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.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
99 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
100 not in 127.0.0.0/8 to help in spotting list domains taken over by a
101 domain-parking registrar.
103 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
104 Previously when a whitespace character was specified it was not inserted
105 after removing the newline.
107 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
108 the domain part of the recipient address. This overrides any tls_sni
109 option set, which was previously used.
111 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
114 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
115 is_tainted() had an off-by-one error in the overenthusiastic direction.
116 Find and fix by Gavan. Although NetBSD is not a supported platform for
117 4.94 this bug could affect other platforms.
123 JH/01 Avoid costly startup code when not strictly needed. This reduces time
124 for some exim process initialisations. It does mean that the logging
125 of TLS configuration problems is only done for the daemon startup.
127 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
129 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
130 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
133 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
134 DKIM signing is being done. Previously a transport_filter would always
135 disable CHUNKING, falling back to traditional DATA.
137 JH/05 Regard command-line receipients as tainted.
139 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
141 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
142 PAM library frees one of the arguments given to it, despite the
143 documentation. Therefore a plain malloc must be used.
145 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
146 on-stack buffers were used, resulting in a taint trap when DSN information
147 copied from a received message was written into the buffer.
149 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
150 the ordering of its ARC headers. This caused a crash.
152 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
153 a new record was being constructed with information from the peer, a trap
156 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
157 installation would get error messages from DMARC verify, when it hit the
158 nonexistent file indicated by the default. Distros wanting DMARC enabled
159 should both provide the file and set the option.
160 Also enforce no DMARC verification for command-line sourced messages.
162 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
163 could, depending on the platform, hang at the STARTTLS response.
165 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
166 message on a connection. Previously if one message had ARC headers and
167 the following one did not, a crash could result when adding an
168 Authentication-Results: header.
170 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
171 Exim main code and Exim-related utities. The introduction of taint
172 tracking also did many adjustments to string handling. Since then, eximon
173 frequently terminated with an assert failure.
175 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
176 check for 452 responses. This slightly helps the inefficieny of doing
177 a large alias-expansion into a recipient-limited target. The max_rcpt
178 transport option still applies (and at the current default, will override
179 the new feature). The check is done for either cause of synch, and forces
180 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
181 connection. The new facility is not tunable at this time.
183 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
184 library live data was being used, so the results became garbage. Make
185 copies while it is still usable.
187 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
188 on delivery deferred (==) and failed (**) lines (if a delivery was
189 attemtped). Previously it was only on completion (=>) lines.
191 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
192 for the expansion of the server_scram_iter and server_scram_salt options.
194 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
195 are now specifically given a NO_DATA response without hitting the system
196 resolver. The library goes on to do the now-standard TXT lookup.
197 Use of dnsdb lookups is not affected.
199 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
200 only retrieve the errormessage once. Previously two calls to dlerror()
201 were used, and the second one (for mainlog/paniclog) retrieved null
204 JH/20 Taint checking: disallow use of tainted data for
205 - the appendfile transport file and directory options
206 - the pipe transport command
207 - the autoreply transport file, log and once options
208 - file names used by the redirect router (including filter files)
210 - paths used by single-key lookups
211 Previously this was permitted.
213 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
214 adjusted the size of a major service buffer; this failed because the
215 buffer was in use at the time. Change to a compile-time increase in the
216 buffer size, when this authenticator is compiled into exim.
218 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
219 previous fast-mode was untenable in the face of glibs using mmap to
220 support larger malloc requests.
222 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
223 New values supported, if defined on system where compiled:
224 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
225 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
227 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
228 running a limited number of proceses in parallel, a benefit is gained. The
229 amount varies with the platform hardware and load. The use of the option
230 queue_run_in_order means we cannot do this, as ordering becomes
233 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
234 had introduced a string-copy (for ensuring NUL-termination) which was not
235 appropriate for that case, which can include embedded NUL bytes in the
236 block of data. Investigation showed the copy to actually be needless, the
237 data being length-specified.
239 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
240 done during a receiving connection, and both used TLS, global info was
241 used rather than per-connection info for tracking the state of data
242 queued for transmission. This could result in a connection hang.
244 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
245 Previously, when delivering serveral messages down a single connection
246 only the first would provide a SIZE. This was due to the size information
247 not being properly tracked.
249 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
250 TAI (at 37 seconds currently), pretend to be in UTC for time-related
251 expansion and logging. Previously, spurious values such as a future
252 minute could be seen.
254 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
255 it could crash from a null-deref. This could also affect the
256 ${addresses: } operator and ${readsock } item.
258 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
259 message following a mime one, the variable was not reset.
261 JH/30 When an pipelined-connect fails at the first response, assume incorrect
262 cached capability (perhaps the peer reneged?) and immediately retry in
265 JH/31 Fix spurious detection of timeout while writing to transport filter.
267 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
268 an attempt to copy the string was made before checking it.
270 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
271 taint of the lookup key was maintained; we now regard the presence in the
272 filesystem as sufficient validation.
274 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
275 argument is supplied.
277 JH/35 The dsearch lookup now requires that the directory is an absolute path.
278 Previously this was not checked, and nonempty relative paths made an
279 access under Exim's current working directory.
281 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
282 Previously no event was raised.
284 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
285 parameter supplied by the sender MAIL FROM command. Previously it was
286 ignored, and only the check_spool_space option value for the required
289 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
290 the size of the signing public-key. Previously it was instead giving
291 the size of the signature hash.
293 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
294 the default. See the (new) dkim_verify_min_keysizes option.
296 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
297 and an ACL use a lookup for checking either the local_part or domain,
298 stale data could be accessed. Ensure that variable references are
299 dropped between messages.
301 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
302 by the client was not checked as pointing within response data before
303 being used. A malicious client could thus cause an out-of-bounds read and
304 possibly gain authentication. Fix by adding the check.
306 JH/42 Internationalisation: change the default for downconversion in the smtp
307 transport to be "if needed". Previously it was "as previously set" for
308 the message, which usually meant "if needed" for message-submission but
309 "no" for everything else. However, MTAs have been seen using SMTPUTF8
310 even when the envelope addresses did not need it, resulting in forwarding
311 failures to non-supporting MTAs. A downconvert in such cases will be
312 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
313 transport. The change does mean that addresses needing conversion will
314 be converted when previously a delivery failure would occur.
316 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
317 response was received it would be used unchecked in a fail-DSN, violating
318 standards on line-length limits. Truncate if needed.
320 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
321 doesn't work. (Additionally add a new main config option to configure the
328 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
329 side, to decode a TLS 1.3 packet capture.
331 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
332 Previously the default library behaviour applied, sending two, each in
335 JH/03 Debug output for ACL now gives the config file name and line number for
338 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
340 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
342 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
343 buffer overrun for (non-chunking) other transports.
345 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
346 TLS1.3, means that a server rejecting a client certificate is not visible
347 to the client until the first read of encrypted data (typically the
348 response to EHLO). Add detection for that case and treat it as a failed
349 TLS connection attempt, so that the normal retry-in-clear can work (if
350 suitably configured).
352 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
353 and/or domain. Found and fixed by Jason Betts.
355 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
356 configuration). If a CNAME target was not a wellformed name pattern, a
359 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
360 the OS reports them interleaved with other addresses.
362 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
363 used both for input and for a verify callout, both encrypted, SMTP
364 responses being sent by the server could be lost. This resulted in
365 dropped connections and sometimes bounces generated by a peer sending
368 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
369 string. Previously it was possible to cause undefined behaviour in a
370 library routine (usually a crash). Found by "zerons".
372 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
375 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
376 API was removed, so update to use the newer ones.
378 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
379 any timeout set, is taking a long time. Previously we would hang on to a
380 rotated logfile "forever" if the input was arriving with long gaps
381 (a previous attempt to fix addressed lack, for a long time, of initial
384 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
385 shared (NFS) environment. The length of the tempfile name is now
386 4 + 16 ("hdr.$message_exim_id") which might break on file
387 systems which restrict the file name length to lower values.
390 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
391 shared (NFS) environment.
393 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
394 did for all versions <4.90). Notably -M, -m, --invert, -I may be
397 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
398 on some platforms for bit 31.
400 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
401 to changes apparently associated with TLS1.3 handling some of the APIs
402 previously used were either nonfunctional or inappropriate. Strings
403 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
404 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
405 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
406 This affects log line X= elements, the $tls_{in,out}_cipher variables,
407 and the use of specific cipher names in the encrypted= ACL condition.
409 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
411 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
412 verification result was not updated unless hosts_require_ocsp applied.
414 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
415 queue_list_requires_admin set to false, non-admin users were denied the
418 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
419 directory-of-certs mode. Previously they were advertised despite the
422 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
423 A single TCP connection by a client will now hold a TLS connection open
424 for multiple message deliveries, by default. Previously the default was to
427 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
428 default. If built with the facility, DANE will be used. The facility
429 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
431 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
432 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
433 must be defined and you must still, unless you define DISABLE_TLS, manage
434 the the include-dir and library-file requirements that go with that
435 choice. Non-TLS builds are still supported.
437 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
440 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
441 default. If the platform supports and has the facility enabled, it will
442 be requested on all coneections.
444 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
445 controlled by the build-time option SUPPORT_PIPE_CONNECT.
447 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
449 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
450 success-DSN messages. Previously the From: header was always the default
451 one for these; the option was ignored.
453 JH/28 Fix the timeout on smtp response to apply to the whole response.
454 Previously it was reset for every read, so a teergrubing peer sending
455 single bytes within the time limit could extend the connection for a
456 long time. Credit to Qualsys Security Advisory Team for the discovery.
458 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
459 delivery address, which leaked information of the results of local
460 forwarding. Change to the original envelope recipient address, per
463 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
464 requested. Previously not bounce was generated and a log entry of
465 error ignored was made.
467 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
469 JH/32 Introduce a general tainting mechanism for values read from the input
470 channel, and values derived from them. Refuse to expand any tainted
471 values, to catch one form of exploit.
473 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
474 was unused and the unexpanded text used for the test. Found and
475 fixed by Ruben Jenster.
477 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
478 an attempt to use a TLS library read routine dereffed a nul pointer,
481 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
482 them in our smtp response.
484 JH/36 Have the general router option retry_use_local_part default to true when
485 any of the restrictive preconditions are set (to anything). Previously it
486 was only for check_local user. The change removes one item of manual
487 configuration which is required for proper retries when a remote router
488 handles a subset of addresses for a domain.
490 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
491 link count into consideration.
493 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
494 caused the extension of big_buffer, the following lines were ignored.
496 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
497 accordance with RFC 2308. Previously there was no expiry, so a longlived
498 receive process (eg. due to ACL delays) versus a short SOA value could
501 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
503 JH/39 Promote DMARC support to mainline.
505 JH/40 Bug 2452: Add a References: header to DSNs.
507 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
508 parameters. The relevant library call is documented as "Deprecated: This
509 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
510 3.6.0, DH parameters are negotiated following RFC7919."
512 HS/06 Change the default of dnssec_request_domains to "*"
514 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
515 carried on and emitted a BDAT command, even when PIPELINING was not
518 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
519 buffer was used for the filename, resulting in a trap when tainted
520 arguments (eg. $domain) were used.
522 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
523 recommended to avoid a possible server-load attack. The feature can be
524 re-enabled via the openssl_options main cofiguration option.
526 JH/45 local_scan API: documented the current smtp_printf() call. This changed
527 for version 4.90 - adding a "more data" boolean to the arguments.
528 Bumped the ABI version number also, this having been missed previously;
529 release versions 4.90 to 4.92.3 inclusive were effectively broken in
530 respect of usage of smtp_printf() by either local_scan code or libraries
531 accessed via the ${dlfunc } expansion item. Both will need coding
532 adjustment for any calls to smtp_printf() to match the new function
533 signature; a FALSE value for the new argument is always safe.
535 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
536 the file-offset (which the Linux syscall does, and exim expects); this
537 resulted in an indefinite loop.
539 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
540 to do ARC verification. The Authentication-Results: header line added
541 by the configuration then had no ARC item.
547 JH/01 Remove code calling the customisable local_scan function, unless a new
548 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
550 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
551 non-signal-safe functions being used.
553 JH/03 Bug 2269: When presented with a received message having a stupidly large
554 number of DKIM-Signature headers, disable DKIM verification to avoid
555 a resource-consumption attack. The limit is set at twenty.
557 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
558 report of oldest_pass in ${authres } in consequence, and separate out
559 some descriptions of reasons for verification fail.
561 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
562 files in the spool were present and unlocked. A queue-runner could spot
563 them, resulting in a duplicate delivery. Fix that by doing the unlock
564 after the unlink. Investigation by Tim Stewart. Take the opportunity to
565 add more error-checking on spoolfile handling while that code is being
568 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
569 No known attacks, no CVE, this is defensive hardening.
571 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
572 a queue-runner could start a delivery while other operations were ongoing.
573 Cutthrough delivery was a common victim, resulting in duplicate delivery.
574 Found and investigated by Tim Stewart. Fix by using the open message data
575 file handle rather than opening another, and not locally closing it (which
576 releases a lock) for that case, while creating the temporary .eml format
577 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
579 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
580 $sender_verify_failure/$recipient_verify_failure to "random".
582 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
585 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
586 Previously this would segfault.
588 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
591 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
592 like zero, since the resolver should be doing this for us, But we need one
593 as a CNAME but no MX presence gets the CNAME returned; we need to check
594 that doesn't point to an MX to declare it "no MX returned" rather than
595 "error, loop". A new main option is added so the older capability of
596 following some limited number of chain links is maintained.
598 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
600 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
601 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
602 now-deprecated earlier definitions used only specified the range up to TLS
603 1.2 (in the older-version library docs).
605 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
607 JH/15 Rework TLS client-side context management. Stop using a global, and
608 explicitly pass a context around. This enables future use of TLS for
609 connections to service-daemons (eg. malware scanning) while a client smtp
610 connection is using TLS; with cutthrough connections this is quite likely.
612 JH/16 Fix ARC verification to do AS checks in reverse order.
614 JH/17 Support a "tls" option on the ${readsocket } expansion item.
616 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
617 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
618 Previously the "utf8" would be re-prepended for every additional message.
620 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
621 Previously thery were accepted, resulting in issues when attempting to
622 forward messages to a non-supporting MTA.
624 PP/02 Let -n work with printing macros too, not just options.
626 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
627 one parent address was copied, and bogus data was used at delivery-logging
628 time. Either a crash (after delivery) or bogus log data could result.
629 Discovery and analysis by Tim Stewart.
631 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
632 Previously if the string ended mid-character, we did not insert the
633 promised '?' replacement.
635 PP/04 Documentation: current string operators work on bytes, not codepoints.
637 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
638 should pack well giving a smaller memory footprint so better caching and
639 therefore performance. Group the declarations where this can't be done so
640 that the byte-sized flag variables are not interspersed among pointer
641 variables, giving a better chance of good packing by the compiler.
643 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
644 non-null, to avoid issues with sites running BATV. Previously reports were
645 sent with an empty envelope sender so looked like bounces.
647 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
648 The ignore_error flag wasn't being returned from the filter subprocess so
649 was not set for later routers. Investigation and fix by Matthias Kurz.
651 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
652 and a msg:complete for the whole, when a message is manually removed using
653 -Mrm. Developement by Matthias Kurz, hacked on by JH.
655 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
656 a "Gnu special" function, asprintf() in the DB utility binary builds; I
657 hope that is portable enough.
659 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
660 requiring a known-CA anchor certificate; make it now rely entirely on the
661 TLSA as an anchor. Checking the name on the leaf cert against the name
662 on the A-record for the host is still done for TA (but not for EE mode).
664 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
665 included in delivery lines for non-proxied connections, when compiled with
666 SUPPORT_SOCKS and running with proxy logging enabled.
668 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
669 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
670 move the existing event to fire before the normal logging of message
671 failure so that custom logging is bracketed by normal logging.
673 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
674 msg:fail:internal event. Developement by Matthias Kurz.
676 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
677 far too small for todays use of crypto signatures stored there. Go all
678 the way to the max DNS message size of 64kB, even though this might be
679 overmuch for IOT constrained device use.
681 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
682 copy a string over itself. The library routine is documented as not
683 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
685 JH/32 For main options check_spool_space and check_inode_space, where the
686 platform supports 64b integers, support more than the previous 2^31 kB
687 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
688 the previous G, M, k.
690 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
691 $authenticated_fail_id variable on authentication failure. Previously
694 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
695 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
696 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
697 GNUTLS_SEC_PARAM_MEDIUM.
699 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
700 side. Previously we would continue as if no SNI had been received.
702 JH/36 Harden the handling of string-lists. When a list consisted of a sole
703 "<" character, which should be a list-separator specification, we walked
704 off past the nul-terimation.
706 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
707 causes) even when the retry time is not yet met. Previously they were
708 not, meaning that when (say) an account was over-quota and temp-rejecting,
709 and multiple senders' messages were queued, only one sender would get
710 notified on each configured delay_warning cycle.
712 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
714 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
715 cipher-suites, an error can be left on the stack even for a succeeding
716 accept; this results in impossible error messages when a later operation
719 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
720 return error codes indicating retry. Under TLS1.3 this becomes required.
722 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
723 it only wrote the new authenticators, resulting in a lack of tracking of
724 peer changes of ESMTP extensions until the next cache flush.
726 JH/41 Fix the loop reading a message header line to check for integer overflow,
727 and more-often against header_maxsize. Previously a crafted message could
728 induce a crash of the recive process; now the message is cleanly rejected.
730 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
731 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
737 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
738 When redis_servers is set to a list of > 1 element, and the Redis servers
739 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
740 case of MOVED into a DEFER case instead, thus moving the query onto the
741 next server in the list. For a cluster of N elements, all N servers must
742 be defined in redis_servers.
744 GF/02 Catch and remove uninitialized value warning in exiqsumm
745 Check for existence of @ARGV before looking at $ARGV[0]
747 JH/01 Replace the store_release() internal interface with store_newblock(),
748 which internalises the check required to safely use the old one, plus
749 the allocate and data copy operations duplicated in both (!) of the
750 extant use locations.
752 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
753 modifier. This matches the restriction on the commandline.
755 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
756 Previously only the last row was returned.
758 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
759 we assumed that tags in the header were well-formed, and parsed the
760 element content after inspecting only the first char of the tag.
761 Assumptions at that stage could crash the receive process on malformed
764 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
765 While running the DKIM ACL we operate on the Permanent memory pool so that
766 variables created with "set" persist to the DATA ACL. Also (at any time)
767 DNS lookups that fail create cache records using the Permanent pool. But
768 expansions release any allocations made on the current pool - so a dnsdb
769 lookup expansion done in the DKIM ACL releases the memory used for the
770 DNS negative-cache, and bad things result. Solution is to switch to the
771 Main pool for expansions.
772 While we're in that code, add checks on the DNS cache during store_reset,
773 active in the testsuite.
774 Problem spotted, and debugging aided, by Wolfgang Breyha.
776 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
777 When none of the hosts presented to a transport match an already-open
778 connection, close it and proceed with the list. Previously we would
779 queue the message. Spotted by Lena with Yahoo, probably involving
782 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
783 Previously a spurious "250 OK id=" response was appended to the proper
786 JH/08 The "support for" informational output now, which built with Content
787 Scanning support, has a line for the malware scanner interfaces compiled
788 in. Interface can be individually included or not at build time.
790 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
791 by the template makefile "src/EDITME". The "STREAM" support for an older
792 ClamAV interface method is removed.
794 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
795 rows affected is given instead).
797 JH/11 The runtime Berkeley DB library version is now additionally output by
798 "exim -d -bV". Previously only the compile-time version was shown.
800 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
801 SMTP connection. Previously, when one had more receipients than the
802 first, an abortive onward connection was made. Move to full support for
803 multiple onward connections in sequence, handling cutthrough connection
804 for all multi-message initiating connections.
806 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
807 routers. Previously, a multi-recipient message would fail to match the
808 onward-connection opened for the first recipient, and cause its closure.
810 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
811 a timeout on read on a GnuTLS initiating connection, resulting in the
812 initiating connection being dropped. This mattered most when the callout
813 was marked defer_ok. Fix to keep the two timeout-detection methods
816 JH/15 Relax results from ACL control request to enable cutthrough, in
817 unsupported situations, from error to silently (except under debug)
818 ignoring. This covers use with PRDR, frozen messages, queue-only and
821 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
823 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
824 metadata, resulting in a crash in free().
826 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
827 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
828 Broken also in d185889f4, with init system revamp.
830 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
831 Previously we abruptly closed the connection after reading a malware-
832 found indication; now we go on to read the "scan ok" response line,
835 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
836 ACL. Previously, a crash would result.
838 JH/19 Speed up macro lookups during configuration file read, by skipping non-
839 macro text after a replacement (previously it was only once per line) and
840 by skipping builtin macros when searching for an uppercase lead character.
842 JH/20 DANE support moved from Experimental to mainline. The Makefile control
843 for the build is renamed.
