Exim logs a panic warning message (but
7450 However, if deliver_drop_privilege is set, the unprivileged state may well
7451 be exactly what is intended, so the warning has been cut out in that case,
7452 and Exim is allowed to try to write to its normal log files.
7458 1. smtp_return_error_details was not giving details for temporary sender
7459 or receiver verification errors.
7461 2. Diagnose a configuration error if two authenticators have the same public
7464 3. Exim used not to create the message log file for a message until the first
7465 delivery attempt. This could be confusing when incoming messages were held
7466 for policy or load reasons. The message log file is now created at the time
7467 the message is received, and an initial "Received" line is written to it.
7469 4. The automatically generated man page for command line options had a minor
7470 bug that caused no ill effects; however, a more serious problem was that
7471 the procedure for building the man page automatically didn't always
7472 operate. Consequently, release 4.20 contains an out-of-date version. This
7473 shouldn't happen again.
7475 5. When building Exim with embedded Perl support, the script that builds the
7476 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7477 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7479 6. The freeze_tell option was not being used for messages that were frozen on
7480 arrival, either by an ACL or by local_scan().
7482 7. Added the smtp_incomplete_transaction log selector.
7484 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7485 was accepting AUTH without a new EHLO.
7487 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7488 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7491 10. Logging of TCP/IP connections (when configured) now happens in the main
7492 daemon process instead of the child process, so that the TCP/IP connection
7493 count is more accurate (but it can never be perfect).
7495 11. The use of "drop" in a nested ACL was not being handled correctly in the
7496 outer ACL. Now, if condition failure induced by the nested "drop" causes
7497 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7498 or "require"), the connection is dropped.
7500 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7501 that yield "discard" can now be used with an "accept" or a "discard" verb,
7502 but an error is generated for any others (because I can't see a useful way
7503 to define what should happen).
7505 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7506 are now processed in the same way as lines in the Exim configuration file.
7507 In particular, continuation lines are supported.
7509 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7511 15. Added -ti meaning -t -i.
7513 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7514 domains, and warn by logging if others are found.
7516 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7517 returned from GIFCONF when Exim is trying to find the list of interfaces on
7518 a host. The code in os.c has been modified to copy each ifreq to an aligned
7519 structure in all cases.
7521 Also, in some cases, the returned ifreq's were being copied to a 'struct
7522 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7523 means the last couple of bytes of an IPv6 address could be chopped if the
7524 ifreq contained only a normal sockaddr (14 bytes storage).
7526 18. Named domain lists were not supported in the hosts_treat_as_local option.
7527 An entry such as +xxxx was not recognized, and was treated as a literal
7530 19. Ensure that header lines added by a DATA ACL are included in the reject log
7531 if the ACL subsequently rejects the message.
7533 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7534 MD5 (which is deprecated).
7536 21. When testing a filter file using -bf, Exim was writing a message when it
7537 took the sender from a "From " line in the message, but it was not doing so
7538 when it took $return_path from a Return-Path: header line. It now does.
7540 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7541 with a valid header line field name (a series of printing characters
7542 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7544 23. Changed "disc" in the source to "disk" to conform to the documentation and
7545 the book and for uniformity.
7547 24. Ignore Sendmail's -Ooption=value command line item.
7549 25. When execve() failed while trying to run a command in a pipe transport,
7550 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7551 could be confused with a return value of 69 from the command itself. This
7552 has been changed to 127, the value the shell returns if it is asked to run
7553 a non-existent command. The wording for the related log line suggests a
7554 non-existent command as the problem.
7556 26. If received_header_text expands to an empty string, do not add a Received:
7557 header line to the message. (Well, it adds a token one on the spool, but
7558 marks it "old" so that it doesn't get used or transmitted.)
7560 27. Installed eximstats 1.28 (addition of -nt option).
7562 28. There was no check for failure on the call to getsockname() in the daemon
7563 code. This can fail if there is a shortage of resources on the system, with
7564 ENOMEM, for example. A temporary error is now given on failure.
7566 29. Contrary to the C standard, it seems that in some environments, the
7567 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7568 program. Exim now does this explicitly; it affects the formatting of
7569 timestamps using strftime().
7571 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7572 complaints. I've now initialized all the variables, to avoid this.
7574 32. Header lines added by a system filter were not being "seen" during
7575 transport-time rewrites.
7577 33. The info_callback() function passed to OpenSSL is set up with type void
7578 (*)(SSL *, int, int), as described somewhere. However, when calling the
7579 function (actually a macro) that sets it up, the type void(*)() is
7580 expected. I've put in a cast to prevent warnings from picky compilers.
7582 34. If a DNS black list lookup found a CNAME record, but there were no A
7583 records associated with the domain it pointed at, Exim crashed.
7585 35. If a DNS black list lookup returned more than one A record, Exim ignored
7586 all but the first. It now scans all returned addresses if a particular IP
7587 value is being sought. In this situation, the contents of the
7588 $dnslist_value variable are a list of all the addresses, separated by a
7591 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7592 to accept a host name and all its aliases if the forward lookup for any of
7593 them yielded the IP address of the incoming connection. Now it accepts only
7594 those names whose forward lookup yields the correct IP address. Any other
7595 names are discarded. This closes a loophole whereby a rogue DNS
7596 administrator could create reverse DNS records to break through a
7597 wildcarded host restriction in an ACL.