845 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
846 was allocated for every new TLS startup, meaning one per message. Fix
847 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
849 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
850 reported the original. Fix to report (as far as possible) the ACL
851 result replacing the original.
853 JH/23 Fix memory leak during multi-message connections using STARTTLS under
854 OpenSSL. Certificate information is loaded for every new TLS startup,
855 and the resources needed to be freed.
857 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
859 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
862 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
863 DATA response info to the (existing) per-recipient response info for
864 the "C=" log element. It can have useful tracking info from the
865 destination system. Patch from Simon Arlott.
867 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
868 length value. Previously this would segfault.
870 HS/02 Support Avast multiline protoocol, this allows passing flags to
871 newer versions of the scanner.
873 JH/28 Ensure that variables possibly set during message acceptance are marked
874 dead before release of memory in the daemon loop. This stops complaints
875 about them when the debug_store option is enabled. Discovered specifically
876 for sender_rate_period, but applies to a whole set of variables.
877 Do the same for the queue-runner and queue-list loops, for variables set
878 from spool message files. Do the same for the SMTP per-message loop, for
879 certain variables indirectly set in ACL operations.
881 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
882 as a multi-recipient message from a mailinglist manager). The coding had
883 an arbitrary cutoff number of characters while checking for more input;
884 enforced by writing a NUL into the buffer. This corrupted long / fast
885 input. The problem was exposed more widely when more pipelineing of SMTP
886 responses was introduced, and one Exim system was feeding another.
887 The symptom is log complaints of SMTP syntax error (NUL chars) on the
888 receiving system, and refused recipients seen by the sending system
889 (propating to people being dropped from mailing lists).
890 Discovered and pinpointed by David Carter.
892 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
893 replaced by the ${authresults } expansion.
895 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
897 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
898 allows proper process termination in container environments.
900 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
901 Previously the "final dot" had a newline after it; ensure it is CR,LF.
903 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
904 and "err_perm", deprecated since 4.83 when the RFC-defined words
905 "temperror" and "permerror" were introduced.
907 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
908 transport-filters or DKIM-signing. The restriction was lost in the
909 consolidation of verify-callout and delivery SMTP handling.
910 Extend the restriction to also cover ARC-signing.
912 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
913 in defer=pass mode supply a 450 to the initiator. Previously the message
916 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
917 tls_require_ciphers is used as before.
919 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
920 "pass_unscanned". Only tmpfails from the scanner are written to
921 the paniclog, as they may require admin intervention (permission
922 denied, license issues). Other scanner errors (like decompression
923 bombs) do not cause a paniclog entry.
925 JH/36 Fix reinitialisation of DKIM logging variable between messages.
926 Previously it was possible to log spurious information in receive log
929 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
930 triggered odd behaviour from Outlook Express clients.
932 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
935 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
936 since the IETF WG has not yet settled on that versus the original
937 "bare" representation.
939 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
940 Previously the millisecond value corrupted the output.
941 Fix also for syslog_pid=no and log_selector +pid, for which the pid
942 corrupted the output.
948 JH/01 Rework error string handling in TLS interface so that the caller in
949 more cases is responsible for logging. This permits library-sourced
950 string to be attached to addresses during delivery, and collapses
951 pairs of long lines into single ones.
953 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
954 during configuration. Wildcards are allowed and expanded.
956 JH/02 Rework error string handling in DKIM to pass more info back to callers.
957 This permits better logging.
959 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
960 do not close it down and have the child transport start it up again on
961 the passed-on TCP connection. Instead, proxy the child (and any
962 subsequent ones) for TLS via a unix-domain socket channel. Logging is
963 affected: the continued delivery log lines do not have any DNSSEC, TLS
964 Certificate or OCSP information. TLS cipher information is still logged.
966 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
967 identical IP addresses on different listening ports. Will also affect
970 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
971 add noisy ifdef guards to special-case this sillyness.
972 Patch from Bernd Kuhls.
974 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
975 than 255 are no longer allowed.
977 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
978 Disable the session-cache too, which might reduce our load. Since we
979 currrectly use a new context for every connection, both as server and
980 client, there is no benefit for these.
981 GnuTLS appears to not support tickets server-side by default (we don't
982 call gnutls_session_ticket_enable_server()) but client side is enabled
983 by default on recent versions (3.1.3 +) unless the PFS priority string
986 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
987 <https://reproducible-builds.org/specs/source-date-epoch/>.
989 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
990 the check for any unsuccessful recipients did not notice the limit, and
991 erroneously found still-pending ones.
993 JH/08 Pipeline CHUNKING command and data together, on kernels that support
994 MSG_MORE. Only in-clear (not on TLS connections).
996 JH/09 Avoid using a temporary file during transport using dkim. Unless a
997 transport-filter is involved we can buffer the headers in memory for
998 creating the signature, and read the spool data file once for the
999 signature and again for transmission.
1001 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1002 4.77 as the kernel support then wasn't solid, having issues in 64bit
1003 mode. Now, it's been long enough. Add support for FreeBSD also.
1005 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1006 case where the routing stage had gathered several addresses to send to
1007 a host before calling the transport for the first, we previously failed
1008 to close down TLS in the old transport process before passing the TCP
1009 connection to the new process. The new one sent a STARTTLS command
1010 which naturally failed, giving a failed delivery and bloating the retry
1011 database. Investigation and fix prototype from Wolfgang Breyha.
1013 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1014 false-negatives in the check that the sender had not preempted a response
1015 or prompt from Exim (running as a server), due to that code's lack of
1016 awareness of the SMTP input buffering.
1018 PP/04 Add commandline_checks_require_admin option.
1019 Exim drops privileges sanely, various checks such as -be aren't a
1020 security problem, as long as you trust local users with access to their
1021 own account. When invoked by services which pass untrusted data to
1022 Exim, this might be an issue. Set this option in main configuration
1023 AND make fixes to the calling application, such as using `--` to stop
1026 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1027 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1028 responses to those, into a single TLS record each way (this usually means
1029 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1032 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1033 affects you only if you're dancing at the edge of the param size limits.
1034 If you are, and this message makes sense to you, then: raise the
1035 configured limit or use OpenSSL 1.1. Nothing we can do for older
1038 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1039 cmdline element to get the documented default one. Previously it was
1042 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1043 are unsuable for non-message-specific reasons - eg. connection timeout,
1044 banner-time rejection.
1046 JH/16 Fix logging of delivery remote port, when specified by router, under
1049 PP/06 Repair manualroute's ability to take options in any order, even if one
1050 is the name of a transport.
1053 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1055 JH/17 Change the list-building routines interface to use the expanding-string
1056 triplet model, for better allocation and copying behaviour.
1058 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1059 Previously it was constructed the first time a possibly-matching string
1060 was met in the configuration file input during startup; now it is done
1063 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1064 compatible one, to avoid the (poorly documented) possibility of a config
1065 file in the working directory redirecting the DB files, possibly correpting
1066 some existing file. CVE-2017-10140 assigned for BDB.
1068 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1069 cache-hot. Previously, although the result was properly cached, the
1070 initial verify call returned a defer.
1072 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1073 the main verify for receipient in uncached-mode.
1075 JH/22 Retire historical build files to an "unsupported" subdir. These are
1076 defined as "ones for which we have no current evidence of testing".
1078 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1079 if present. Previously it was ignored.
1081 JH/24 Start using specified-initialisers in C structure init coding. This is
1082 a C99 feature (it's 2017, so now considered safe).
1084 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1085 if was a fixed-sized field and bitmask ops via macros; it is now more
1088 PP/07 GitHub PR 56: Apply MariaDB build fix.
1089 Patch provided by Jaroslav Škarvada.
1091 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1092 during Coverity cleanups [4.87 JH/47]
1093 Diagnosis and fix provided by Michael Fischer v. Mollard.
1095 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1096 the right size to place the terminating semicolon on its own folded
1097 line, the header hash was calculated to an incorrect value thanks to
1098 the (relaxed) space the fold became.
1100 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1101 and confused the parent.
1103 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1104 which could crash as a result. This could lead to undeliverable messages.
1106 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1109 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1110 matters on fast-turnover and PID-randomising systems, which were getting
1111 out-of-order delivery.
1113 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1114 a possibly-overlapping copy. The symptom was that "Remote host closed
1115 connection in response to HELO" was logged instead of the actual 4xx
1118 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1119 Previously only that bufferd was discarded, resulting in SYMTP command
1122 JH/32 DKIM: when a message has multiple signatures matching an identity given
1123 in dkim_verify_signers, run the dkim acl once for each. Previously only
1124 one run was done. Bug 2189.
1126 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1127 "panic the current process" to "deliberately defer". The panic log is
1128 still written with the problem list name; the mail and reject logs now
1129 get a temp-reject line for the message that was being handled, saying
1130 something like "domains check lookup or other defer". The SMTP 451
1131 message is still "Temporary local problem".
1133 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1134 A crafted sequence of BDAT commands could result in in-use memory beeing
1135 freed. CVE-2017-16943.
1137 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1138 from SMTP input. Previously it was always done; now only done for DATA
1139 and not BDAT commands. CVE-2017-16944.
1141 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1142 to the message (such as an overlong header line). Previously this was
1143 not done and we did not exit BDAT mode. Followon from the previous item
1144 though a different problem.
1150 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1151 than -2003 did; needs libidn2 in addition to libidn.
1153 JH/02 The path option on a pipe transport is now expanded before use.
1155 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1156 Patch provided by "Björn", documentation fix added too.
1158 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1159 missing a wire-to-host endian conversion.
1161 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1162 close after a BDAT command line could be taken as a following command,
1163 giving a synch failure. Fix by only checking for synch immediately
1164 before acknowledging the chunk.
1166 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1167 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1168 macro. Patches provided by Josh Soref.
1170 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1171 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1172 by IANA as a service extension. However, John Klensin suggests that we
1175 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1176 the dkim code may be unix-mode line endings rather than smtp wire-format
1177 CRLF, so prepend a CR to any bare LF.
1179 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1180 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1182 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1183 the first were themselves being wrongly included in the feed into dkim
1184 processing; with most chunk sizes in use this resulted in an incorrect
1185 body hash calculated value.
1187 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1188 DKIM signature block, for verification. Although advised against by
1189 standards it is specifically not ruled illegal.
1191 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1193 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1194 missing a body hash (the bh= tag).
1196 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1197 It seems that HAProxy sends the Proxy Protocol information in clear and
1198 only then does a TLS startup, so do the same.
1200 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1201 TCP connections (such as for Spamd) unless the daemon successfully set
1202 Fast Open mode on its listening sockets. This fixes breakage seen on
1203 too-old kernels or those not configured for Fast Open, at the cost of
1204 requiring both directions being enabled for TFO, and TFO never being used
1205 by non-daemon-related Exim processes.
1207 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1208 endings, at least on the first header line. Try to canonify any that get
1209 past that check, despite the cost.
1211 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1212 now limited to an arbitrary five deep, while parsing addresses with the
1213 strip_excess_angle_brackets option enabled.
1215 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1216 instead leave the unprompted TLS handshake in socket buffer for the
1217 TLS library to consume.
1219 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1221 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1223 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1224 region in one operation, for speed, and this leaves assigned pointers
1225 dangling. Add checks run only under the testsuite which checks all
1226 variables at a store-reset and panics on a dangling pointer; add code
1227 explicitly nulling out all the variables discovered. Fixes one known
1228 bug: a transport crash, where a dangling pointer for $sending_ip_address
1229 originally assigned in a verify callout, is re-used.
1231 PP/06 Drop '.' from @INC in various Perl scripts.
1233 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1235 PP/08 Reduce a number of compilation warnings under clang; building with
1236 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1237 should be warning-free.
1239 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1241 HS/01 Fix portability problems introduced by PP/08 for platforms where
1242 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1244 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1245 chunk. This allows us to accept broken chunked messages. We need a more
1246 general solution here.
1248 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1249 already-broken messages in the queue.
1251 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1253 JH/19 Fix reference counting bug in routing-generated-address tracking.
1259 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1260 supports it and a size is available (ie. the sending peer gave us one).
1262 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1263 years of being deprecated). The replacements are the ACLs
1264 acl_smtp_mime and acl_not_smtp_mime.
1266 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1267 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1268 or even an in-clear connection were permitted. Now, if the host lookup
1269 was dnssec and dane was requested then the host is only used if the
1270 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1271 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1272 if one fails this test.
1273 This means that a poorly-configured remote DNS will make it incommunicado;
1274 but it protects against a DNS-interception attack on it.
1276 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1277 non-noisy when a race steals the message being considered.
1279 JH/05 If main configuration option tls_certificate is unset, generate a
1280 self-signed certificate for inbound TLS connections.
1282 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1283 in rewrites and routers.
1285 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1286 and logged a warning sing 4.83; now they are a configuration file error.
1288 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1289 (lacking @domain). Apply the same qualification processing as RCPT.
1291 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1293 JH/10 Support ${sha256:} applied to a string (as well as the previous
1296 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1297 a cutthrough deliver is pending, as we always want to make a connection.
1298 This also avoids re-routing the message when later placing the cutthrough
1299 connection after a verify cache hit.
1300 Do not update it with the verify result either.
1302 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1303 when routing results in more than one destination address.
1305 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1306 signing (which inhibits the cutthrough capability). Previously only
1307 the presence of an option was tested; now an expansion evaluating as
1308 empty is permissible (obviously it should depend only on data available
1309 when the cutthrough connection is made).
1311 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1312 the relevant preceding SMTP command did not note the pipelining mode.
1314 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1315 Previously they were not counted.
1317 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1318 as one having no matching records. Previously we deferred the message
1319 that needed the lookup.
1321 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1322 distinguished as "(=".
1324 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1325 for missing MX records. Previously it only worked for missing A records.
1327 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1329 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1330 after the data-go-ahead and data-ack. Patch from Jason Betts.
1332 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1333 even for a "none" policy. Patch from Tony Meyer.
1335 JH/22 Fix continued use of a connection for further deliveries. If a port was
1336 specified by a router, it must also match for the delivery to be
1339 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1340 When one of the recipients of a message was unsuitable for the connection
1341 (has no matching addresses), we lost track of needing to mark it
1342 deferred. As a result mail would be lost.
1344 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1346 JH/25 Decoding ACL controls is now done using a binary search; the source code
1347 takes up less space and should be simpler to maintain. Merge the ACL
1348 condition decode tables also, with similar effect.
1350 JH/26 Fix problem with one_time used on a redirect router which returned the
1351 parent address unchanged. A retry would see the parent address marked as
1352 delivered, so not attempt the (identical) child. As a result mail would
1355 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1356 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1357 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1360 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1361 Main config options check_{log,spool}_{inodes,space} are now
1362 100 inodes, 10MB unless set otherwise in the configuration.
1364 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1365 Previously it only applied to the main-section connection policy
1368 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1370 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1371 by me. Added RFC7919 DH primes as an alternative.
1373 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1374 are not in the system include path.
1376 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1377 GnuTLS, when a session startup failed (eg because the client disconnected)
1378 Exim did stdio operations after fclose. This was exposed by a recent
1379 change which nulled out the file handle after the fclose.
1381 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1382 signed directly by the cert-signing cert, rather than an intermediate
1383 OCSP-signing cert. This is the model used by LetsEncrypt.
1385 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1387 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1388 an incoming connection.
1390 HS/02 Bug 1802: Do not half-close the connection after sending a request
1393 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1394 fallback to "prime256v1".
1396 JH/34 SECURITY: Use proper copy of DATA command in error message.
1397 Could leak key material. Remotely exploitable. CVE-2016-9963.
1403 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1404 and 3.4.4 - once the server is enabled to respond to an OCSP request
1405 it does even when not requested, resulting in a stapling non-aware
1406 client dropping the TLS connection.
1408 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1409 support variable-length bit vectors. No functional change.
1411 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1412 The I= interface field on outgoing lines is now after the H= remote
1413 host field, same as incoming lines. There is a separate
1414 outgoing_interface log selector which allows you to disable the
1417 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1418 If not running log_selector +smtp_connection the mainlog would be held
1419 open indefinitely after a "too many connections" event, including to a
1420 deleted file after a log rotate. Leave the per net connection logging
1421 leaving it open for efficiency as that will be quickly detected by the
1422 check on the next write.
1424 HS/01 Bug 1671: Fix post transport crash.
1425 Processing the wait-<transport> messages could crash the delivery
1426 process if the message IDs didn't exist for some reason. When
1427 using 'split_spool_directory=yes' the construction of the spool
1428 file name failed already, exposing the same netto behaviour.
1430 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1431 mime_regex ACL conditions.
1433 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1434 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1435 to HELO, local diagnostic string.
1437 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1438 log line to debug. Even when configured with a tls authenticator many
1439 client connections are expected to not authenticate in this way, so
1440 an authenticate fail is not an error.
1442 HS/02 Add the Exim version string to the process info. This way exiwhat
1443 gives some more detail about the running daemon.
1445 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1446 matter for fast-change records such as DNSBLs.
1448 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1449 retry record, even if constant. There may be multiple transports with
1450 different interface settings and the retry behaviour needs to be kept
1453 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1455 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1457 JH/10 Bug 840: fix log_defer_output option of pipe transport
1459 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1460 hosts_randomize. This matters a lot when combined with mua_wrapper.
1462 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1463 ${quote_pgsql:<string>} operator.
1465 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1467 JH/14 Change header file name for freeradius-client. Relevant if compiling
1468 with Radius support; from the Gentoo tree and checked under Fedora.
1470 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1472 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1473 option transport_filter, meaning no filtering.
1475 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1477 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1478 defaults to "*" (all hosts). The variable is now available when not built
1479 with TLS, default unset, mainly to enable keeping the testsuite sane.
1480 If a server certificate is not supplied (via tls_certificate) an error is
1481 logged, and clients will find TLS connections fail on startup. Presumably
1482 they will retry in-clear.
1483 Packagers of Exim are strongly encouraged to create a server certificate
1484 at installation time.
1486 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1487 with the $config_file variable.
1489 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1490 in transport context, after the attempt, and per-recipient. The latter type
1491 is per host attempted. The event data is the error message, and the errno
1492 information encodes the lookup type (A vs. MX) used for the (first) host,
1493 and the trailing two digits of the smtp 4xx response.
1495 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1496 to write to mainlog (or rejectlog, paniclog) in the window between file
1497 creation and permissions/ownership being changed. Particularly affects
1498 installations where exicyclog is run as root, rather than exim user;
1499 result is that the running daemon panics and dies.
1501 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1503 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1504 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1505 "pri" and "weight". Note that the previous implicit priority given by the
1506 list order is no longer honoured.
1508 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1509 for DKIM processing.
1511 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1512 by defining SUPPORT_SOCKS.
1514 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1515 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1516 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1517 variables are renamed to proxy_{local,external}_{address,port}.
1519 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1520 for a build by defining SUPPORT_I18N
1522 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1523 of the query string, and make ${quote_redis:} do that quoting.
1525 JH/29 Move Events support from Experimental to mainline, enabled by default
1526 and removable for a build by defining DISABLE_EVENT.
1528 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1530 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1531 cached by the daemon.
1533 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1534 by defining LOOKUP_REDIS. The libhiredis library is required.
1536 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1537 keys are given for lookup.
1539 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1540 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1541 only supported when built with TLS support. The PolarSSL SHA routines
1542 are still used when the TLS library is too old for convenient support.
1544 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1545 openssl_options), for security. OpenSSL forces this from version 1.1.0
1546 server-side so match that on older versions.
1548 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1549 allocation for $value could be released as the expansion processing
1550 concluded, but leaving the global pointer active for it.
1552 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1553 and to use the domains and local_parts ACL conditions.
1555 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1556 incorrectly not doubled on cutthrough transmission, hence seen as a
1557 body-termination at the receiving system - resulting in truncated mails.
1558 Commonly the sender saw a TCP-level error, and retransmitted the message
1559 via the normal store-and-forward channel. This could result in duplicates
1560 received - but deduplicating mailstores were liable to retain only the
1561 initial truncated version.
1563 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1565 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1567 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1568 we're in there, support oversigning also; bug 1309.
1570 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1572 HS/04 Add support for keep_environment and add_environment options.
1574 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1575 either intentional arithmetic overflow during PRNG, or testing config-
1578 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1579 delivery resulted in actual delivery. Cancel cutthrough before DATA
1582 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1583 rejects a recipient: pass the reject to the originator.
1585 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1586 Many were false-positives and ignorable, but it's worth fixing the
1589 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1590 for the new environment-manipulation done at startup. Move the routines
1591 from being local to tls.c to being global via the os.c file.
1593 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1594 an extract embedded as result-arg for a map, the first arg for extract
1595 is unavailable so we cannot tell if this is a numbered or keyed
1596 extraction. Accept either.
1602 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1605 JH/02 The smtp transport option "multi_domain" is now expanded.
1607 JH/03 The smtp transport now requests PRDR by default, if the server offers
1610 JH/04 Certificate name checking on server certificates, when exim is a client,
1611 is now done by default. The transport option tls_verify_cert_hostnames
1612 can be used to disable this per-host. The build option
1613 EXPERIMENTAL_CERTNAMES is withdrawn.
1615 JH/05 The value of the tls_verify_certificates smtp transport and main options
1616 default to the word "system" to access the system default CA bundle.
1617 For GnuTLS, only version 3.0.20 or later.
1619 JH/06 Verification of the server certificate for a TLS connection is now tried
1620 (but not required) by default. The verification status is now logged by
1621 default, for both outbound TLS and client-certificate supplying inbound
1624 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1627 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1628 Status Notification (bounce) messages are now MIME format per RFC 3464.
1629 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1630 under the control of the dsn_advertise_hosts option, and routers may
1631 have a dsn_lasthop option.
1633 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1634 default, modifiable by a malware= option. The list separator for
1635 the options can now be changed in the usual way. Bug 68.
1637 JH/10 The smtp_receive_timeout main option is now expanded before use.
1639 JH/11 The incoming_interface log option now also enables logging of the
1640 local interface on delivery outgoing connections.
1642 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1643 if the interface and destination host and port all match.
1645 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1648 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1649 Patch from Andrew Lewis.
1651 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1652 now supports optional time-restrictions, weighting, and priority
1653 modifiers per server. Patch originally by <rommer@active.by>.
1655 JH/16 The spamd_address main option now supports a mixed list of local
1656 and remote servers. Remote servers can be IPv6 addresses, and
1657 specify a port-range.
1659 JH/17 Bug 68: The spamd_address main option now supports an optional
1660 timeout value per server.
1662 JH/18 Bug 1581: Router and transport options headers_add/remove can
1663 now have the list separator specified.
1665 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1668 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1671 JH/21 Support for the A6 type of dns record is withdrawn.
1673 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1674 rather than the verbs used.
1676 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1677 from 255 to 1024 chars.
1679 JH/24 Verification callouts now attempt to use TLS by default.
1681 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1682 are generic router options now. The defaults didn't change.
1684 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1685 Original patch from Alexander Shikoff, worked over by JH.
1687 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1688 files if ZCAT_COMMAND is not executable.
1690 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1692 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1694 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1695 Normally benign, it bites when the pair was led to by a CNAME;
1696 modern usage is to not canonicalize the domain to a CNAME target
1697 (and we were inconsistent anyway for A-only vs AAAA+A).
1699 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1701 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1702 when evaluating $sender_host_dnssec.
1704 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1705 $sender_helo_dnssec variable.
1707 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1709 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1711 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1713 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1714 documented as working, but never had. Support all but $spam_report.
1716 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1717 added for tls authenticator.
1719 HS/03 Add perl_taintmode main config option
1725 TL/01 When running the test suite, the README says that variables such as
1726 no_msglog_check are global and can be placed anywhere in a specific
1727 test's script, however it was observed that placement needed to be near
1728 the beginning for it to behave that way. Changed the runtest perl
1729 script to read through the entire script once to detect and set these
1730 variables, reset to the beginning of the script, and then run through
1731 the script parsing/test process like normal.
1733 TL/02 The BSD's have an arc4random API. One of the functions to induce
1734 adding randomness was arc4random_stir(), but it has been removed in
1735 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1736 function when detected.
1738 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1739 cause callback expansion.
1741 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1742 syntax errors in an expansion can be treated as a string instead of
1743 logging or causing an error, due to the internal use of bool_lax
1744 instead of bool when processing it.
1746 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1747 server certificates when making smtp deliveries.
1749 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1751 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1753 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1755 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1756 Merged patch from Sebastian Wiedenroth.
1758 JH/05 Fix results-pipe from transport process. Several recipients, combined
1759 with certificate use, exposed issues where response data items split
1760 over buffer boundaries were not parsed properly. This eventually
1761 resulted in duplicates being sent. This issue only became common enough
1762 to notice due to the introduction of connection certificate information,
1763 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1765 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1766 size buffer was used, resulting in syntax errors when an expansion
1769 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1770 version 3.3.6 or later.
1772 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1773 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1774 both become "event_action", the variables become $event_name, $event_data
1775 and $event_defer_errno. There is a new variable $verify_mode, usable in
1776 routers, transports and related events. The tls:cert event is now also
1777 raised for inbound connections, if the main configuration event_action
1780 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1781 early OCSP support, but no stapling (appears to be less than 1.0.0).
1783 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1784 server certificate names available under the smtp transport option
1785 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1788 JH/10 Time-related extraction expansions from certificates now use the main
1789 option "timezone" setting for output formatting, and are consistent
1790 between OpenSSL and GnuTLS compilations. Bug 1541.
1792 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1793 encoded parameter in the incoming message. Bug 1558.
1795 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1796 include certificate info, eximon was claiming there were spoolfile
1799 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1801 JH/14 Log delivery-related information more consistently, using the sequence
1802 "H=<name> [<ip>]" wherever possible.
1804 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1805 are problematic for Debian distribution, omit them from the release
1808 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1810 JH/16 Fix string representation of time values on 64bit time_t architectures.
1813 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1814 output list separator was used.
1819 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1820 checkers that were complaining about end of non-void function with no
1823 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1824 This was a regression introduced in 4.83 by another bugfix.
1826 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1828 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1829 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1835 TF/01 Correctly close the server side of TLS when forking for delivery.
1837 When a message was received over SMTP with TLS, Exim failed to clear up
1838 the incoming connection properly after forking off the child process to
1839 deliver the message. In some situations the subsequent outgoing
1840 delivery connection happened to have the same fd number as the incoming
1841 connection previously had. Exim would try to use TLS and fail, logging
1842 a "Bad file descriptor" error.
1844 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1845 utilities have not been installed.
1847 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1848 temporary space as the ACL may create new global variables.
1850 TL/01 LDAP support uses per connection or global context settings, depending
1851 upon the detected version of the libraries at build time.
1853 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1854 to extract and use the src ip:port in logging and expansions as if it
1855 were a direct connection from the outside internet. PPv2 support was
1856 updated based on HAProxy spec change in May 2014.
1858 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1860 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1861 Properly escape header and check for NULL return.
1863 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1866 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1868 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1869 characters in header names, implemented as a verify condition.
1870 Contributed by Michael Fischer v. Mollard.
1872 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1873 results permerror and temperror. Previous values are deprecated but
1874 still accepted. In a future release, err_perm and err_temp will be
1875 completely removed, which will be a backward incompatibility if the
1876 ACL tests for either of these two old results. Patch contributed by
1877 user bes-internal on the mailing list.
1879 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1881 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1882 selectors, in both main and reject logs.
1884 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1887 JH/07 Add malware type "sock" for talking to simple daemon.
1889 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1891 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1892 routers/transports under cutthrough routing.
1894 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1895 numbers. Touch up "bool" conditional to keep the same definition.
1897 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1899 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1901 JH/12 Expand items in router/transport headers_add or headers_remove lists
1902 individually rather than the list as a whole. Bug 1452.
1904 Required for reasonable handling of multiple headers_ options when
1905 they may be empty; requires that headers_remove items with embedded
1906 colons must have them doubled (or the list-separator changed).
1908 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1909 view the policy declared in the DMARC record. Currently, $dmarc_status
1910 is a combined value of both the record presence and the result of the
1913 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1915 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1916 dnslookup router and the smtp transport (applying to the forward
1919 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1920 of ldap servers used for a specific lookup. Patch provided by Heiko
1923 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1924 New variable $lookup_dnssec_authenticated for observability.
1926 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1927 Patch submitted by Lars Timman.
1929 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1931 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1932 Requires trusted mode and valid format message id, aborts otherwise.
1933 Patch contributed by Heiko Schlichting.
1935 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1936 certextract with support for various fields. Bug 1358.
1938 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1939 is requested by default, modifiable by smtp transport option
1942 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1943 operate on certificate variables to give certificate fingerprints
1944 Also new ${sha256:cert_variable}.
1946 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1948 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1949 Christian Aistleitner.
1951 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1953 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1954 file. Patch from Wolfgang Breyha.
1956 JH/25 Expand the coverage of the delivery $host and $host_address to
1957 client authenticators run in verify callout. Bug 1476.
1959 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1960 align with daemon_smtp_ports. Bug 72.
1962 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1963 support and error reporting did not work properly.
1965 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1966 and is readable. Patch from Andrew Colin Kissa.
1968 TL/14 Enhance documentation of ${run expansion and how it parses the
1969 commandline after expansion, particularly in the case when an
1970 unquoted variable expansion results in an empty value.
1972 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1974 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1975 by renaming away from it. Addresses GH issue 15, reported by
1978 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1980 TL/15 SECURITY: prevent double expansion in math comparison functions
1981 (can expand unsanitized data). Not remotely exploitable.
1988 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1990 PP/02 Make -n do something, by making it not do something.
1991 When combined with -bP, the name of an option is not output.
1993 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1996 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1997 $sender_host_name and config options to manage this, and basic check
2000 PP/05 DSCP support for outbound connections and control modifier for inbound.
2002 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2003 (Only plugin which currently uses this is kerberos4, which nobody should
2004 be using, but we should make it available and other future plugins might
2005 conceivably use it, even though it would break NAT; stuff *should* be
2006 using channel bindings instead).
2008 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2009 name; added for Sendmail compatibility; requires admin caller.
2010 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2011 just ignore it); requires trusted caller.
2012 Also parse but ignore: -Ac -Am -X<logfile>
2015 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2017 TL/02 Add +smtp_confirmation as a default logging option.
2019 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2020 Patch by Magnus Holmgren from 2007-02-20.
2022 TL/04 Bugzilla 1281 - Spec typo.
2023 Bugzilla 1283 - Spec typo.
2024 Bugzilla 1290 - Spec grammar fixes.
2026 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2028 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2030 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2031 mailing list by by Dmitry Isaikin.
2033 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2035 JH/02 Support "G" suffix to numbers in ${if comparisons.
2037 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2039 NM/01 Bugzilla 1197 - Spec typo
2040 Bugzilla 1196 - Spec examples corrections
2042 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2044 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2045 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2048 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2049 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2051 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2052 "acl {{name}{arg}...}", and optional args on acl condition
2055 JH/05 Permit multiple router/transport headers_add/remove lines.
2057 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2059 JH/07 Avoid using a waiting database for a single-message-only transport.
2060 Performance patch from Paul Fisher. Bugzilla 1262.
2062 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2065 JH/09 Add $headers_added variable, with content from use of ACL modifier
2066 add_header (but not yet added to the message). Bugzilla 199.
2068 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2069 Pulled from Bugzilla 817 by Wolfgang Breyha.
2071 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2073 (nb: this is the same fix as in Exim 4.80.1)
2075 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2078 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2081 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2083 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2084 advertise SMTP AUTH mechanism to us, instead of a generic
2085 protocol violation error. Also, make Exim more robust to bad
2086 data from the Dovecot auth socket.
2088 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2090 When a queue runner is handling a message, Exim first routes the
2091 recipient addresses, during which it prunes them based on the retry
2092 hints database. After that it attempts to deliver the message to
2093 any remaining recipients. It then updates the hints database using
2096 So if a recipient address works intermittently, it can get repeatedly
2097 deferred at routing time. The retry hints record remains fresh so the
2098 address never reaches the final cutoff time.
2100 This is a fairly common occurrence when a user is bumping up against
2101 their storage quota. Exim had some logic in its local delivery code
2102 to deal with this. However it did not apply to per-recipient defers
2103 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2105 This change adds a proper retry rule check during routing so that the
2106 final cutoff time is checked against the message's age. We only do
2107 this check if there is an address retry record and there is not a
2108 domain retry record; this implies that previous attempts to handle
2109 the address had the retry_use_local_parts option turned on. We use
2110 this as an approximation for the destination being like a local
2111 delivery, as in LMTP.
2113 I suspect this new check makes the old local delivery cutoff check
2114 redundant, but I have not verified this so I left the code in place.
2116 TF/02 Correct gecos expansion when From: is a prefix of the username.
2118 Test 0254 submits a message to Exim with the header
2122 When I ran the test suite under the user fanf2, Exim expanded
2123 the header to contain my full name, whereas it should have added
2124 a Resent-Sender: header. It erroneously treats any prefix of the
2125 username as equal to the username.
2127 This change corrects that bug.
2129 GF/01 DCC debug and logging tidyup
2130 Error conditions log to paniclog rather than rejectlog.
2131 Debug lines prefixed by "DCC: " to remove any ambiguity.
2133 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2135 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2136 Bug spotted by Jeremy Harris; was flawed since initial commit.
2137 Would have resulted in OCSP responses post-SNI triggering an Exim
2138 NULL dereference and crash.
2140 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2142 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2143 Bug detection, analysis and fix by Samuel Thibault.
2144 Bugzilla 1331, Debian bug #698092.
2146 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2148 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2149 Server implementation by Todd Lyons, client by JH.
2150 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2151 config variable "prdr_enable" controls whether the server
2152 advertises the facility. If the client requests PRDR a new
2153 acl_data_smtp_prdr ACL is called once for each recipient, after
2154 the body content is received and before the acl_smtp_data ACL.
2155 The client is controlled by both of: a hosts_try_prdr option
2156 on the smtp transport, and the server advertisement.
2157 Default client logging of deliveries and rejections involving
2158 PRDR are flagged with the string "PRDR".
2160 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2161 fclose(). Diagnosis by Todd Lyons.
2163 PP/17 Update configure.default to handle IPv6 localhost better.
2164 Patch by Alain Williams (plus minor tweaks).
2167 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2168 This is now consistent with GnuTLS, and is now documented: the
2169 previous undocumented portable approach to treating the option as
2170 unset was to force an expansion failure. That still works, and
2171 an empty string is now equivalent.
2173 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2174 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2175 not performing validation itself.
2177 PP/20 Added force_command boolean option to pipe transport.
2178 Patch from Nick Koston, of cPanel Inc.
2180 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2183 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2185 PP/21 Fix eximon continuous updating with timestamped log-files.
2186 Broken in a format-string cleanup in 4.80, missed when I repaired the
2187 other false fix of the same issue.
2188 Report and fix from Heiko Schlichting.
2191 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2192 Report from Prashanth Katuri.
2194 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2195 It's SecureTransport, so affects any MacOS clients which use the
2196 system-integrated TLS libraries, including email clients.
2198 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2199 using a MIME ACL for non-SMTP local injection.
2200 Report and assistance in diagnosis by Warren Baker.
2202 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2204 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2206 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2207 last id that failed so it may be referenced in subsequent ACL's.
2209 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2212 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2213 ldap library initialization, allowing self-signed CA's to be
2214 used. Also properly sets require_cert option later in code by
2215 using NULL (global ldap config) instead of ldap handle (per
2216 session). Bug diagnosis and testing by alxgomz.
2218 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2219 the src/util/ subdirectory.
2221 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2222 renamed to Transport Post Delivery Action by Jeremy Harris, as
2225 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2226 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2227 redis_servers = needs to be configured which will be used by the redis
2228 lookup. Patch from Warren Baker, of The Packet Hub.
2230 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2232 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2233 hostname or reverse DNS when processing a host list. Used suggestions
2234 from multiple comments on this bug.
2236 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2238 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2239 Missed a few lines, added it to make the runtest require no keyboard
2242 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2243 contains upper case chars. Make router use caseful_local_part.
2245 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2246 support when GnuTLS has been built with p11-kit.
2252 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2254 This, or similar/improved, will also be change PP/11 of 4.82.
2260 PP/01 Handle short writes when writing local log-files.
2261 In practice, only affects FreeBSD (8 onwards).
2262 Bugzilla 1053, with thanks to Dmitry Isaikin.
2264 NM/01 Bugzilla 949 - Documentation tweak
2266 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2269 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2271 PP/02 Implemented gsasl authenticator.
2273 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2275 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2276 `pkg-config foo` for cflags/libs.
2278 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2279 with rest of GSASL and with heimdal_gssapi.
2281 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2282 `pkg-config foo` for cflags/libs for the TLS implementation.
2284 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2285 properties get this fed in as external SSF. A number of robustness
2286 and debugging improvements to the cyrus_sasl authenticator.
2288 PP/08 cyrus_sasl server now expands the server_realm option.
2290 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2291 Patch by Jeremy Harris.
2293 PP/10 Added dbmjz lookup type.
2295 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2297 PP/12 MAIL args handles TAB as well as SP, for better interop with
2298 non-compliant senders.
2299 Analysis and variant patch by Todd Lyons.
2301 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2302 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2303 Patch from Dirk Mueller <dmueller@suse.com>
2305 PP/13 tls_peerdn now print-escaped for spool files.
2306 Observed some $tls_peerdn in wild which contained \n, which resulted
2307 in spool file corruption.
2309 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2310 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2311 or write after TLS renegotiation, which otherwise led to messages
2314 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2315 as a tracking header (ie: a signed header comes before the signature).
2316 Patch from Wolfgang Breyha.
2318 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2319 comma-sep list; embedded commas doubled.
2321 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2323 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2325 Report and patch from Dmitry Banschikov.
2327 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2328 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2329 failures. We appear to now support TLS1.1+ with Exim.
2331 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2332 lets Exim select keys and certificates based upon TLS SNI from client.
2333 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2334 before an outbound SMTP session. New log_selector, +tls_sni.
2336 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2337 NULL dereference. Report and patch from Alun Jones.
2339 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2340 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2341 Not seeing resolver debug output on NetBSD, but suspect this is a
2342 resolver implementation change.
2344 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2345 Left warnings. Added "eximon gdb" invocation mode.
2347 PP/21 Defaulting "accept_8bitmime" to true, not false.
2349 PP/22 Added -bw for inetd wait mode support.
2351 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2352 locate the relevant includes and libraries. Made this the default.
2354 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2355 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2357 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2358 This may cause build issues on older platforms.
2360 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2361 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2362 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2363 Added SNI support via GnuTLS too.
2364 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2366 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2368 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2369 Applied second patch from Janne, implementing suggestion to default
2370 multiple-strings-in-record handling to match SPF spec.
2372 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2374 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2375 read-only, out of scope).
2376 Patch from Wolfgang Breyha, report from Stuart Northfield.
2378 PP/29 Fix three issues highlighted by clang analyser static analysis.
2379 Only crash-plausible issue would require the Cambridge-specific
2380 iplookup router and a misconfiguration.
2381 Report from Marcin Mirosław.
2383 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2385 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2386 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2387 As part of this, removing so much warning spew let me fix some minor
2388 real issues in debug logging.
2390 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2391 assignment on my part. Fixed.
2393 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2394 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2395 Janne Snabb (who went above and beyond: thank you).
2397 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2398 string otherwise requires a connection and a bunch more work and it's
2399 relatively easy to get wrong. Should also expose TLS library linkage
2402 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2403 64-bit ${eval} (JH/03).
2405 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2406 GNU libc to support some of the 64-bit stuff, should not lead to
2407 conflicts. Defined before os.h is pulled in, so if a given platform
2408 needs to override this, it can.
2410 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2411 protection layer was required, which is not implemented.
2412 Bugzilla 1254, patch from Wolfgang Breyha.
2414 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2415 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2416 tls_dhparam take prime identifiers. Also unbreak combination of
2417 OpenSSL+DH_params+TLSSNI.
2419 PP/39 Disable SSLv2 by default in OpenSSL support.
2425 PP/01 Solaris build fix for Oracle's LDAP libraries.
2426 Bugzilla 1109, patch from Stephen Usher.
2428 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2430 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2433 TF/02 Fix a couple more cases where we did not log the error message
2434 when unlink() failed. See also change 4.74-TF/03.
2436 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2437 lock up or crash if it happened to be inside a call to libc when it
2438 got a SIGUSR1 from exiwhat.
2440 The SIGUSR1 handler appends the current process status to the process
2441 log which is later printed by exiwhat. It used to use the general
2442 purpose logging code to do this, but several functions it calls are
2443 not safe for signals.
2445 The new output code in the SIGUSR1 handler is specific to the process
2446 log, and simple enough that it's easy to inspect for signal safety.
2447 Removing some special cases also simplifies the general logging code.
2448 Removing the spurious timestamps from the process log simplifies
2451 TF/04 Improved ratelimit ACL condition.
2453 The /noupdate option has been deprecated in favour of /readonly which
2454 has clearer semantics. The /leaky, /strict, and /readonly update modes
2455 are mutually exclusive. The update mode is no longer included in the
2456 database key; it just determines when the database is updated. (This
2457 means that when you upgrade Exim will forget old rate measurements.)
2459 Exim now checks that the per_* options are used with an update mode that
2460 makes sense for the current ACL. For example, when Exim is processing a
2461 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2462 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2463 must specify per_mail/readonly. If you omit the update mode it defaults to
2464 /leaky where that makes sense (as before) or /readonly where required.
2466 The /noupdate option is now undocumented but still supported for
2467 backwards compatibility. It is equivalent to /readonly except that in
2468 ACLs where /readonly is required you may specify /leaky/noupdate or
2469 /strict/noupdate which are treated the same as /readonly.