7599 37. If a user filter or a system filter that ran in a subprocess used any of
7600 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7601 the wrong values were passed to the pipe command ($thisaddress had the
7602 value of $0, $0 had the value of $1, etc). This bug was introduced by
7603 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7605 38. Improved the line breaking for long SMTP error messages from ACLs.
7606 Previously, if there was no break point between 40 and 75 characters, Exim
7607 left the rest of the message alone. Two changes have been made: (a) I've
7608 reduced the minimum length to 35 characters; (b) if it can't find a break
7609 point between 35 and 75 characters, it looks ahead and uses the first one
7610 that it finds. This may give the occasional overlong line, but at least the
7611 remaining text gets split now.
7613 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7614 file descriptors might be low, and that setting 1000 would always raise it.
7615 It turns out that in some environments, the limit is already over 1000 and
7616 that lowering it causes trouble. So now Exim takes care not to decrease it.
7618 40. When delivering a message, the value of $return_path is set to $sender_
7619 address at the start of routing (routers may change the value). By an
7620 oversight, this default was not being set up when an address was tested by
7621 -bt or -bv, which affected the outcome if any router or filter referred to
7624 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7625 log only when "message" or "log_message" are set. However, if one of the
7626 conditions was an address verification, or a call to a nested ACL, the
7627 messages generated by the underlying test were being passed through. This
7628 no longer happens. The underlying message is available in $acl_verify_
7629 message for both "message" and "log_message" expansions, so it can be
7630 passed through if needed.
7632 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7633 new expansion $bh_ to give the encoded byte string without charset
7634 translation. Translation happens only if iconv() is available; HAVE_ICONV
7635 indicates this at build time. HEADERS_CHARSET gives the charset to
7636 translate to; headers_charset can change it in the configuration, and
7637 "headers charset" can change it in an individual filter file.
7639 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7640 that creates RFC 2047 encoded "words" labels them as that charset instead
7641 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7642 expansion operator; (ii) when Exim creates a From: line for a local
7643 message; (iii) when a header line is rewritten to include a "phrase" part.
7645 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7646 buggy, causing it to skip the first lines of messages whose message ID
7647 ended in 'D'. This would not have bitten before Exim release 4.14, because
7648 message IDs were unlikely to end in 'D' before then. The effect was to have
7649 incorrect size information for certain domains.
7651 45. #include "config.h" was missing at the start of the crypt16.c module. This
7652 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7655 46. If there was a timeout during a "random" callout check, Exim treated it as
7656 a failure of the random address, and carried on sending RSET and the real
7657 address. If the delay was just some slowness somewhere, the response to the
7658 original RCPT would be taken as a response to RSET and so on, causing
7659 mayhem of various kinds.
7661 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7662 when I implemented it. It didn't allow for the fact that some option values
7663 may legitimately be negative (e.g. size_addition), and it didn't even do
7664 the right test for positive values.
7666 48. Domain names in DNS records are case-independent. Exim always looks them up
7667 in lower case. Some resolvers return domain names in exactly the case they
7668 appear in the zone file, that is, they may contain uppercase letters. Not
7669 all resolvers do this - some return always lower case. Exim was treating a
7670 change of case by a resolver as a change of domain, similar to a widening
7671 of a domain abbreviation. This triggered its re-routing code and so it was
7672 trying to route what was effectively the same domain again. This normally
7673 caused routing to fail (because the router wouldn't handle the domain
7674 twice). Now Exim checks for this case specially, and just changes the
7675 casing of the domain that it ultimately uses when it transmits the message
7678 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7681 50. If a filter generated a file delivery with a non-absolute name (possible if
7682 no home directory exists for the router), the forbid_file option was not
7685 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7686 the existing equality matching.
7688 52. Exim was using ints instead of ino_t variables in some places where it was
7689 dealing with inode numbers.
7691 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7692 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7693 called TMPDIR, and if it finds it is different, it changes its value.
7695 54. The smtp_printf() function is now made available to local_scan() so
7696 additional output lines can be written before returning. There is also an
7697 smtp_fflush() function to enable the detection of a dropped connection.
7698 The variables smtp_input and smtp_batched_input are exported to
7701 55. Changed the default runtime configuration: the message "Unknown user"
7702 has been removed from the ACL, and instead placed on the localuser router,
7703 using the cannot_route_message feature. This means that any verification
7704 failures that generate their own messages won't get overridden. Similarly,
7705 the "Unrouteable address" message that was in the ACL for unverifiable
7706 relay addresses has also been removed.
7708 56. Added hosts_avoid_esmtp to the smtp transport.
7710 57. The exicyclog script was not checking for the esoteric option
7711 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7712 will work only if exicyclog is run under the appropriate euid.