2471 A useful new feature is the /count= option. This is a generalization
2472 of the per_byte option, so that you can measure the throughput of other
2473 aggregate values. For example, the per_byte option is now equivalent
2474 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2476 The per_rcpt option has been generalized using the /count= mechanism
2477 (though it's more complicated than the per_byte equivalence). When it is
2478 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2479 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2480 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2481 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2482 Note that using per_rcpt with a non-readonly update mode in more than
2483 one ACL will cause the recipients to be double-counted. (The per_mail
2484 and per_byte options don't have this problem.)
2486 The handling of very low rates has changed slightly. If the computed rate
2487 is less than the event's count (usually one) then this event is the first
2488 after a long gap. In this case the rate is set to the same as this event's
2489 count, so that the first message of a spam run is counted properly.
2491 The major new feature is a mechanism for counting the rate of unique
2492 events. The new per_addr option counts the number of different
2493 recipients that someone has sent messages to in the last time period. It
2494 behaves like per_rcpt if all the recipient addresses are different, but
2495 duplicate recipient addresses do not increase the measured rate. Like
2496 the /count= option this is a general mechanism, so the per_addr option
2497 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2498 example, measure the rate that a client uses different sender addresses
2499 with the options per_mail/unique=$sender_address. There are further
2500 details in the main documentation.
2502 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2504 TF/06 Removed a few PCRE remnants.
2506 TF/07 Automatically extract Exim's version number from tags in the git
2507 repository when doing development or release builds.
2509 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2510 Bugzilla 879. Patch from Paul Fisher.
2512 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2513 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2516 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2518 PP/05 Variable $av_failed, true if the AV scanner deferred.
2519 Bugzilla 1078. Patch from John Horne.
2521 PP/06 Stop make process more reliably on build failure.
2522 Bugzilla 1087. Patch from Heiko Schlittermann.
2524 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2525 Bugzilla 1089. Patch from Heiko Schlittermann.
2527 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2528 Bugzilla 1131. Patch from Holger Weiß.
2530 PP/09 Handle IPv6 addresses with SPF.
2531 Bugzilla 860. Patch from Wolfgang Breyha.
2533 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2535 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2538 PP/11 match_* no longer expand right-hand-side by default.
2539 New compile-time build option, EXPAND_LISTMATCH_RHS.
2540 New expansion conditions, "inlist", "inlisti".
2542 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2544 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2546 PP/14 fix log_write() format string regression from TF/03.
2547 Bugzilla 1152. Patch from Dmitry Isaikin.
2553 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2555 PP/02 Harmonised TLS library version reporting; only show if debugging.
2556 Layout now matches that introduced for other libraries in 4.74 PP/03.
2558 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2560 PP/04 New "dns_use_edns0" global option.
2562 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2565 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2566 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2568 TK/01 Updated PolarSSL code to 0.14.2.
2569 Bugzilla 1097. Patch from Andreas Metzler.
2571 PP/07 Catch divide-by-zero in ${eval:...}.
2572 Fixes bugzilla 1102.
2574 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2577 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2578 format-string attack -- SECURITY: remote arbitrary code execution.
2580 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2581 time unintentionally subject to list matching rules, letting the header
2582 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2583 arbitrary string expansion). This allowed for information disclosure.
2585 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2586 INT_MIN/-1 -- value coerced to INT_MAX.
2592 NM/01 Workaround for PCRE version dependency in version reporting
2595 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2596 This fixes portability to compilers other than gcc, notably
2597 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2599 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2600 makefiles for portability to HP-UX and POSIX correctness.
2602 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2603 Also via indented variable definition in the Makefile.
2604 (Debugging by Oliver Heesakkers).
2606 PP/02 Restore caching of spamd results with expanded spamd_address.
2607 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2609 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2610 Improves build reliability. Fix from: Frank Elsner
2612 NM/02 Fix wide character breakage in the rfc2047 coding
2613 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2615 NM/03 Allow underscore in dnslist lookups
2616 Fixes bug 1026. Patch from Graeme Fowler
2618 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2619 Code patches from Adam Ciarcinski of NetBSD.
2621 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2624 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2625 is logged, to avoid truncation. Patch from John Horne.
2627 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2628 Patch from Jakob Hirsch.
2630 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2631 SQL string expansion failure details.
2632 Patch from Andrey Oktyabrski.
2634 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2635 Patch from Simon Arlott.
2637 PP/09 New lookups functionality failed to compile on old gcc which rejects
2638 extern declarations in function scope.
2639 Patch from Oliver Fleischmann
2641 PP/10 Use sig_atomic_t for flags set from signal handlers.
2642 Check getgroups() return and improve debugging.
2643 Fixed developed for diagnosis in bug 927 (which turned out to be
2646 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2647 Patch from Mark Zealey.
2649 PP/12 Bugzilla 1056: Improved spamd server selection.
2650 Patch from Mark Zealey.
2652 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2653 Based on patch from Heiko Schlittermann.
2655 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2656 Patch from Uwe Doering, sign-off by Michael Haardt.
2658 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2659 variable declaration deep within a block. Bug and patch from
2662 PP/15 lookups-Makefile IRIX compatibility coercion.
2664 PP/16 Make DISABLE_DKIM build knob functional.
2666 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2667 Patch by Simon Arlott
2669 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2670 variable argument macros. Our copy now differs from upstream.
2676 TF/01 Failure to get a lock on a hints database can have serious
2677 consequences so log it to the panic log.
2679 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2680 controlled using the smtp_confirmation log selector.
2682 TF/03 Include the error message when we fail to unlink a spool file.
2684 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2685 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2686 for maintaining out-of-tree patches for some time.
2688 PP/01 Bugzilla 139: Documentation and portability issues.
2689 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2690 Handle per-OS dynamic-module compilation flags.
2692 PP/02 Let /dev/null have normal permissions.
2693 The 4.73 fixes were a little too stringent and complained about the
2694 permissions on /dev/null. Exempt it from some checks.
2695 Reported by Andreas M. Kirchwitz.
2697 PP/03 Report version information for many libraries, including
2698 Exim version information for dynamically loaded libraries. Created
2699 version.h, now support a version extension string for distributors
2700 who patch heavily. Dynamic module ABI change.
2702 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2703 privilege escalation vulnerability whereby the Exim run-time user
2704 can cause root to append content of the attacker's choosing to
2707 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2710 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2711 If dropping privileges for untrusted macros, we disabled normal logging
2712 on the basis that it would fail; for the Exim run-time user, this is not
2713 the case, and it resulted in successful deliveries going unlogged.
2714 Fixed. Reported by Andreas Metzler.
2720 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2721 only prepend for the Resent-* case. Fixes regression introduced in
2722 Exim 4.70 by NM/22 for Bugzilla 607.
2724 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2725 increasing numbers of administrators be bitten by this.
2727 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2729 PP/03 Bugzilla 994: added openssl_options main configuration option.
2731 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2733 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2735 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2737 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2738 without a peer certificate, leading to a segfault because of an
2739 assumption that peers always have certificates. Be a little more
2740 paranoid. Problem reported by Martin Tscholak.
2742 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2743 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2744 NB: ClamAV planning to remove STREAM in "middle of 2010".
2745 CL also introduces -bmalware, various -d+acl logging additions and
2746 more caution in buffer sizes.
2748 PP/09 Implemented reverse_ip expansion operator.
2750 PP/10 Bugzilla 937: provide a "debug" ACL control.
2752 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2754 PP/12 Bugzilla 973: Implement --version.
2756 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2758 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2760 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2762 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2763 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2764 ignore trailing whitespace.
2766 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2768 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2771 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2772 Notification from Dr Andrew Aitchison.
2774 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2775 ExtendedDetectionInfo response format.
2776 Notification from John Horne.
2778 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2781 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2782 XSL and documented dependency on system catalogs, with examples of how
2785 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2788 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2789 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2790 configuration file which is writeable by the Exim user or group.
2792 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2793 of configuration files to cover files specified with the -C option if
2794 they are going to be used with root privileges, not just the default
2797 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2798 option (effectively making it always true).
2800 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2801 files to be used while preserving root privileges.
2803 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2804 that rogue child processes cannot use them.
2806 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2807 run-time user, instead of root.
2809 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2810 Exim run-time user without dropping privileges.
2812 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2813 result string, instead of calling string_vformat() twice with the same
2816 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2817 for other users. Others should always drop root privileges if they use
2818 -C on the command line, even for a whitelisted configure file.
2820 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2822 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2828 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2829 $data_path, and $header_path variables; fixed documentation bugs and
2832 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2833 exipick to access non-standard spools, including the "frozen" queue
2836 NM/01 Bugzilla 965: Support mysql stored procedures.
2837 Patch from Alain Williams
2839 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2841 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2842 Patch from Andreas Metzler
2844 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2845 Patch from Kirill Miazine
2847 NM/05 Bugzilla 671: Added umask to procmail example.
2849 JJ/03 installed exipick 20100323.0, fixing doc bug
2851 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2852 directory. Notification and patch from Dan Rosenberg.
2854 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2856 TK/02 Improve log output when DKIM signing operation fails.
2858 MH/01 Treat the transport option dkim_domain as a colon separated
2859 list, not as a single string, and sign the message with each element,
2860 omitting multiple occurences of the same signer.
2862 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2863 Bugzilla 985, 986. Patch by Simon Arlott
2865 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2866 Patch by Simon Arlott
2868 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2869 MBX locking. Notification from Dan Rosenberg.
2875 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2877 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2879 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2881 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2883 NM/04 Bugzilla 915: Flag broken perl installation during build.
2889 TK/01 Added patch by Johannes Berg that expands the main option
2890 "spamd_address" if it starts with a dollar sign.
2892 TK/02 Write list of recipients to X-Envelope-Sender header when building
2893 the mbox-format spool file for content scanning (suggested by Jakob
2896 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2897 (http://www.dcc-servers.net/) support via dccifd. Activated by
2898 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2900 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2901 by Mark Daniel Reidel <mr@df.eu>.
2903 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2904 When building exim an external PCRE library is now needed -
2905 PCRE is a system library on the majority of modern systems.
2906 See entry on PCRE_LIBS in EDITME file.
2908 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2909 conversation. Added nologin parameter to request.
2910 Patch contributed by Kirill Miazine.
2912 TF/01 Do not log submission mode rewrites if they do not change the address.
2914 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2916 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2917 log files in place. Contributed by Roberto Lima.
2919 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2921 TF/03 Bugzilla 615: When checking the local_parts router precondition
2922 after a local_part_suffix or local_part_prefix option, Exim now
2923 does not use the address's named list lookup cache, since this
2924 contains cached lookups for the whole local part.
2926 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2927 Robert Millan. Documentation is in experimental-spec.txt.
2929 TF/04 Bugzilla 668: Fix parallel build (make -j).
2931 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2933 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2934 Patch provided by Jan Srzednicki.
2936 TF/05 Leading white space used to be stripped from $spam_report which
2937 wrecked the formatting. Now it is preserved.
2939 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2940 that they are available at delivery time.
2942 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2944 TF/08 TLS error reporting now respects the incoming_interface and
2945 incoming_port log selectors.
2947 TF/09 Produce a more useful error message if an SMTP transport's hosts
2948 setting expands to an empty string.
2950 NM/06 Bugzilla 744: EXPN did not work under TLS.
2951 Patch provided by Phil Pennock.
2953 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2954 Patch provided by Richard Godbee.
2956 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2957 acl_smtp_notquit, added index entry.
2959 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2960 Patch provided by Eugene Bujak.
2962 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2963 accept(). Patch provided by Maxim Dounin.
2965 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2966 Patch provided by Phil Pennock.
2968 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2970 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2971 Patch provided by Brad "anomie" Jorsch.
2973 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2974 Patch provided by Dean Brooks.
2976 TK/05 Add native DKIM support (does not depend on external libraries).
2978 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2979 Patch provided by Graeme Fowler.
2981 NM/16 Bugzilla 851: Documentation example syntax fix.
2983 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2985 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2988 NM/19 Bugzilla 745: TLS version reporting.
2989 Patch provided by Phil Pennock.
2991 NM/20 Bugzilla 167: bool: condition support.
2992 Patch provided by Phil Pennock.
2994 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2995 clients. Patch provided by Phil Pennock.
2997 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2998 Patch provided by Brad "anomie" Jorsch.
3000 NM/23 Bugzilla 687: Fix misparses in eximstats.
3001 Patch provided by Heiko Schlittermann.
3003 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3004 Patch provided by Heiko Schlittermann.
3006 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3007 plus update to original patch.
3009 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3011 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3012 Patch provided by David Brownlee.
3014 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3016 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3018 NM/30 Bugzilla 888: TLS documentation bugfixes.
3020 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3022 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3023 Unlike the original bugzilla I have changed all shell scripts in src tree.
3025 NM/33 Bugzilla 898: Transport filter timeout fix.
3026 Patch by Todd Rinaldo.
3028 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3029 Patch by Serge Demonchaux.
3031 NM/35 Bugzilla 39: Base64 decode bug fixes.
3032 Patch by Jakob Hirsch.
3034 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3036 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3038 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3040 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3046 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3047 ALT-N's libdkim that I have put here:
3048 http://duncanthrax.net/exim-experimental/
3050 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3051 were called 'true' and it seems that C99 defines that as a reserved
3052 keyword to be used with 'bool' variable types. That means you could
3053 not include C99-style headers which use bools without triggering
3054 build errors in sieve.c.
3056 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3057 as mailq or other aliases. Changed the --help handling significantly
3058 to do whats expected. exim_usage() emits usage/help information.
3060 SC/01 Added the -bylocaldomain option to eximstats.
3062 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3064 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3066 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3072 PH/01 Another patch from the Sieve maintainer.
3074 PH/02 When an IPv6 address is converted to a string for single-key lookup
3075 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3076 dots are used instead of colons so that keys in lsearch files need not
3077 contain colons. This was done some time before quoting was made available
3078 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3079 (notated using the quote facility) so as to distinguish them from IPv4
3080 keys. This meant that lookups for IP addresses in host lists did not work
3081 for iplsearch lookups.
3083 This has been fixed by arranging for IPv6 addresses to be expressed with
3084 colons if the lookup type is iplsearch. This is not incompatible, because
3085 previously such lookups could never work.
3087 The situation is now rather anomalous, since one *can* have colons in
3088 ordinary lsearch keys. However, making the change in all cases is
3089 incompatible and would probably break a number of configurations.
3091 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3094 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3095 conversion specification without a maximum field width, thereby enabling
3096 a rogue spamd server to cause a buffer overflow. While nobody in their
3097 right mind would setup Exim to query an untrusted spamd server, an
3098 attacker that gains access to a server running spamd could potentially
3099 exploit this vulnerability to run arbitrary code as the Exim user.
3101 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3102 $primary_hostname instead of what libspf2 thinks the hosts name is.
3104 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3105 a directory entry by the name of the lookup key. Previously, if a
3106 symlink pointed to a non-existing file or a file in a directory that
3107 Exim lacked permissions to read, a lookup for a key matching that
3108 symlink would fail. Now it is enough that a matching directory entry
3109 exists, symlink or not. (Bugzilla 503.)
3111 PH/03 The body_linecount and body_zerocount variables are now exported in the
3114 PH/04 Added the $dnslist_matched variable.
3116 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3117 This means they are set thereafter only if the connection becomes
3120 PH/06 Added the client_condition to authenticators so that some can be skipped
3121 by clients under certain conditions.
3123 PH/07 The error message for a badly-placed control=no_multiline_responses left
3124 "_responses" off the end of the name.
3126 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3128 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3129 (without spaces) instead of just copying the configuration text.
3131 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3133 PH/11 Added $max_received_linelength.
3135 PH/12 Added +ignore_defer and +include_defer to host lists.
3137 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3138 way in which PCRE > 7.0 is built.
3140 PH/14 Implemented queue_only_load_latch.
3142 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3143 MAIL command. The effect was to mangle the value on 64-bit systems.
3145 PH/16 Another patch from the Sieve maintainer.
3147 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3149 PH/18 If a system quota error occurred while trying to create the file for
3150 a maildir delivery, the message "Mailbox is full" was not appended to the
3151 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3152 only to a quota excession during the actual writing of the file.
3154 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3155 characters?) which causes problems in log lines. The DN values are now
3156 passed through string_printing() before being added to log lines.
3158 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3159 and InterBase are left for another time.)
3161 PH/21 Added message_body_newlines option.
3163 PH/22 Guard against possible overflow in moan_check_errorcopy().
3165 PH/23 POSIX allows open() to be a macro; guard against that.
3167 PH/24 If the recipient of an error message contained an @ in the local part
3168 (suitably quoted, of course), incorrect values were put in $domain and
3169 $local_part during the evaluation of errors_copy.
3175 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3176 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3179 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3180 issue a MAIL command.
3182 PH/02 In an ACL statement such as
3184 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3186 if a client was not listed at all, or was listed with a value other than
3187 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3188 the condition was not true (as it should be), so access was not denied.
3189 The bug was that the ! inversion was incorrectly passed on to the second
3190 item. This has been fixed.
3192 PH/03 Added additional dnslists conditions == and =& which are different from
3193 = and & when the dns lookup returns more than one IP address.
3195 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3196 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3198 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3199 FSYNC, which compiles an option called disable_fsync that allows for
3200 bypassing fsync(). The documentation is heavily laced with warnings.
3202 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3204 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3205 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3206 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3207 including adding "make clean"; (3) Added -fPIC when compiling the test
3208 dynamically loaded module, to get rid of a warning.
3210 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3211 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3212 = 0s. The bug is otherwise harmless.
3214 PH/07 There was a bug in the dovecot authenticator such that the value of
3215 $auth1 could be overwritten, and so not correctly preserved, after a
3216 successful authentication. This usually meant that the value preserved by
3217 the server_setid option was incorrect.
3219 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3221 PH/09 Installed PCRE release 7.0.
3223 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3224 run for batched SMTP input. It is now run at the start of every message
3225 in the batch. While fixing this I discovered that the process information
3226 (output by running exiwhat) was not always getting set for -bs and -bS
3227 input. This is fixed, and it now also says "batched" for BSMTP.
3229 PH/11 Added control=no_pipelining.
3231 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3232 patch, slightly modified), and move the expansion of helo_data till after
3233 the connection is made in the smtp transport (so it can use these
3236 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3238 PH/14 Added log_selector = +pid.
3240 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3242 PH/16 Add ${if forany and ${if forall.
3244 PH/17 Added dsn_from option to vary the From: line in DSNs.
3246 PH/18 Flush SMTP output before performing a callout, unless control =
3247 no_callout_flush is set.
3249 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3250 was true (the default) a successful delivery failed to delete the retry
3251 item, thus causing premature timeout of the address. The bug is now
3254 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3256 PH/21 Long custom messages for fakedefer and fakereject are now split up
3257 into multiline responses in the same way that messages for "deny" and
3258 other ACL rejections are.
3260 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3261 with slight modification.
3263 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3264 draft, changing the syntax and factoring some duplicate code".
3266 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3267 for deliveries of the second and subsequent messages over the same SMTP
3270 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3271 ${reduce, with only minor "tidies".
3273 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3275 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3276 expansion side effects.
3278 PH/27 When a message times out after an over-quota error from an Exim-imposed
3279 quota, the bounce message says "mailbox is full". This message was not
3280 being given when it was a system quota that was exceeded. It now should
3283 MH/03 Made $recipients available in local_scan(). local_scan() already has
3284 better access to the recipient list through recipients_list[], but
3285 $recipients can be useful in postmaster-provided expansion strings.
3287 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3288 in the case of a MAIL command with additional options following the
3289 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3290 were accidentally chopped off.
3292 PH/29 SMTP synchronization checks are implemented when a command is read -
3293 there is a check that no more input is waiting when there shouldn't be
3294 any. However, for some commands, a delay in an ACL can mean that it is
3295 some time before the response is written. In this time, more input might
3296 arrive, invalidly. So now there are extra checks after an ACL has run for
3297 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3298 pipelining has not been advertised.
3300 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3302 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3303 returned in the SMTP response when smtp_return_error_details was false.
3304 This has been fixed.
3306 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3307 instead of the C I/O that was originally supplied, because problems were
3308 reported on Solaris.
3310 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3311 Exim which did not show up earlier: it was assuming that a call to
3312 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3313 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3314 was a macro that became an assignment, so it seemed to work. This has
3315 changed to a proper function call with a void return, hence the compile
3316 error. Exim's code has been fixed.
3318 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3321 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3323 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3325 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3326 because it turns out to be too big or there is a timeout) there is no
3327 "Completed" line in the log. When some messages of this type were
3328 selected by exigrep, they were listed as "not completed". Others were
3329 picked up by some special patterns. I have improved the selection
3330 criteria to be more general.
3332 PH/38 The host_find_failed option in the manualroute router can now be set
3333 to "ignore", to completely ignore a host whose IP address cannot be
3334 found. If all hosts are ignored, the behaviour is controlled by the new
3335 host_all_ignored option.
3337 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3338 homing or because of multiple MX records with /mx) generated more than
3339 one IP address, and the following item turned out to be the local host,
3340 all the secondary addresses of the first item were incorrectly removed
3341 from the list, along with the local host and any following hosts (which
3342 is what is supposed to happen).
3344 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3345 whoever called Exim into the -H file. In the case of the daemon it was
3346 behaving confusingly. When first started, it used values for whoever
3347 started the daemon, but after a SIGHUP it used the Exim user (because it
3348 calls itself on a restart). I have changed the code so that it now always
3351 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3352 message are rejected with the same error (e.g. no authentication or bad
3353 sender address), and a DATA command is nevertheless sent (as can happen
3354 with PIPELINING or a stupid MUA), the error message that was given to the
3355 RCPT commands is included in the rejection of the DATA command. This is
3356 intended to be helpful for MUAs that show only the final error to their
3359 PH/42 Another patch from the Sieve maintainer.
3361 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3362 Eximstats - Fixed some broken HTML links and added missing column headers
3364 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3365 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3367 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3369 PH/43 Yet another patch from the Sieve maintainer.
3371 PH/44 I found a way to check for a TCP/IP connection going away before sending
3372 the response to the final '.' that terminates a message, but only in the
3373 case where the client has not sent further data following the '.'
3374 (unfortunately, this is allowed). However, in many cases there won't be
3375 any further data because there won't be any more messages to send. A call
3376 to select() can be used: if it shows that the input is "ready", there is
3377 either input waiting, or the socket has been closed. An attempt to read
3378 the next input character can distinguish the two cases. Previously, Exim
3379 would have sent an OK response which the client would never have see.
3380 This could lead to message repetition. This fix should cure that, at
3381 least in a lot of common cases.
3383 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3384 advertised in response to EHLO.
3390 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3391 fixed by 4.65/MH/01 (is this a record?) are fixed:
3393 (i) An empty string was always treated as zero by the numeric comparison
3394 operators. This behaviour has been restored.
3396 (ii) It is documented that the numeric comparison operators always treat
3397 their arguments as decimal numbers. This was broken in that numbers
3398 starting with 0 were being interpreted as octal.
3400 While fixing these problems I realized that there was another issue that
3401 hadn't been noticed. Values of message_size_limit (both the global option
3402 and the transport option) were treated as octal if they started with 0.
3403 The documentation was vague. These values are now always treated as
3404 decimal, and I will make that clear in the documentation.
3410 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3411 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3414 MH/01 Don't check that the operands of numeric comparison operators are
3415 integers when their expansion is in "skipping" mode (fixes bug
3416 introduced by 4.64-PH/07).
3418 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3419 child addresses, Exim now panics and dies. Previously, because the count
3420 is held in a short int, deliveries were likely to be lost. As such a
3421 large number of recipients for a single message is ridiculous
3422 (performance will be very, very poor), I have chosen to impose a limit
3423 rather than extend the field.
3429 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3430 leftover -K file (the existence of which was triggered by #402).
3431 While we were at it, introduced process PID as part of the -K
3432 filename. This should rule out race conditions when creating
3435 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3436 processing considerably. Previous code took too long for large mails,
3437 triggering a timeout which in turn triggers #401.
3439 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3440 in the DK code in transports.c. sendfile() is not really portable,
3441 hence the _LINUX specificness.
3443 TF/01 In the add_headers option to the mail command in an Exim filter,
3444 there was a bug that Exim would claim a syntax error in any
3445 header after the first one which had an odd number of characters
3448 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3449 callout verification, Exim cached a "reject" for the entire domain. This
3450 is correct for most verifications, but it is not correct for a recipient
3451 verification with use_sender or use_postmaster set, because in that case
3452 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3453 case of MAIL FROM:<> rejection from other early rejections (e.g.
3454 rejection of HELO). When verifying a recipient using a non-null MAIL
3455 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3456 Whatever the result of the callout, the value of the domain cache is
3457 left unchanged (for any other kind of callout, getting as far as trying
3458 RCPT means that the domain itself is ok).
3460 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3463 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3464 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3465 session, the daemon ignores ECONNECT errors and logs others; it now
3466 ignores EPIPE as well.
3468 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3469 (quoted-printable decoding).
3471 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3472 later the small subsequent patch to fix an introduced bug.
3474 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3476 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3478 PH/08 An error is now given if message_size_limit is specified negative.
3480 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3481 to be given (somewhat) arbitrary names.
3483 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3486 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3487 miscellaneous code fixes
3489 PH/10 Added the log_reject_target ACL modifier to specify where to log
3492 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3493 hostname. This is wrong, because it relates to the incoming message (and
3494 probably the interface on which it is arriving) and not to the outgoing
3495 callout (which could be using a different interface). This has been
3496 changed to use the value of the helo_data option from the smtp transport
3497 instead - this is what is used when a message is actually being sent. If
3498 there is no remote transport (possible with a router that sets up host
3499 addresses), $smtp_active_hostname is used.
3501 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3502 tweaks were necessary in order to get it to work (see also 21 below):
3503 (a) The code assumed that strncpy() returns a negative number on buffer
3504 overflow, which isn't the case. Replaced with Exim's string_format()
3506 (b) There were several signed/unsigned issues. I just did the minimum
3507 hacking in of casts. There is scope for a larger refactoring.
3508 (c) The code used strcasecmp() which is not a standard C function.
3509 Replaced with Exim's strcmpic() function.
3510 (d) The code set only $1; it now sets $auth1 as well.
3511 (e) A simple test gave the error "authentication client didn't specify
3512 service in request". It would seem that Dovecot has changed its
3513 interface. Fortunately there's a specification; I followed it and
3514 changed what the client sends and it appears to be working now.
3516 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3519 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3520 address A is aliased to B and C, where B exists and C does not. Without
3521 -v the output is "A verified" because verification stops after a
3522 successful redirection if more than one address is generated. However,
3523 with -v the child addresses are also verified. Exim was outputting "A
3524 failed to verify" and then showing the successful verification for C,
3525 with its parentage. It now outputs "B failed to verify", showing B's
3526 parentage before showing the successful verification of C.
3528 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3529 look up a TXT record in a specific list after matching in a combined
3532 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3533 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3534 they consult the DNS. I had assumed they would set it the way they
3535 wanted; and indeed my experiments on Linux seem to show that in some
3536 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3537 To be on the safe side, however, I have now made the interface to
3538 host_find_byname() similar to host_find_bydns(), with an argument
3539 containing the DNS resolver options. The host_find_byname() function now
3540 sets these options at its start, just as host_find_bydns() does. The smtp
3541 transport options dns_qualify_single and dns_search_parents are passed to
3542 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3543 of host_find_byname() use the default settings of RES_DEFNAMES
3544 (qualify_single) but not RES_DNSRCH (search_parents).
3546 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3547 spool_read_header() do less string testing, by means of a preliminary
3548 switch on the second character of optional "-foo" lines. (This is
3549 overdue, caused by the large number of possibilities that now exist.
3550 Originally there were few.) While I was there, I also converted the
3551 str(n)cmp tests so they don't re-test the leading "-" and the first
3552 character, in the hope this might squeeze out yet more improvement.
3554 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3555 flag allowing group syntax was set by the header_syntax check but not
3556 turned off, possible causing trouble later; (2) The flag was not being
3557 set at all for the header_verify test, causing "group"-style headers to
3558 be rejected. I have now set it in this case, and also caused header_
3559 verify to ignore an empty address taken from a group. While doing this, I
3560 came across some other cases where the code for allowing group syntax
3561 while scanning a header line wasn't quite right (mostly, not resetting
3562 the flag correctly in the right place). These bugs could have caused
3563 trouble for malformed header lines. I hope it is now all correct.
3565 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3566 with the "reply" argument non-NULL. The code, however (which originally
3567 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3568 but it didn't always do it. This confused somebody who was copying the
3569 code for some other use. I have removed all the tests.
3571 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3572 feature that was used to support insecure browsers during the U.S. crypto
3573 embargo. It requires special client support, and Exim is probably the
3574 only MTA that supported it -- and would never use it because real RSA is
3575 always available. This code has been removed, because it had the bad
3576 effect of slowing Exim down by computing (never used) parameters for the
3577 RSA_EXPORT functionality.
3579 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3580 authenticator to fail if there's a tab character in the incoming data
3581 (there should never be unless someone is messing about, as it's supposed
3582 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3583 if the connection is using TLS or if the remote IP is the same as the
3584 local IP, and the "valid-client-cert option" if a client certificate has
3587 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3588 authenticators. This can be used for authorization after authentication
3589 succeeds. (In the case of plaintext, it servers for both authentication
3592 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3593 if any retry times were supplied.
3595 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3596 connection, where there is no client IP address to check. In this
3597 situation, the verify now always succeeds.
3599 PH/25 Applied John Jetmore's -Mset patch.
3601 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3603 PH/27 In a string expansion for a processed (not raw) header when multiple
3604 headers of the same name were present, leading whitespace was being
3605 removed from all of them, but trailing whitespace was being removed only
3606 from the last one. Now trailing whitespace is removed from each header
3607 before concatenation. Completely empty headers in a concatenation (as
3608 before) are ignored.
3610 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3611 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3613 PH/29 [Removed. This was a change that I later backed out, and forgot to
3614 correct the ChangeLog entry (that I had efficiently created) before
3615 committing the later change.]
3617 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3618 address errors (4xx response to RCPT) over the same connection as other
3619 messages routed to the same hosts. Such deliveries are always "forced",
3620 so retry times are not inspected. This resulted in far too many retries
3621 for the affected addresses. The effect occurred only when there were more
3622 hosts than the hosts_max_try setting in the smtp transport when it had
3623 the 4xx errors. Those hosts that it had tried were not added to the list
3624 of hosts for which the message was waiting, so if all were tried, there
3625 was no problem. Two fixes have been applied:
3627 (i) If there are any address or message errors in an SMTP delivery, none
3628 of the hosts (tried or untried) are now added to the list of hosts
3629 for which the message is waiting, so the message should not be a
3630 candidate for sending over the same connection that was used for a
3631 successful delivery of some other message. This seems entirely
3632 reasonable: after all the message is NOT "waiting for some host".
3633 This is so "obvious" that I'm not sure why it wasn't done
3634 previously. Hope I haven't missed anything, but it can't do any
3635 harm, as the worst effect is to miss an optimization.
3637 (ii) If, despite (i), such a delivery is accidentally attempted, the
3638 routing retry time is respected, so at least it doesn't keep
3639 hammering the server.
3641 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3642 in ${readsocket because some servers need this prod.
3644 PH/32 Added some extra debug output when updating a wait-xxx database.
3646 PH/33 The hint "could be header name not terminated by colon", which has been
3647 given for certain expansion errors for a long time, was not being given
3648 for the ${if def:h_colon_omitted{... case.
3650 PH/34 The spec says: "With one important exception, whenever a domain list is
3651 being scanned, $domain contains the subject domain." There was at least
3652 one case where this was not true.
3654 PH/35 The error "getsockname() failed: connection reset by peer" was being
3655 written to the panic log as well as the main log, but it isn't really
3656 panic-worthy as it just means the connection died rather early on. I have
3657 removed the panic log writing for the ECONNRESET error when getsockname()
3660 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3661 runs only) independently of the message's sender address. This meant
3662 that, if the 4xx error was in fact related to the sender, a different
3663 message to the same recipient with a different sender could confuse
3664 things. In particular, this can happen when sending to a greylisting
3665 server, but other circumstances could also provoke similar problems.
3666 I have changed the default so that the retry time for these errors is now
3667 based a combination of the sender and recipient addresses. This change
3668 can be overridden by setting address_retry_include_sender=false in the
3671 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3672 remote server are returned as part of bounce messages. This was not
3673 happening for LMTP over a pipe (the lmtp transport), but now it is the
3674 same for both kinds of LMTP.
3676 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3677 in header lines that were in fact CNAMEs. This is no longer the case.
3679 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3680 and queue runs started by the daemon processed all messages. This has
3681 been fixed so that -R and -S can now usefully be given with -q<time>.
3683 PH/40 Import PCRE release 6.7 (fixes some bugs).
3685 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3687 PH/42 Give an error if -q is specified more than once.
3689 PH/43 Renamed the variables $interface_address and $interface_port as
3690 $received_ip_address and $received_port, to make it clear that these
3691 values apply to message reception, and not to the outgoing interface when
3692 a message is delivered. (The old names remain recognized, of course.)
3694 PH/44 There was no timeout on the connect() call when using a Unix domain
3695 socket in the ${readsocket expansion. There now is.
3697 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3698 be meaningful with "accept".
3700 SC/01 Eximstats V1.43
3701 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3703 SC/02 Eximstats V1.44
3704 Use a glob alias rather than an array ref in the generated
3705 parser. This improves both readability and performance.
3707 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3708 Collect SpamAssassin and rejection statistics.
3709 Don't display local sender or destination tables unless
3710 there is data to show.
3711 Added average volumes into the top table text output.
3713 SC/04 Eximstats V1.46
3714 Collect data on the number of addresses (recipients)
3715 as well as the number of messages.
3717 SC/05 Eximstats V1.47
3718 Added 'Message too big' to the list of mail rejection
3719 reasons (thanks to Marco Gaiarin).
3721 SC/06 Eximstats V1.48
3722 Mainlog lines which have GMT offsets and are too short to
3723 have a flag are now skipped.
3725 SC/07 Eximstats V1.49 (Alain Williams)
3726 Added the -emptyok flag.
3728 SC/08 Eximstats V1.50
3729 Fixes for obtaining the IP address from reject messages.
3731 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3732 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3733 whitespace changes from 4.64-PH/27
3735 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3738 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3739 are found, allow negative numbers in numeric criteria)
3741 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3743 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3744 to match changes made in 4.64-PH/43
3746 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3748 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3749 "domain" is actually an IP address, and force a failure. This locks out
3750 those revolvers/nameservers that support "A-for-A" lookups, in
3751 contravention of the specifications.
3753 PH/48 When a host name was looked up from an IP address, and the subsequent
3754 forward lookup of the name timed out, the host name was left in
3755 $sender_host_name, contrary to the specification.
3757 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3758 restricted to single-key lookups, Exim was not diagnosing an error if
3759 * or *@ was used with a query-style lookup.
3761 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3763 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3764 long ago, but noone interested enough thought of it. Let's just say that
3765 the "1.1" means that there are some new functions that weren't there at
3766 some point in the past.
3768 PH/51 Error processing for expansion failure of helo_data from an smtp
3769 transport during callout processing was broken.
3771 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3772 tested/used via the -bh/-bhc/-bs options.
3774 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3775 bug, fixed in subsequent PCRE releases).
3777 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3778 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3780 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3786 SC/01 Use a glob alias rather than an array ref in eximstats generated
3787 parser. This improves both readability and performance.
3789 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3790 Don't display local sender or destination tables in eximstats unless
3791 there is data to show.
3792 Added average volumes into the eximstats top table text output.
3794 SC/03 Collect data on the number of addresses (recipients) as well
3795 as the number of messages in eximstats.
3797 TF/01 Correct an error in the documentation for the redirect router. Exim
3798 does (usually) call initgroups() when daemonizing.
3800 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3801 with consistent privilege compared to when running as a daemon.
3803 TF/03 Note in the spec that $authenticated_id is not set for local
3804 submissions from trusted users.
3806 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3807 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3809 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3810 by adding some example configuration directives to the default
3811 configuration file. A little bit of work is required to uncomment the
3812 directives and define how usernames and passwords are checked, but
3813 there is now a framework to start from.
3815 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3816 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3817 without this. I don't know how relevant this is to other LDAP libraries.
3819 PH/02 Add the verb name to the "unknown ACL verb" error.
3821 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3823 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3825 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3826 directory not expanded when it should be if an expanded home directory
3827 was set for the address (which is overridden by the transport).
3829 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3832 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3833 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3834 because it is too late at that time, and has no effect.
3836 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3837 security issue with \' (bugzilla #107). I could not use the
3838 PQescapeStringConn() function, because it needs a PGconn value as one of
3841 PH/08 When testing addresses using -bt, indicate those final addresses that
3842 are duplicates that would not cause an additional delivery. At least one
3843 person was confused, thinking that -bt output corresponded to deliveries.
3844 (Suppressing duplicates isn't a good idea as you lose the information
3845 about possibly different redirections that led to the duplicates.)
3847 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3848 systems where poll() doesn't work, in particular OS X.
3850 PH/10 Added more information to debugging output for retry time not reached.
3852 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3853 operations in malware.c.
3855 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3858 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3859 syslog_duplication set false, log lines that would normally be written
3860 both the the main log and to the reject log were not written to syslog at
3863 PH/14 In the default configuration, change the use of "message" in ACL warn
3864 statements to "add_header".
3866 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3867 not followed by a command (e.g. "seen endif").
3869 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3870 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3873 PH/17 Added extra conditions to the default value of delay_warning_condition
3877 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3878 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3879 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3882 The Auto-Submitted: and various List- headers are standardised, whereas I
3883 don't think Precedence: ever was.
3885 PH/18 Refactored debugging code in route_finduser() to show more information,
3886 in particular, the error code if getpwnam() issues one.
3888 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3889 This is apparently needed in addition to the PH/07 change above to avoid
3890 any possible encoding problems.
3892 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3893 but not after initializing Perl.
3895 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3896 output them only if debugging. By default they are written stderr,
3897 apparently, which is not desirable.
3899 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3902 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3905 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3907 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3908 authenticated or an ident call has been made. Suppress the default
3909 values for $authenticated_id and $authenticated_sender (but permit -oMai
3910 and -oMas) when testing with -bh.
3912 PH/24 Re-jigged the order of the tests in the default configuration so that the
3913 tests for valid domains and recipients precede the DNS black list and CSA
3914 tests, on the grounds that those ones are more expensive.
3916 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3917 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3918 This bug exists in every version of Exim that I still have, right back to
3921 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3922 However, an attempt to turn on case-sensitivity in a regex key by
3923 including (?-i) didn't work because the subject string was already
3924 lowercased, and the effects were non-intuitive. It turns out that a
3925 one-line patch can be used to allow (?-i) to work as expected.
3931 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3932 other effects) broke the use of negated acl sub-conditions.
3934 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3937 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3938 "Deny" causes Exim to reject the incoming connection with a 554 error.
3939 Unfortunately, if there is a major crisis, such as a disk failure,
3940 tcp-wrappers gives "deny", whereas what one would like would be some
3941 kind of temporary error. A kludge has been added to help with this.
3942 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3943 554 error is used if errno is still zero or contains ENOENT (which occurs
3944 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3947 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3949 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3950 errors. Otherwise a message that provokes a temporary error (when other
3951 messages do not) can cause a whole host to time out.
3953 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3954 addresses were routed directly to files or pipes from a redirect router.
3955 File deliveries just didn't batch; pipe deliveries might have suffered
3958 PH/06 A failure to get a lock for a hints database would erroneously always say
3959 "Failed to get write lock", even when it was really a read lock.
3961 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3962 of 0600. This has been changed to use the value of the lockfile_mode
3963 option (which defaults to 0600).
3965 PH/08 Applied small patch from the Sieve maintainer.
3967 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3968 folder from quota calculations, a direct delivery into this folder messed
3969 up the contents of the maildirsize file. This was because the regex was
3970 used only to exclude .Trash (or whatever) when the size of the mailbox
3971 was calculated. There was no check that a delivery was happening into an
3972 excluded directory. This bug has been fixed by ignoring all quota
3973 processing for deliveries into excluded directories.
3975 PH/10 Added the maildirfolder_create_regex option to appendfile.
3981 PH/01 The code for finding all the local interface addresses on a FreeBSD
3982 system running IPv6 was broken. This may well have applied to all BSD
3983 systems, as well as to others that have similar system calls. The broken
3984 code found IPv4 interfaces correctly, but gave incorrect values for the
3985 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3986 that it would not match correctly against @[] and not recognize the IPv6
3989 PH/02 The ipliteral router was not recognizing addresses of the form user@
3990 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3992 PH/03 Added disable_ipv6.
3994 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3995 decoded form, because it is most often used to construct To: headers
3996 lines in autoreplies, and the decoded form may well be syntactically
3997 invalid. However, $reply_address has leading white space removed, and all
3998 newlines turned into spaces so that the autoreply transport does not
4001 PH/05 If group was specified without a user on a router, and no group or user
4002 was specified on a transport, the group from the router was ignored.
4004 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4005 for visible compile-time settings that can be used to change these
4006 numbers, for those that want even more. Backwards compatibility with old
4007 spool files has been maintained. However, going back to a previous Exim
4008 release will lost any variables that are in spool files.
4010 PH/07 Two small changes when running in the test harness: increase delay when
4011 passing a TCP/IP connection to a new process, in case the original
4012 process has to generate a bounce, and remove special handling of
4013 127.0.0.2 (sic), which is no longer necessary.
4015 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4016 be the same on different OS.
4018 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4021 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4022 whether --show-vars was specified or not
4024 JJ/02 exipick: Added support for new ACL variable spool format introduced
4027 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4028 syntactically invalid From: or Reply-to: line, and a filter used this to
4029 generate an autoreply, and therefore failed to obtain an address for the
4030 autoreply, Exim could try to deliver to a non-existent relative file
4031 name, causing unrelated and misleading errors. What now happens is that
4032 it logs this as a hard delivery error, but does not attempt to create a
4035 PH/11 The exinext utility has a -C option for testing purposes, but although
4036 the given file was scanned by exinext itself; it wasn't being passed on
4037 when Exim was called.
4039 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4040 an end-of-file indication when reading a command response.
4042 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4043 compiled. In many other places in Exim, IPv6 addresses are always
4044 recognized, so I have changed this. It also means that IPv4 domain
4045 literals of the form [IPV4:n.n.n.n] are now always recognized.
4047 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4048 used if the router is not running as root, for example, when verifying at
4049 ACL time, or when using -bh. The debugging output from this situation was
4050 non-existent - all you got was a failure to exec. I have made two
4053 (a) Failures to set uid/gid, the current directory, or a process leader
4054 in a subprocess such as that created by queryprogram now generate
4055 suitable debugging output when -d is set.
4057 (b) The queryprogram router detects when it is not running as root,
4058 outputs suitable debugging information if -d is set, and then runs
4059 the subprocess without attempting to change uid/gid.
4061 PH/15 Minor change to Makefile for building test_host (undocumented testing
4064 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4065 additional section of a DNS packet that returns MX or SRV records.
4066 Instead, it always explicitly searches for A/AAAA records. This avoids
4067 major problems that occur when a DNS server includes only records of one
4068 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4069 fixed another bug: if SRV records were looked up and the corresponding
4070 address records were *not* found in the additional section, the port
4071 values from the SRV records were lost.
4073 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4074 using the correct key (the original address) when searching the retry
4075 rules in order to find which one to use for generating the retry hint.
4077 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4078 adding the default one. Similarly, if it contains a Reply-To: header, the
4079 errors_reply_to option, if set, is not used.
4081 PH/19 When calculating a retry time, Exim used to measure the "time since
4082 failure" by looking at the "first failed" field in the retry record. Now
4083 it does not use this if it is later than than the arrival time of the
4084 message. Instead it uses the arrival time. This makes for better
4085 behaviour in cases where some deliveries succeed, thus re-setting the
4086 "first failed" field. An example is a quota failure for a huge message
4087 when small messages continue to be delivered. Without this change, the
4088 "time since failure" will always be short, possible causing more frequent
4089 delivery attempts for the huge message than are intended.
4090 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4092 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4093 $1, $2, $3) because the numerical variables can be reset during some
4094 expansion items (e.g. "match"), thereby losing the authentication data.
4096 PH/21 Make -bV show the size of off_t variables so that the test suite can
4097 decide whether to run tests for quotas > 2G.
4099 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4100 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4101 filecount value is greater than 2G or if a quota value is greater than 2G
4102 on a system where the size of off_t is not greater than 4, a panic error
4105 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4106 never match. The debug and -bh output now contains an explicit error
4107 message indicating a malformed IPv4 address or mask.
4109 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4110 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4111 PH/23 above applies.
4113 PH/25 Do not write to syslog when running in the test harness. The only
4114 occasion when this arises is a failure to open the main or panic logs
4115 (for which there is an explicit test).
4117 PH/26 Added the /no_tell option to "control=freeze".
4119 PH/27 If a host name lookup failed very early in a connection, for example, if
4120 the IP address matched host_lookup and the reverse lookup yielded a name
4121 that did not have a forward lookup, an error message of the form "no IP
4122 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4123 could be logged. Now it outputs the IP address instead of "NULL".
4125 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4126 allows the sender and the authenticated sender to be set when
4127 submitting a message from within Exim. Since child_open_exim() is
4128 documented for local_scan(), the new function should be too.
4130 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4131 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4132 results in an empty string is now treated as unset.
4134 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4136 PH/31 Added sender_verify_fail logging option.
4138 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4139 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4140 tidied the source and removed it altogether.
4142 PH/33 When a queue run was abandoned because the load average was too high, a
4143 log line was always written; now it is written only if the queue_run log
4144 selector is set. In addition, the log line for abandonment now contains
4145 information about the queue run such as the pid. This is always present
4146 in "start" and "stop" lines but was omitted from the "abandon" line.
4148 PH/34 Omit spaces between a header name and the colon in the error message that
4149 is given when verify = headers_syntax fails (if there are lots of them,
4150 the message gets confusing).
4152 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4153 names, as there are now some PTR records that contain slashes. This check
4154 is only to protect against broken name servers that fall over on strange
4155 characters, so the fact that it applies to all lookups doesn't matter.
4157 PH/36 Now that the new test suite is complete, we can remove some of the
4158 special code in Exim that was needed for the old test suite. For example,
4159 sorting DNS records because real resolvers return them in an arbitrary
4160 order. The new test suite's fake resolver always returns records in the
4163 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4164 bounces) except when queue_only is set, to avoid logging races between
4165 the different processes.
4167 PH/38 Panic-die if .include specifies a non-absolute path.
4169 PH/39 A tweak to the "H" retry rule from its user.
4171 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4172 a label. They prevented compilation on older perls.
4174 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4175 a warning to be raised on newish perls.
4177 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4178 on queue. Changes to match documented behaviour of showing count of
4179 messages matching specified criteria.
4181 PH/40 Changed the default ident timeout from 30s to 5s.
4183 PH/41 Added support for the use of login_cap features, on those BSD systems
4184 that have them, for controlling the resources used by pipe deliveries.
4186 PH/42 The content-scanning code uses fopen() to create files in which to put
4187 message data. Previously it was not paying any attention to the mode of
4188 the files. Exim runs with umask(0) because the rest of the code creates
4189 files with open(), and sets the required mode explicitly. Thus, these
4190 files were ending up world-writeable. This was not a big issue, because,
4191 being within the spool directory, they were not world-accessible. I have
4192 created a function called modefopen, which takes an additional mode
4193 argument. It sets umask(777), creates the file, chmods it to the required
4194 mode, then resets the umask. All the relevant calls to fopen() in the
4195 content scanning code have been changed to use this function.
4197 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4198 to 24 hours. This avoids potential overflow problems when processing G
4199 and H retry rules. I suspect nobody ever tinkers with this value.
4201 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4203 PH/45 When the plaintext authenticator is running as a client, the server's
4204 challenges are checked to ensure they are valid base64 strings. By
4205 default, the authentication attempt is cancelled if an invalid string is
4206 received. Setting client_ignore_invalid_base64 true ignores these errors.
4207 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4208 they are received. Thus, the responses can be made to depend on the
4209 challenges. If an invalid string is ignored, an empty string is placed in
4212 PH/46 Messages that are created by the autoreply transport now contains a
4213 References: header, in accordance with RFCs 2822 and 3834.
4215 PH/47 Added authenticated_sender_force to the smtp transport.
4217 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4219 PH/49 Installed latest patch from the Sieve maintainer.
4221 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4222 was also set, the appendfile transport was unnecessarily scanning a
4223 directory of message files (e.g. for maildir delivery) to find the count
4224 of files (along with the size), even though it did not need this
4225 information. It now does the scan only if it needs to find either the
4226 size of the count of files.
4228 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4230 PH/52 Two bugs concerned with error handling when the smtp transport is
4233 (i) Exim was not creating retry information for temporary errors given
4234 for individual recipients after the DATA command when the smtp transport
4235 was used in LMTP mode. This meant that they could be retried too
4236 frequently, and not timed out correctly.
4238 (ii) Exim was setting the flag that allows error details to be returned
4239 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4240 recipients that were returned after the DATA command.
4242 PH/53 This is related to PH/52, but is more general: for any failing address,
4243 when detailed error information was permitted to be returned to the
4244 sender, but the error was temporary, then after the final timeout, only
4245 "retry timeout exceeded" was returned. Now it returns the full error as
4246 well as "retry timeout exceeded".
4248 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4249 do this, and (what is worse) MTAs that accept it.
4251 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4252 will now be deprecated.
4254 PH/56 New os.c-cygwin from the Cygwin maintainer.
4256 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4257 formats (previously only available in exim formats via -bpr, -bpru,
4258 and -bpra. Now also available in native and exiqgrep formats)
4260 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4261 with very large, slow to parse queues
4263 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4265 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4267 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4268 responses to authentication challenges, though it was showing the
4269 challenges; (ii) I've removed the CR characters from the debug output for
4272 PH/58 Allow for the insertion of a newline as well as a space when a string
4273 is turned into more than one encoded-word during RFC 2047 encoding. The
4274 Sieve code now uses this.
4276 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4277 data_4xx, lost_connection, tls_required.
4279 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4280 message was being sent as an SMTP response.
4282 PH/61 Add -l and -k options to exicyclog.
4284 PH/62 When verifying, if an address was redirected to one new address, so that
4285 verification continued, and the new address failed or deferred after
4286 having set something in $address_data, the value of $address_data was not
4287 passed back to the ACL. This was different to the case when no
4288 redirection occurred. The value is now passed back in both cases.
4290 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4291 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4292 use login_cap.h, so on its own it isn't the distinguishing feature. The
4293 new name refers directly to the setclassresources() function.
4295 PH/65 Added configuration files for NetBSD3.
4297 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4299 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4300 is preferred over IPv4.
4302 PH/68 The bounce_return_message and bounce_return_body options were not being
4303 honoured for bounces generated during the reception of non-SMTP messages.
4304 In particular, this applied to messages rejected by the ACL. This bug has
4305 been fixed. However, if bounce_return_message is true and bounce_return_
4306 body is false, the headers that are returned for a non-SMTP message
4307 include only those that have been read before the error was detected.
4308 (In the case of an ACL rejection, they have all been read.)
4310 PH/69 The HTML version of the specification is now built in a directory called
4311 spec_html instead of spec.html, because the latter looks like a path with
4312 a MIME-type, and this confuses some software.
4314 PH/70 Catch two compiler warnings in sieve.c.
4316 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4317 function verify_get_ident() calls ip_connect() to connect a socket, but
4318 if the "connect()" function timed out, ip_connect() used to close the
4319 socket. However, verify_get_ident() also closes the socket later, and in
4320 between Exim writes to the log, which may get opened at this point. When
4321 the socket was closed in ip_connect(), the log could get the same file
4322 descriptor number as the socket. This naturally causes chaos. The fix is
4323 not to close the socket in ip_connect(); the socket should be closed by
4324 the function that creates it. There was only one place in the code where
4325 this was missing, in the iplookup router, which I don't think anybody now
4326 uses, but I've fixed it anyway.
4328 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4329 well as to direct DNS lookups. Otherwise the handling of names in host
4330 lists is inconsistent and therefore confusing.
4336 PH/01 Two changes to the default runtime configuration:
4338 (1) Move the checks for relay_from_hosts and authenticated clients from
4339 after to before the (commented out) DNS black list checks.
4341 (2) Add control=submission to the relay_from_hosts and authenticated
4342 clients checks, on the grounds that messages accepted by these
4343 statements are most likely to be submissions.
4345 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4347 (1) Generate an error if the third argument for the ${prvs expansion is
4350 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4353 (3) Reset the variables that are obtained from the first argument of
4354 ${prvscheck and used in the second argument before leaving the code,
4355 because their memory is reclaimed, so using them afterwards may do
4358 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4359 one (it's much easier than Tom thought :-).
4361 (5) Because of (4), we can now allow for the use of $prvscheck_result
4362 inside the third argument.
4364 PH/03 For some reason, the default setting of PATH when running a command from
4365 a pipe transport was just "/usr/bin". I have changed it to
4368 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4369 anything to be listed in the output from -bV.
4371 PH/05 When a filter generated an autoreply, the entire To: header line was
4372 quoted in the delivery log line, like this:
4374 => >A.N.Other <ano@some.domain> <original@ddress> ...
4376 This has been changed so that it extracts the operative address. There
4377 may be more than one such address. If so, they are comma-separated, like
4380 => >ano@some.domain,ona@other.domain <original@ddress> ...
4382 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4383 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4384 not being looked up in the rDNS to get a host name, Exim was showing the
4385 IP address twice in Received: lines, even though the IP addresses were
4386 identical. For example:
4388 Received: from [1.2.3.4] (helo=[1.2.3.4])
4390 However, if the real host name was known, it was omitting the HELO data
4391 if it matched the actual IP address. This has been tidied up so that it
4392 doesn't show the same IP address twice.
4394 PH/07 When both +timestamp and +memory debugging was on, the value given by
4395 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4396 called by the debug printing, thereby overwriting the timestamp buffer.
4397 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4399 PH/08 When the original message was included in an autoreply transport, it
4400 always said "this is a copy of the message, including all the headers",
4401 even if body_only or headers_only was set. It now gives an appropriate
4404 PH/09 Applied a patch from the Sieve maintainer which:
4406 o fixes some comments
4407 o adds the (disabled) notify extension core
4408 o adds some debug output for the result of if/elsif tests
4409 o points to the current vacation draft in the documentation
4410 and documents the missing references header update
4414 o fixes a bug in processing the envelope test (when testing
4415 multiple envelope elements, the last element determined the
4418 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4419 Electronic Mail") by including:
4421 Auto-submitted: auto-generated
4423 in the messages that it generates (bounce messages and others, such as
4424 warnings). In the case of bounce messages for non-SMTP messages, there was
4425 also a typo: it was using "Auto_submitted" (underscore instead of
4426 hyphen). Since every message generated by Exim is necessarily in response
4427 to another message, thes have all been changed to:
4429 Auto-Submitted: auto-replied
4431 in accordance with these statements in the RFC:
4433 The auto-replied keyword:
4435 - SHOULD be used on messages sent in direct response to another
4436 message by an automatic process,
4438 - MUST NOT be used on manually-generated messages,
4440 - MAY be used on Delivery Status Notifications (DSNs) and Message
4441 Disposition Notifications (MDNs),
4443 - MUST NOT be used on messages generated by automatic or periodic
4444 processes, except for messages which are automatic responses to
4447 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4448 to the default Received: header definition.
4450 PH/12 Added log selector acl_warn_skipped (default on).
4452 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4453 variables because (a) they are in the wrong storage pool and (b) even if
4454 they were copied, it wouldn't work properly because of the caching.
4456 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4457 checking when decoding. Apparently there are clients that generate
4458 overlong encoded strings. Why am I not surprised?
4460 PH/15 If the first argument of "${if match_address" was not empty, but did not
4461 contain an "@" character, Exim crashed. Now it writes a panic log message
4462 and treats the condition as false.
4464 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4466 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4467 extension "envelope-auth". The code is finished and in agreement with
4468 other implementations, but there is no documentation so far and in fact,
4469 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4470 not changing the active code.
4472 Print executed "if" and "elsif" statements when debugging is used. This
4473 helps a great deal to understand what a filter does.
4475 Document more things not specified clearly in RFC3028. I had all this
4476 sorted out, when out of a sudden new issues came to my mind. Oops."
4478 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4481 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4482 canonical form (as documented). However, after a host name lookup from
4483 the IP address, check_host() was doing a simple string comparison with
4484 addresses acquired from the DNS when checking that the found name did
4485 have the original IP as one of its addresses. Since any found IPv6
4486 addresses are likely to be in abbreviated form, the comparison could
4487 fail. Luckily, there already exists a function for doing the comparison
4488 by converting both addresses to binary, so now that is used instead of
4489 the text comparison.
4491 PH/20 There was another similar case to PH/19, when a complete host name was
4492 given in a host list; looking up its IP address could give an abbreviated
4493 form, whereas the current host's name might or might not be abbreviated.
4494 The same fix has been applied.
4500 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4501 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4504 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4505 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4507 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4509 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4510 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4511 Also, for "Auto-Submitted", treat anything other than "no" as
4512 non-personal, in accordance with RFC 3834. (Previously it treated
4513 anything starting "auto-" as non-personal.)
4515 TF/01 The control=submission/name=... option had a problem with syntax
4516 errors if the name included a slash character. The /name= option
4517 now slurps the rest of the string, so it can include any characters
4518 but it must come last in the list of options (after /sender_retain
4521 PH/05 Some modifications to the interface to the fake nameserver for the new
4529 TK/01 Added the "success_on_redirect" address verification option. See
4530 NewStuff for rationale and an example.
4532 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4534 PH/02 Patch to exigrep to allow it to work on syslog lines.
4536 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4537 fread() to skip over the body file's header line, because in Cygwin the
4538 header line is locked and is inaccessible.
4540 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4541 co-exist for some time) to make it clear that it is the Exim ID that is
4542 referenced, not the Message-ID: header line.
4544 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4545 string_format() function, because snprintf() does not exist on all
4548 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4549 ${stat: expansion item.
4551 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4552 protocol synchronization error", to keep the pedants happy.
4554 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4555 well as for IRIX systems, when gcc is being used. See the host.c source
4558 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4560 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4563 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4564 transport and to the smtp transport in LMTP mode.
4566 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4568 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4569 run a filter in a subprocess. This could lead to confusion in subsequent
4570 lookups in the parent process. There should also be a search_tidyup() at
4571 the end of the subprocess.
4573 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4574 only if the host matched helo_try_verify_hosts, which caused the
4575 verification to occur when the EHLO/HELO command was issued. The ACL just
4576 tested the remembered result. Now, if a previous verification attempt has
4577 not happened, "verify = helo" does it there and then.
4579 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4581 TK/03 Fix log output including CR from clamd.
4583 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4584 exist provoked a memory error which could cause a segfault.
4586 PH/15 Installed PCRE 6.2
4588 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4590 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4591 of the problem. Specifically, suggested +O2 rather than +O1 for the
4594 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4596 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4597 record, the port was not correctly logged when the outgoing_port log
4598 selector was set (it logged the transort's default port).
4600 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4601 fallback_hosts, and "hosts" in the smtp transport.
4603 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4604 host errors such as "Connection refused".
4606 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4607 authentication with radiusclient 0.4.9:
4609 - Error returned from rc_read_config was caught wrongly
4610 - Username/password not passed on to radius server due to wrong length.
4612 The presumption is that some radiusclient API changes for 4.51/PH/17
4613 were not taken care of correctly. The code is still untested by me (my
4614 Linux distribution still has 0.3.2 of radiusclient), but it was
4615 contributed by a Radius user.
4617 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4618 expanding the "port" option of the smtp transport.
4620 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4621 while reading a MIME header. Thanks to Tom Hughes for a patch.
4623 PH/24 Include config.h inside local_scan.h so that configuration settings are
4626 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4627 that in an ACL for RCPT (for example), you can examine exactly what was
4630 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4631 commands, but it was not correctly comparing the address with the actual
4632 client host address. Thus, it would show the EHLO address in Received:
4633 header lines when this was not necessary.
4635 PH/27 Added the % operator to ${eval:}.
4637 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4638 it should be ignoring failures (because with -C, for example, it has lost
4639 privilege). It wasn't ignoring creation failures other than "already
4642 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4645 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4646 because an input line was too long, either on its own, or by virtue of
4647 too many continuations, the temporary file was not being removed, and the
4648 return code was incorrect.
4650 PH/31 Missing "BOOL" in function definition in filtertest.c.
4652 PH/32 Applied Sieve patches from the maintainer.
4654 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4656 PH/33 Added "verify = not_blind".
4658 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4659 Local/Makefile (with some defaults set). These are used in built scripts
4660 such as exicyclog, but they have never been used in the exim_install
4661 script (though there are many overriding facilities there). I have
4662 arranged that the exim_install script now takes note of these two
4665 PH/35 Installed configuration files for Dragonfly.
4667 PH/36 When a locally submitted message by a trusted user did not contain a
4668 From: header, and the sender address was obtained from -f or from an SMTP
4669 MAIL command, and the trusted user did not use -F to supply a sender
4670 name, $originator_name was incorrectly used when constructing a From:
4671 header. Furthermore, $originator_name was used for submission mode
4672 messages from external hosts without From: headers in a similar way,
4673 which is clearly wrong.
4675 PH/37 Added control=suppress_local_fixups.
4677 PH/38 When log_selector = +received_sender was set, and the addition of the
4678 sender made the log line's construction buffer exactly full, or one byte
4679 less than full, an overflow happened when the terminating "\n" was
4682 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4683 when the result of a list match is failure because a DNS lookup failed.
4685 PH/40 RM_COMMAND is now used in the building process.
4687 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4688 the "build-* directories that it finds.
4690 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4691 address in a domain literal was a prefix of an interface address.
4693 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4694 when verifying a sender address, unless rewrite_headers is false.
4696 PH/44 Wrote a long comment about why errors_to addresses are verified as
4697 recipients, not senders.
4699 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4700 the ratelimit ACL was added.
4702 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4704 PH/46 Added extra information about PostgreSQL errors to the error string.
4706 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4707 avoiding the need to install special zones in a real server. This is
4708 backwards compatible; if it can't find the fake resolver, it drops back.
4709 Thus, both old and new test suites can be run.
4711 TF/02 Added util/ratelimit.pl
4713 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4714 clock is set back in time.
4716 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4717 Candler <B.Candler@pobox.com>.
4719 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4720 OK for addresses that are the result of redirections.
4722 PH/48 A number of further additions for the benefit of the new test suite,
4723 including a fake gethostbyname() that interfaces to the fake DNS resolver
4726 TF/06 The fix for widen_domains has also been applied to qualify_single and
4727 search_parents which are the other dnslookup options that can cause
4730 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4733 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4735 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4736 service" if it encountered an MX record with an empty target hostname.
4737 The message is now "an MX or SRV record indicated no SMTP service".
4739 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4740 if the DNS of the sending site is misconfigured. This is quite a
4741 common situation. This change restores the behaviour of treating a
4742 helo verification defer as a failure.
4744 PH/51 If self=fail was set on a router, the bounce message did not include the
4745 actual error message.
4751 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4753 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4754 command itself ended in error, the underlying message about the transport
4755 filter timeout was being overwritten with the pipe command error. Now the
4756 underlying error message should be appended to the second error message.
4758 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4760 PH/02 Reduce the amount of output that "make" produces by default. Full output
4761 can still be requested.
4763 PH/03 The warning log line about a condition test deferring for a "warn" verb
4764 was being output only once per connection, rather than after each
4765 occurrence (because it was using the same function as for successful
4766 "warn" verbs). This seems wrong, so I have changed it.
4768 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4769 it should not have, which might have caused a crash in the right
4770 circumstances, but probably never did.
4772 PH/04 Installed a modified version of Tony Finch's patch to make submission
4773 mode fix the return path as well as the Sender: header line, and to
4774 add a /name= option so that you can make the user's friendly name appear
4777 TF/03 Added the control = fakedefer ACL modifier.
4779 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4780 Mark Lowes for thorough testing.
4782 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4784 TK/03 Merged latest SRS patch from Miles Wilton.
4786 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4787 with the definition in sysexits.h (which is #included earlier).
4788 Fortunately, Exim does not actually use EX_OK. The code used to try to
4789 preserve the sysexits.h value, by assuming that macro definitions were
4790 scanned for macro replacements. I have been disabused of this notion,
4791 so now the code just undefines EX_OK before #including unistd.h.
4793 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4794 in the smtp transport. When a block could not be written in a single
4795 write() function, the timeout was being re-applied to each part-write.
4796 This seems wrong - if the receiver was accepting one byte at a time it
4797 would take for ever. The timeout is now adjusted when this happens. It
4798 doesn't have to be particularly precise.
4800 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4801 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4803 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4804 without a domain if the check to <postmaster@domain> fails.
4806 SC/01 Eximstats: added -xls and the ability to specify output files
4807 (patch written by Frank Heydlauf).
4809 SC/02 Eximstats: use FileHandles for outputting results.
4811 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4813 SC/04 Eximstats: fixed display of large numbers with -nvr option
4815 SC/05 Eximstats: fixed merging of reports with empty tables.
4817 SC/06 Eximstats: added the -include_original_destination flag
4819 SC/07 Eximstats: removed tabs and trailing whitespace.
4821 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4823 TK/06 MBOX spool code: Add real "From " MBOX separator line
4824 so the .eml file is really in mbox format (even though
4825 most programs do not really care). Patch from Alex Miller.
4827 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4828 The latter is generated from $received_to and is only set if the
4829 message has one envelope recipient. SA can use these headers,
4830 obviously out-of-the-box. Patch from Alex Miller.
4832 PH/08 The ${def test on a variable was returning false if the variable's
4833 value was "0", contrary to what the specification has always said!
4834 The result should be true unless the variable is empty.
4836 PH/09 The syntax error of a character other than { following "${if
4837 def:variable_name" (after optional whitespace) was not being diagnosed.
4838 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4839 accidental colon was present, for example, could give incorrect results.
4841 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4842 result is used (not including appendfile, where other changes are about
4845 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4846 This involved changing a lot of size variables from int to off_t. It
4847 should work with maildirs and everything.
4849 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4850 spamd dying while we are connected to it.
4852 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4855 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4856 function for BDB 4.3.
4858 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4860 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4861 the sources for PCRE have been re-arranged and more files are now
4864 PH/15 The code I had for printing potentially long long variables in PH/11
4865 above was not the best (it lost precision). The length of off_t variables
4866 is now inspected at build time, and an appropriate printing format (%ld
4867 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4868 to be "long long int" or "long int". This is needed for the internal
4869 formatting function string_vformat().
4871 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4872 the configuration file to be ":syslog", then the script "guesses" where
4873 the logs files are, rather than using the compiled in default. In our
4874 case the guess is not the same as the compiled default, so the script
4875 suddenly stopped working when I started to use syslog. The patch checks
4876 to see if log_file_path is "". If so, it attempts to read it from exim
4877 with no configuration file to get the compiled in version, before it
4878 falls back to the previous guessing code."
4880 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4881 implementing BATV in an Exim configuration. See NewStuff for the gory
4884 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4885 Makefile that are specific to HP-UX.
4887 PH/18 If the "use_postmaster" option was set for a recipient callout together
4888 with the "random" option, the postmaster address was used as the MAIL
4889 FROM address for the random test, but not for the subsequent recipient
4890 test. It is now used for both.
4892 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4893 patch removes a few documentation additions to RFC 3028, because the
4894 latest draft now contains them. It adds the new en;ascii-case comparator
4895 and a new error check for 8bit text in MIME parts. Comparator and
4896 require names are now matched exactly. I enabled the subaddress
4897 extension, but it is not well tested yet (read: it works for me)."
4899 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4900 rework some of the code of TK/09 above to avoid the hardwired use of
4901 "%lld" and "long long". Replaced the call to snprintf() with a call to
4904 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4905 records point to non-existent hosts", "retry timeout exceeded", and
4906 "retry time not reached for any host after a long failure period".
4908 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4909 experimental DomainKeys support:
4911 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4912 (2) On an error such as an illegally used "control", the wrong name for
4913 the control was given.
4915 These problems did NOT occur unless DomainKeys support was compiled.
4917 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4919 PH/24 Added ${if match_ip condition.
4921 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4922 databases so that it will be absolutely obvious if a crash occurs in the
4923 DB library. This is a regular occurrence (often caused by mis-matched
4926 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4927 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4928 picked up on a user's system that detects such things. There doesn't seem
4929 to be a gcc warning option for this - only an attribute that has to be
4930 put on the function's prototype. It seems that in Fedora Core 4 they have
4931 set this on a number of new functions. No doubt there will be more in due
4934 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4935 specify a transport. However, if an address that was verified by such a
4936 router was the subject of a callout, Exim crashed because it tried to
4937 read the rcpt_include_affixes from the non-existent transport. Now it
4938 just assumes that the setting of that option is false. This bug was
4939 introduced by 4.51/PH/31.
4941 PH/28 Changed -d+all to exclude +memory, because that information is very
4942 rarely of interest, but it makes the output a lot bigger. People tend to
4943 do -d+all out of habit.
4945 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4946 code in os-type was giving problems when libc.so lives in lib64, like on
4949 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4950 aren't the modern standard, and it seems that some systems' include files
4951 don't always have them. Exim was already checking for some of the newer
4952 ones like T_AAAA, and defining it itself. I've added checks for all the
4953 record types that Exim uses.
4955 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4956 not automatically generating a new one, as it is supposed to. This
4957 prevented TLS from working. If the file did exist, but contained invalid
4958 data, a new version was generated, as expected. It was only the case of a
4959 non-existent file that was broken.
4961 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4962 with a change in libdomainkeys > 0.64.
4964 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4965 from DNS. If the selector record carries the flag, it now has
4966 precedence over the domain-wide flag.
4968 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4970 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4971 the use of a port name that isn't defined in /etc/services) occurred, the
4972 message was deferred as in a normal delivery, and thus remained on the
4973 spool, instead of being failed because of the mua_wrapper setting. This
4974 is now fixed, and I tidied up some of the mua_wrapper messages at the
4977 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4978 the messages in a hash of arrays rather than using individual hashes.
4979 This is a bit cleaner and results in dramatic memory savings, albeit
4980 at a slight CPU cost.
4982 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4983 as requested by Marc Sherman.
4985 SC/10 Eximstats: added histograms for user specified patterns as requested
4988 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4990 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4991 fopen() in the content-scanning modules that did not already have it.
4997 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4998 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5000 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5002 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5004 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5005 to test Sieve filters that use "vacation".
5007 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5008 that changes the way the GnuTLS parameters are stored in the cache file.
5009 The new format can be generated externally. For backward compatibility,
5010 if the data in the cache doesn't make sense, Exim assumes it has read an
5011 old-format file, and it generates new data and writes a new file. This
5012 means that you can't go back to an older release without removing the
5015 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5016 work if there are any delivery delays because "one_time" forces the
5017 parent to be marked "delivered", so its unseen clone is never tried
5018 again. For this reason, Exim now forbids the simultaneous setting of
5021 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5022 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5023 entry if you want to know the details. The fix, however, neglected to
5024 consider the case when local delivery batching is involved. The test for
5025 "previously delivered" was not happening when checking to see if an
5026 address could be batched with a previous (undelivered) one; under
5027 certain circumstances this could lead to multiple deliveries to the same
5030 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5031 in its include files, and this causes problems building Exim.
5033 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5034 header_syntax) but Exim was just ignoring anything given after a slash.
5035 In particular, this caused confusion with an attempt to use "verify =
5036 reverse_host_lookup/defer_ok". An error is now given when options are
5037 supplied for verify items that do not have them. (Maybe reverse_host_
5038 lookup should have a defer_ok option, but that's a different point.)
5040 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5041 defined by RFC 821) to 2048, because there were problems with some AUTH
5042 commands, and RFC 1869 says the size should be increased for extended
5043 SMTP commands that take arguments.
5045 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5048 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5049 "unknown" error; now it says that the functionality isn't in the binary.
5051 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5052 an address' error message when a string expansion fails (syntax or
5053 whatever). Otherwise the password may appear in the log. Following change
5054 PH/42 below, there is no longer a chance of it appearing in a bounce
5057 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5059 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5060 of its addresses was ever tried. (Bugzilla bug #2.)
5062 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5063 the result incorrectly in the debug output. (It correctly added a newline
5064 to what was transported.)
5066 TF/01 Added $received_time.
5068 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5069 commented out examples of how to interface to a virus scanner and to
5070 SpamAssassin. Also added commented examples of av_scanner and
5071 spamd_address settings.
5073 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5074 and controls are allowed in which ACLs. There were a couple of minor
5075 errors. Some of the entries in the conditions table (which is a table of
5076 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5077 negation of where the condition IS allowed.
5079 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5081 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5082 header file does not have a version number, so I've had to invent a new
5083 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5084 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5085 radiusclient), but it was contributed by a Radius user.
5087 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5088 files or directories, for OpenSSL.
5090 PH/19 When an Exim process that is running as root has to create an Exim log
5091 file, it does so in a subprocess that runs as exim:exim so as to get the
5092 ownership right at creation (otherwise, other Exim processes might see
5093 the file with the wrong ownership). There was no test for failure of this
5094 fork() call, which would lead to the process getting stuck as it waited
5095 for a non-existent subprocess. Forks do occasionally fail when resources
5096 run out. I reviewed all the other calls to fork(); they all seem to check
5099 PH/20 When checking for unexpected SMTP input at connect time (before writing
5100 the banner), Exim was not dealing correctly with a non-positive return
5101 from the read() function. If the client had disconnected by this time,
5102 the result was a log entry for a synchronization error with an empty
5103 string after "input=" when read() returned zero. If read() returned -1
5104 (an event I could not check), uninitialized data bytes were printed.
5105 There were reports of junk text (parts of files, etc) appearing after
5108 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5110 PH/22 Added support for macro redefinition, and (re)definition in between
5111 driver and ACL definitions.
5113 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5114 forgetting to use the resulting value; it was using the unexpanded value.
5116 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5117 hadn't been configured. The fix is from Juergen Kreileder, who
5118 understands it better than I do:
5120 "Here's what I see happening with three configured cyrus_sasl
5121 authenticators configured (plain, login, cram-md5):
5123 On startup auth_cyrus_sasl_init() gets called for each of these.
5124 This means three calls to sasl_listmech() without a specified mech_list.
5125 => SASL tests which mechs of all available mechs actually work
5126 => three warnings about OTP not working
5127 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5129 With the patch, sasl_listmech() also gets called three times. But now
5130 SASL's mech_list option is set to the server_mech specified in the the
5131 authenticator. Or in other words, the answer from sasl_listmech()
5132 gets limited to just the mech you're testing for (which is different
5134 => the return list contains just 'plain' or 'login', 'cram-md5' or
5135 nothing depending on the value of ob->server_mech.
5137 I've just tested the patch: Authentication still works fine,
5138 unavailable mechs specified in the exim configuration are still
5139 caught, and the auth.log warnings about OTP are gone."
5141 PH/25 When debugging is enabled, the contents of the command line are added
5142 to the debugging output, even when log_selector=+arguments is not
5145 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5146 answer is "GNU", and only if the return is "GNU/something" is the answer
5149 PH/27 $acl_verify_message is now set immediately after the failure of a
5150 verification in an ACL, and so is available in subsequent modifiers. In
5151 particular, the message can be preserved by coding like this:
5153 warn !verify = sender
5154 set acl_m0 = $acl_verify_message
5156 Previously, $acl_verify_message was set only while expanding "message"
5157 and "log_message" when a very denied access.
5159 PH/28 Modified OS/os.c-Linux with
5161 -#ifndef OS_LOAD_AVERAGE
5162 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5164 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5165 nomenclature these days.)
5167 PH/29 Installed patch from the Sieve maintainer that adds the options
5168 sieve_useraddress and sieve_subaddress to the redirect router.
5170 PH/30 In these circumstances:
5171 . Two addresses routed to the same list of hosts;
5172 . First host does not offer TLS;
5173 . First host accepts first address;
5174 . First host gives temporary error to second address;
5175 . Second host offers TLS and a TLS session is established;
5176 . Second host accepts second address.
5177 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5178 and peerdn, if requested) that were in fact used only for the second
5181 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5182 attention to any local part prefix or suffix that was matched by the
5183 router that accepted the address. It now behaves in the same way as it
5184 does for delivery: the affixes are removed from the local part unless
5185 rcpt_include_affixes is set on the transport.
5187 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5188 timeout during the DATA phase of an incoming message.
5190 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5191 applied a patch sanctioned by the Sieve maintainer.
5193 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5194 the uid or gid is negative. A case of a negative gid caused this to be
5195 noticed. The fix allows for either to be negative.
5197 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5198 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5201 PH/36 The change PH/12 above was broken. Fixed it.
5203 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5204 the grounds that routing the same address twice would always produce the
5205 same answer. This might have been true once, but it is certainly no
5206 longer true now. Routing a child address may depend on the previous
5207 routing that produced that child. Some complicated redirection strategies
5208 went wrong when messages had multiple recipients, and made Exim's
5209 behaviour dependent on the order in which the addresses were given.
5211 I have moved the duplicate checking until after the routing is complete.
5212 Exim scans the addresses that are assigned to local and remote
5213 transports, and removes any duplicates. This means that more work will be
5214 done, as duplicates will always all be routed, but duplicates are
5215 presumably rare, so I don't expect this is of any significance.
5217 For deliveries to pipes, files, and autoreplies, the duplicate checking
5218 still happens during the routing process, since they are not going to be
5221 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5222 It corrects a timeout issue with spamd. This is Ian's comment: "The
5223 background is that sometimes spamd either never reads data from a
5224 connection it has accepted, or it never writes response data. The exiscan
5225 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5226 blindly assumes that writes won't block so it may never time out."
5228 PH/39 Allow G after quota size as well as K and M.
5230 PH/40 The value set for $authenticated_id in an authenticator may not contain
5231 binary zeroes or newlines because the value is written to log lines and
5232 to spool files. There was no check on this. Now the value is run through
5233 the string_printing() function so that such characters are converted to
5234 printable escape sequences.
5236 PH/41 $message_linecount is a new variable that contains the total number of
5237 lines in the message. Compare $body_linecount, which is the count for the
5240 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5241 bounce and delay warning messages, except in certain special cases, which
5244 (a) An SMTP error message from a remote host;
5245 (b) A message specified in a :fail: redirection;
5246 (c) A message specified in a "fail" command in a system filter;
5247 (d) A message specified in a FAIL return from the queryprogram router;
5248 (e) A message specified by the cannot_route_message router option.
5250 In these cases only, Exim does include the error details in bounce and
5251 warning messages. There are also a few cases where bland messages such
5252 as "unrouteable address" or "local delivery error" are given.
5254 PH/43 $value is now also set for the "else" part of a ${run expansion.
5256 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5257 being worked on, but at least Exim now implements the latest version to
5260 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5261 process to complete was treated as a delivery failure, a timeout while
5262 writing the message to the pipe was logged, but erroneously treated as a
5263 successful delivery. Such timeouts include transport filter timeouts. For
5264 consistency with the overall process timeout, these timeouts are now
5265 treated as errors, giving rise to delivery failures by default. However,
5266 there is now a new Boolean option for the pipe transport called
5267 timeout_defer, which, if set TRUE, converts the failures into defers for
5268 both kinds of timeout. A transport filter timeout is now identified in
5271 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5272 systems where "make" and "gmake" are different, calling "gmake" at top
5273 level broke things. I've arranged for the value of $(MAKE) to be passed
5274 from the Makefile to this script so that it can call the same version of
5278 A note about Exim versions 4.44 and 4.50
5279 ----------------------------------------
5281 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5282 changes of various kinds. As a consequence, a big documentation update was
5283 needed. This delayed the release for rather longer than seemed good, especially
5284 in the light of a couple of (minor) security issues. Therefore, the changes
5285 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5286 release. So 4.44 and 4.50 are in effect two different branches that both start
5289 I have left the 4.50 change log unchanged; it contains all the changes since
5290 4.43. The change log for 4.44 is below; many of its items are identical to
5291 those for 4.50. This seems to be the most sensible way to preserve the
5292 historical information.
5298 1. Minor wording change to the doc/README.SIEVE file.
5300 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5301 computation of the current number of files was incorrect.
5303 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5304 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5307 4. Give more explanation in the error message when the command for a transport
5308 filter fails to execute.
5310 5. There are several places where Exim runs a non-Exim command in a
5311 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5312 was being done only for the command run by the queryprogram router. It is
5313 now done for all such subprocesses. The other cases are: ${run, transport
5314 filters, and the commands run by the lmtp and pipe transports.
5316 6. Added CONFIGURE_GROUP build-time option.
5318 7. Some older OS have a limit of 256 on the maximum number of file
5319 descriptors. Exim was using setrlimit() to set 1000 as a large value
5320 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5321 systems. I've change it so that if it can't get 1000, it tries for 256.
5323 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5324 was an oversight, and furthermore, ever since the addition of extra
5325 controls (e.g. 4.43/32), the checks on when to allow different forms of
5326 "control" were broken. There should now be diagnostics for all cases when a
5327 control that does not make sense is encountered.
5329 9. Added the /retain_sender option to "control=submission".
5331 10. $recipients is now available in the predata ACL (oversight).
5333 11. Tidy the search cache before the fork to do a delivery from a message
5334 received from the command line. Otherwise the child will trigger a lookup
5335 failure and thereby defer the delivery if it tries to use (for example) a
5336 cached ldap connection that the parent has called unbind on.
5338 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5339 of $address_data from the recipient verification was clobbered by the
5340 sender verification.
5342 13. The value of address_data from a sender verification is now available in
5343 $sender_address_data in subsequent conditions in the ACL statement.
5345 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5347 15. Added a new option "connect=<time>" to callout options, to set a different
5350 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5351 was its contents. (It was OK if the option was not defined at all.)
5353 17. A "Completed" log line is now written for messages that are removed from
5354 the spool by the -Mrm option.
5356 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5357 information about exactly what failed.
5359 19. Added -dd to debug only the daemon process.
5361 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5362 handles timeouts, both on the server side and network timeouts. Renamed the
5363 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5365 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5366 It is now set to "smtps".
5368 22. $host_address is now set to the target address during the checking of
5369 ignore_target_hosts.
5371 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5372 being passed; this would have caused $sender_host_name to have been used if
5373 matching the list had actually called for a host name (not very likely,
5374 since this list is usually IP addresses). A host name is now passed as
5377 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5378 code that specifies a non-restarting handler (typically sigaction() in
5379 modern systems) in an attempt to fix a rare and obscure crash bug.
5381 25. Narrowed the window for a race in the daemon that could cause it to ignore
5382 SIGCHLD signals. This is not a major problem, because they are used only to
5383 wake it up if nothing else does.
5385 26. A malformed maildirsize file could cause Exim to calculate negative values
5386 for the mailbox size or file count. Odd effects could occur as a result.
5387 The maildirsize information is now recalculated if the size or filecount
5390 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5391 support for a long time. Removed HAVE_SYS_VFS_H.
5393 28. Installed the latest version of exipick from John Jetmore.
5395 29. In an address list, if the pattern was not a regular expression, an empty
5396 subject address (from a bounce message) matched only if the pattern was an
5397 empty string. Non-empty patterns were not even tested. This was the wrong
5398 because it is perfectly reasonable to use an empty address as part of a
5399 database query. An empty address is now tested by patterns that are
5400 lookups. However, all the other forms of pattern expect the subject to
5401 contain a local part and a domain, and therefore, for them, an empty
5402 address still always fails if the pattern is not itself empty.
5404 30. Exim went into a mad DNS loop when attempting to do a callout where the
5405 host was specified on an smtp transport, and looking it up yielded more
5406 than one IP address.
5408 31. Re-factored the code for checking spool and log partition space into a
5409 function that finds that data and another that does the check. The former
5410 is then used to implement four new variables: $spool_space, $log_space,
5411 $spool_inodes, and $log_inodes.
5413 32. The RFC2047 encoding function was originally intended for short strings
5414 such as real names; it was not keeping to the 75-character limit for
5415 encoded words that the RFC imposes. It now respects the limit, and
5416 generates multiple encoded words if necessary. To be on the safe side, I
5417 have increased the buffer size for the ${rfc2047: expansion operator from
5420 33. It is now permitted to omit both strings after an "if" condition; if the
5421 condition is true, the result is "true". As before, when the second string
5422 is omitted, a false condition yields an empty string. This makes it less
5423 cumbersome to write custom ACL and router conditions.
5425 34. Failure to deliver a bounce message always caused it to be frozen, even if
5426 there was an errors_to setting on the router. The errors_to setting is now
5429 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5430 canonical form (fully expanded) before being placed in
5431 $sender_host_address.
5433 36. The table in the code that translates DNS record types into text (T_A to
5434 "A" for instance) was missing entries for NS and CNAME. It is just possible
5435 that this could have caused confusion if both these types were looked up
5436 for the same domain, because the text type is used as part of Exim's
5437 per-process caching. But the chance of anyone hitting this buglet seems
5440 37. The dnsdb lookup has been extended in a number of ways.
5442 (1) There is a new type, "zns", which walks up the domain tree until it
5443 finds some nameserver records. It should be used with care.
5445 (2) There is a new type, "mxh", which is like "mx" except that it returns
5446 just the host names, not the priorities.
5448 (3) It is now possible to give a list of domains (or IP addresses) to be
5449 looked up. The behaviour when one of the lookups defers can be
5450 controlled by a keyword.
5452 (4) It is now possible to specify the separator character for use when
5453 multiple records are returned.
5455 38. The dnslists ACL condition has been extended: it is now possible to supply
5456 a list of IP addresses and/or domains to be looked up in a particular DNS
5459 39. Added log_selector=+queue_time_overall.
5461 40. When running the queue in the test harness, wait just a tad after forking a
5462 delivery process, to get repeatability of debugging output.
5464 41. Include certificate and key file names in error message when GnuTLS fails
5465 to set them up, because the GnuTLS error message doesn't include the name
5466 of the failing file when there is a problem reading it.
5468 42. Allow both -bf and -bF in the same test run.
5470 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5472 44. The "Exiscan patch" is now merged into the mainline Exim source.
5474 45. Sometimes the final signoff response after QUIT could fail to get
5475 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5476 < 0 before doing a fflush(). This bug looks as though it goes back to the
5477 introduction of TLS in release 3.20, but "sometimes" must have been rare
5478 because the tests only now provoked it.
5480 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5481 (this can affect the format of dates).
5483 47. exim_tidydb, when checking for the continued existence of a message for
5484 which it has found a message-specific retry record, was not finding
5485 messages that were in split spool directories. Consequently, it was
5486 deleting retry records that should have stayed in existence.
5488 48. Steve fixed some bugs in eximstats.
5490 49. The SPA authentication driver was not abandoning authentication and moving
5491 on to the next authenticator when an expansion was forced to fail,
5492 contradicting the general specification for all authenticators. Instead it
5493 was generating a temporary error. It now behaves as specified.
5495 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5496 (the order specifies the preference for clients). The order is now AES256,
5497 AES128, 3DES, ARCFOUR128.
5499 51. Small patch to Sieve code - explicitly set From: when generating an
5502 52. Exim crashed if a remote delivery caused a very long error message to be
5503 recorded - for instance if somebody sent an entire SpamAssassin report back
5504 as a large number of 550 error lines. This bug was coincidentally fixed by
5505 increasing the size of one of Exim's internal buffers (big_buffer) that
5506 happened as part of the Exiscan merge. However, to be on the safe side, I
5507 have made the code more robust (and fixed the comments that describe what
5510 53. Now that there can be additional text after "Completed" in log lines (if
5511 the queue_time_overall log selector is set), a one-byte patch to exigrep
5512 was needed to allow it to recognize "Completed" as not the last thing in
5515 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5516 patch that reportedly fixes this has been added. I am not expert enough to
5517 create a test for it. This is what the patch creator wrote:
5519 "I found a little strange behaviour of ldap code when working with
5520 Windows 2003 AD Domain, where users was placed in more than one
5521 Organization Units. When I tried to give exim partial DN, the exit code
5522 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5523 But simultaneously result of request was absolutely normal ldap result,
5524 so I produce this patch..."
5526 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5527 REFERENCE, so I have modified the code to exclude the patch when that macro
5530 55. Some experimental protocols are using DNS PTR records for new purposes. The
5531 keys for these records are domain names, not reversed IP addresses. The
5532 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5533 leaves it alone. Component reversal etc. now happens only for IP addresses.
5536 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5538 57. Double the size of the debug message buffer (to 2048) so that more of very
5539 long debug lines gets shown.
5541 58. The exicyclog utility now does better if the number of log files to keep
5542 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5544 59. Two changes related to the smtp_active_hostname option:
5546 (1) $smtp_active_hostname is now available as a variable.
5547 (2) The default for smtp_banner uses $smtp_active_hostname instead
5548 of $primary_hostname.
5550 60. The host_aton() function is supposed to be passed a string that is known
5551 to be a valid IP address. However, in the case of IPv6 addresses, it was
5552 not checking this. This is a hostage to fortune. Exim now panics and dies
5553 if the condition is not met. A case was found where this could be provoked
5554 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5555 components; fortuitously, this particular loophole had already been fixed
5556 by change 4.50/55 above.
5558 If there are any other similar loopholes, the new check in host_aton()
5559 itself should stop them being exploited. The report I received stated that
5560 data on the command line could provoke the exploit when Exim was running as
5561 exim, but did not say which command line option was involved. All I could
5562 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5563 running as the user.
5566 61. There was a buffer overflow vulnerability in the SPA authentication code
5567 (which came originally from the Samba project). I have added a test to the
5568 spa_base64_to_bits() function which I hope fixes it.
5571 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5572 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5574 63. The daemon start-up calls getloadavg() while still root for those OS that
5575 need the first call to be done as root, but it missed one case: when
5576 deliver_queue_load_max is set with deliver_drop_privilege. This is
5577 necessary for the benefit of the queue runner, because there is no re-exec
5578 when deliver_drop_privilege is set.
5580 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5581 This has been fixed.
5583 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5584 was in use, was not putting the data itself into the right store pool;
5585 consequently, it could be overwritten for a subsequent message in the same
5586 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5589 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5591 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5592 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5593 the function were treating the return as a boolean value, which happened to
5594 work because 0=false and not-0=true, but is not correct code.
5596 68. The host_aton() function was not handling scoped IPv6 addresses (those
5597 with, for example, "%eth0" on the end) correctly.
5599 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5600 negated items (that is, ~something) in unsigned ints. Some compilers
5601 apparently mutter when there is no cast.
5603 70. If an address verification called from an ACL failed, and did not produce a
5604 user-specific message (i.e. there was only a "system" message), nothing was
5605 put in $acl_verify_message. In this situation, it now puts the system
5608 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5609 session; change 4.31/43 added the unwanted input to the log line - except
5610 that it did not do this in the start of session case. It now does.
5612 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5613 This is wrong and can cause the other end to generate a synchronization
5614 error if it is another Exim or anything else that does the synchronization
5615 check. A QUIT command is no longer sent after a timeout.
5617 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5618 during host lookups.
5620 74. The defer_ok option of callout verification was not working if it was used
5621 when verifying addresses in header lines, that is, for this case:
5623 verify = header_sender/callout=defer_ok
5625 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5626 those file descriptors could be used for SMTP connections. If anything
5627 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5628 could be sent to the SMTP client, causing chaos. The daemon now opens
5629 stdin, stdout, and stderr to /dev/null when it puts itself into the
5632 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5633 log by default. The user can override this with suitable Perl magic.
5635 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5636 the log message when discard triggers, was not working for the DATA ACL or
5637 for the non-SMTP ACL.
5639 78. Error message wording change in sieve.c.
5641 79. If smtp_accept_max_per_host was set, the number of connections could be
5642 restricted to fewer than expected, because the daemon was trying to set up
5643 a new connection before checking whether the processes handling previous
5644 connections had finished. The check for completed processes is now done
5645 earlier. On busy systems, this bug wouldn't be noticed because something
5646 else would have woken the daemon, and it would have reaped the completed
5649 80. If a message was submitted locally by a user whose login name contained one
5650 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5651 It caused a spool format error. I have fixed the spool reading code. A
5652 related problem was that the "from" clause in the Received: line became
5653 illegal because of the space(s). It is now covered by ${quote_local_part.
5655 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5658 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5660 83. Patch from Sieve maintainer to fix unterminated string problem in
5661 "vacation" handling.
5663 84. Some minor changes to the Linux configuration files to help with other
5664 OS variants using glibc.
5666 85. One more patch for Sieve to update vacation handling to latest spec.
5669 ----------------------------------------------------
5670 See the note above about the 4.44 and 4.50 releases.
5671 ----------------------------------------------------
5677 1. Change 4.43/35 introduced a bug that caused file counts to be
5678 incorrectly computed when quota_filecount was set in an appendfile
5681 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5682 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5685 3. Give more explanation in the error message when the command for a transport
5686 filter fails to execute.
5688 4. There are several places where Exim runs a non-Exim command in a
5689 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5690 was being done only for the command run by the queryprogram router. It is
5691 now done for all such subprocesses. The other cases are: ${run, transport
5692 filters, and the commands run by the lmtp and pipe transports.
5694 5. Some older OS have a limit of 256 on the maximum number of file
5695 descriptors. Exim was using setrlimit() to set 1000 as a large value
5696 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5697 systems. I've change it so that if it can't get 1000, it tries for 256.
5699 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5700 was an oversight, and furthermore, ever since the addition of extra
5701 controls (e.g. 4.43/32), the checks on when to allow different forms of
5702 "control" were broken. There should now be diagnostics for all cases when a
5703 control that does not make sense is encountered.
5705 7. $recipients is now available in the predata ACL (oversight).
5707 8. Tidy the search cache before the fork to do a delivery from a message
5708 received from the command line. Otherwise the child will trigger a lookup
5709 failure and thereby defer the delivery if it tries to use (for example) a
5710 cached ldap connection that the parent has called unbind on.
5712 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5713 of $address_data from the recipient verification was clobbered by the
5714 sender verification.
5716 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5717 was its contents. (It was OK if the option was not defined at all.)
5719 11. A "Completed" log line is now written for messages that are removed from
5720 the spool by the -Mrm option.
5722 12. $host_address is now set to the target address during the checking of
5723 ignore_target_hosts.
5725 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5726 being passed; this would have caused $sender_host_name to have been used if
5727 matching the list had actually called for a host name (not very likely,
5728 since this list is usually IP addresses). A host name is now passed as
5731 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5732 code that specifies a non-restarting handler (typically sigaction() in
5733 modern systems) in an attempt to fix a rare and obscure crash bug.
5735 15. Narrowed the window for a race in the daemon that could cause it to ignore
5736 SIGCHLD signals. This is not a major problem, because they are used only to
5737 wake it up if nothing else does.
5739 16. A malformed maildirsize file could cause Exim to calculate negative values
5740 for the mailbox size or file count. Odd effects could occur as a result.
5741 The maildirsize information is now recalculated if the size or filecount
5744 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5745 support for a long time. Removed HAVE_SYS_VFS_H.
5747 18. Updated exipick to current release from John Jetmore.
5749 19. Allow an empty sender to be matched against a lookup in an address list.
5750 Previously the only cases considered were a regular expression, or an
5753 20. Exim went into a mad DNS lookup loop when doing a callout where the
5754 host was specified on the transport, if the DNS lookup yielded more than
5757 21. The RFC2047 encoding function was originally intended for short strings
5758 such as real names; it was not keeping to the 75-character limit for
5759 encoded words that the RFC imposes. It now respects the limit, and
5760 generates multiple encoded words if necessary. To be on the safe side, I
5761 have increased the buffer size for the ${rfc2047: expansion operator from
5764 22. Failure to deliver a bounce message always caused it to be frozen, even if
5765 there was an errors_to setting on the router. The errors_to setting is now
5768 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5769 canonical form (fully expanded) before being placed in
5770 $sender_host_address.
5772 24. Updated eximstats to version 1.33
5774 25. Include certificate and key file names in error message when GnuTLS fails
5775 to set them up, because the GnuTLS error message doesn't include the name
5776 of the failing file when there is a problem reading it.
5778 26. Expand error message when OpenSSL has problems setting up cert/key files.
5781 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5782 (this can affect the format of dates).
5784 28. exim_tidydb, when checking for the continued existence of a message for
5785 which it has found a message-specific retry record, was not finding
5786 messages that were in split spool directories. Consequently, it was
5787 deleting retry records that should have stayed in existence.
5789 29. eximstats updated to version 1.35
5790 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5791 1.35 - bugfix such that pie charts by volume are generated correctly
5793 30. The SPA authentication driver was not abandoning authentication and moving
5794 on to the next authenticator when an expansion was forced to fail,
5795 contradicting the general specification for all authenticators. Instead it
5796 was generating a temporary error. It now behaves as specified.
5798 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5799 (the order specifies the preference for clients). The order is now AES256,
5800 AES128, 3DES, ARCFOUR128.
5802 31. Small patch to Sieve code - explicitly set From: when generating an
5805 32. Exim crashed if a remote delivery caused a very long error message to be
5806 recorded - for instance if somebody sent an entire SpamAssassin report back
5807 as a large number of 550 error lines. This bug was coincidentally fixed by
5808 increasing the size of one of Exim's internal buffers (big_buffer) that
5809 happened as part of the Exiscan merge. However, to be on the safe side, I
5810 have made the code more robust (and fixed the comments that describe what
5813 33. Some experimental protocols are using DNS PTR records for new purposes. The
5814 keys for these records are domain names, not reversed IP addresses. The
5815 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5816 leaves it alone. Component reversal etc. now happens only for IP addresses.
5819 34. The host_aton() function is supposed to be passed a string that is known
5820 to be a valid IP address. However, in the case of IPv6 addresses, it was
5821 not checking this. This is a hostage to fortune. Exim now panics and dies
5822 if the condition is not met. A case was found where this could be provoked
5823 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5824 components; fortuitously, this particular loophole had already been fixed
5825 by change 4.50/55 or 4.44/33 above.
5827 If there are any other similar loopholes, the new check in host_aton()
5828 itself should stop them being exploited. The report I received stated that
5829 data on the command line could provoke the exploit when Exim was running as
5830 exim, but did not say which command line option was involved. All I could
5831 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5832 running as the user.
5835 35. There was a buffer overflow vulnerability in the SPA authentication code
5836 (which came originally from the Samba project). I have added a test to the
5837 spa_base64_to_bits() function which I hope fixes it.
5840 36. The daemon start-up calls getloadavg() while still root for those OS that
5841 need the first call to be done as root, but it missed one case: when
5842 deliver_queue_load_max is set with deliver_drop_privilege. This is
5843 necessary for the benefit of the queue runner, because there is no re-exec
5844 when deliver_drop_privilege is set.
5846 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5847 was in use, was not putting the data itself into the right store pool;
5848 consequently, it could be overwritten for a subsequent message in the same
5849 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5852 38. Sometimes the final signoff response after QUIT could fail to get
5853 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5854 < 0 before doing a fflush(). This bug looks as though it goes back to the
5855 introduction of TLS in release 3.20, but "sometimes" must have been rare
5856 because the tests only now provoked it.
5862 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5863 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5864 Now it returns an integer. A number of calls were still expecting a T/F
5865 return. Fortuitously, in all cases, the tests worked in OK situations,
5866 which is the norm. However, things would have gone wrong on any write
5867 failures on the smtp file descriptor. This function is used when sending
5868 messages over SMTP and also when doing verify callouts.
5870 2. When Exim is called to do synchronous delivery of a locally submitted
5871 message (the -odf or -odi options), it no longer closes stderr before doing
5874 3. Implemented the mua_wrapper option.
5876 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5878 5. Implemented the functions header_remove(), header_testname(),
5879 header_add_at_position(), and receive_remove_recipient(), and exported them
5882 6. If an ACL "warn" statement specified the addition of headers, Exim already
5883 inserted X-ACL-Warn: at the start if there was no header name. However, it
5884 was not making this test for the second and subsequent header lines if
5885 there were newlines in the string. This meant that an invalid header could
5886 be inserted if Exim was badly configured.
5888 7. Allow an ACL "warn" statement to add header lines at the start or after all
5889 the Received: headers, as well as at the end.
5891 8. Added the rcpt_4xx retry error code.
5893 9. Added postmaster_mailfrom=xxx to callout verification option.
5895 10. Added mailfrom=xxxx to the callout verification option, for verify=
5898 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5899 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5901 12. Inserted some casts to stop certain compilers warning when using pointer
5902 differences as field lengths or precisions in printf-type calls (mostly
5903 affecting debugging statements).
5905 13. Added optional readline() support for -be (dynamically loaded).
5907 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5908 same clock tick as a message's arrival, so that its received time was the
5909 same as the "first fail" time on the retry record, and that message
5910 remained on the queue past the ultimate address timeout, every queue runner
5911 would try a delivery (because it was past the ultimate address timeout) but
5912 after another failure, the ultimate address timeout, which should have then
5913 bounced the address, did not kick in. This was a "< instead of <=" error;
5914 in most cases the first failure would have been in the next clock tick
5915 after the received time, and all would be well.
5917 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5918 being recognized when the domain list was tested by the match_domain
5919 condition in an expansion string.
5921 16. Added the ${str2b64: operator.
5923 17. Exim was always calling setrlimit() to set a large limit for the number of
5924 processes, without checking whether the existing limit was already
5925 adequate. (It did check for the limit on file descriptors.) Furthermore,
5926 errors from getrlimit() and setrlimit() were being ignored. Now they are
5927 logged to the main and panic logs, but Exim does carry on, to try to do its
5928 job under whatever limits there are.
5930 18. Imported PCRE 5.0.
5932 19. Trivial typo in log message " temporarily refused connection" (the leading
5935 20. If the log selector return_path_on_delivery was set and an address was
5936 redirected to /dev/null, the delivery process crashed because it assumed
5937 that a return path would always be set for a "successful" delivery. In this
5938 case, the whole delivery is bypassed as an optimization, and therefore no
5941 21. Internal re-arrangement: the function for sending a challenge and reading
5942 a response while authentication was assuming a zero-terminated challenge
5943 string. It's now changed to take a pointer and a length, to allow for
5944 binary data in such strings.
5946 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5948 23. Exim was not respecting finduser_retries when seeking the login of the
5949 uid under which it was called; it was always trying 10 times. (The default
5950 setting of finduser_retries is zero.) Also, it was sleeping after the final
5951 failure, which is pointless.
5953 24. Implemented tls_on_connect_ports.
5955 25. Implemented acl_smtp_predata.
5957 26. If the domain in control=submission is set empty, Exim assumes that the
5958 authenticated id is a complete email address when it generates From: or
5959 Sender: header lines.
5961 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5962 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5963 chown and chgrp in /bin and hostname in /usr/bin.
5965 28. Exim was keeping the "process log" file open after each use, just as it
5966 does for the main log. This opens the possibility of it remaining open for
5967 long periods when the USR1 signal hits a daemon. Occasional processlog
5968 errors were reported, that could have been caused by this. Anyway, it seems
5969 much more sensible not to leave this file open at all, so that is what now
5972 29. The long-running daemon process does not normally write to the log once it
5973 has entered its main loop, and it closes the log before doing so. This is
5974 so that log files can straightforwardly be renamed and moved. However,
5975 there are a couple of unusual error situations where the daemon does write
5976 log entries