7714 58. Following a discussion on the list, the rules by which Exim recognises line
7715 endings on incoming messages have been changed. The -dropcr and drop_cr
7716 options are now no-ops, retained only for backwards compatibility. The
7717 following line terminators are recognized: LF CRLF CR. However, special
7718 processing applies to CR:
7720 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7721 nor a local message in the state where . is a terminator.
7723 (ii) If a bare CR is encountered in a header line, an extra space is added
7724 after the line terminator so as not to end the header. The reasoning
7725 behind this is that bare CRs in header lines are most likely either
7726 to be mistakes, or people trying to play silly games.
7728 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7729 was being incorrectly given as 18 bytes larger than it should have been.
7730 This is a VOB (very old bug).
7732 60. This may never have affected anything current, but just in case it has:
7733 When the local host is found other than at the start of a list of hosts,
7734 the local host, those with the same MX, and any that follow, are discarded.
7735 When the list in question was part of a longer list of hosts, the following
7736 hosts (not currently being processed) were also being discarded. This no
7737 longer happens. I'm not sure if this situation could ever has previously
7740 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7743 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7744 header. This is something that is recommended in a new Internet Draft, and
7745 is something that is documented as being done by Sendmail. There are two
7746 possible values. For messages generated by the autoreply transport, Exim
7749 Auto-Submitted: auto-replied
7751 whereas for all other generated messages (e.g. bounces) it adds
7753 Auto-Submitted: auto-generated
7755 63. The "personal" condition in filters now includes a test for the
7756 Auto-Submitted: header. If it contains the string "auto-" the message it
7757 not considered personal.
7759 64. Added rcpt_include_affixes as a generic transport option.
7761 65. Added queue_only_override (default true).
7763 66. Added the syslog_duplication option.
7765 67. If what should have been the first header line of a message consisted of
7766 a space followed by a colon, Exim was mis-interpreting it as a header line.
7767 It isn't of course - it is syntactically invalid and should therefore be
7768 treated as the start of the message body. The misbehaviour could have
7769 caused a number of strange effects, including loss of data in subsequent
7770 header lines, and spool format errors.
7772 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7773 client host had authenticated. This control can now be exercised by an ACL
7774 for more flexibility.
7776 69. By default, callouts do not happen when testing with -bh. There is now a
7777 variant, -bhc, which does actually run the callout code, including
7778 consulting and updating the callout cache.
7780 70. Added support for saslauthd authentication, courtesy of Alexander
7783 71. If statvfs() failed on the spool or log directories while checking their
7784 size for availability, Exim confusingly gave the error "space shortage".
7785 Furthermore, in debugging mode it crashed with a floating point exception.
7786 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7787 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7788 serious problem, Exim now writes to the main and panic logs when this
7789 happens, with details of the failure. It then refuses to accept the
7790 incoming message, giving the message "spool directory problem" or "log
7791 directory problem" with a 421 code for SMTP messages.
7793 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7794 0, 1, and 2 exist, because some OS complain for execs without them (see
7795 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7796 descriptors. However, the code omitted to check that the open succeeded,
7797 causing mysterious errors if for some reason the permissions on /dev/null
7798 got screwed. Now Exim writes a message to the main and panic logs, and
7799 bombs out if it can't open /dev/null.
7801 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7802 interact so that it is all more flexible. It is supposed to remain
7803 backwards compatible. Also added extra_local_interfaces.
7805 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7806 to bomb out with an assertion failure - to the client this appears as a
7807 connection drop. This problem occurs in the part of the code that was taken
7808 from the Samba project. Fortunately, the assertion is in a very simple
7809 function, so I have fixed this by reproducing the function inline in the
7810 one place where it is called, and arranging for authentication to fail
7811 instead of killing the process with assert().
7813 75. The SPA client code was not working when the server requested OEM rather
7814 than Unicode encoding.
7816 76. Added code to make require_files with a specific uid setting more usable in
7817 the case where statting the file as root fails - usually a non-root-mounted
7818 NFS file system. When this happens and the failure is EACCES, Exim now
7819 forks a subprocess and does the per-uid checking as the relevant uid.
7821 77. Added process_log_path.
7823 78. If log_file_path was not explicitly set, a setting of check_log_space or
7824 check_log_inodes was ignored.
7826 79. If a space check for the spool or log partitions fails, the incident is now
7827 logged. Of course, in the latter case the data may get lost...
7829 80. Added the %p formatting code to string_format() so that it can be used to
7830 print addresses in debug_print(). Adjusted all the address printing in the
7831 debugging in store.c to use %p rather than %d.
7833 81. There was a concern that a line of code in smtp_in.c could overflow a
7834 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7835 initially expressed, the concern was not well-founded, because trailing
7836 spaces are removed early. However, if the trailing spaces were followed by
7837 a NULL, they did not get removed, so the overflow was possible. Two fixes
7840 (a) I re-wrote the offending code in a cleaner fashion.
7841 (b) If an incoming SMTP command contains a NULL character, it is rejected
7844 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7845 runs initgroups(), so that if the Exim user is in any additional groups,
7846 they will be used during message reception.
7852 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob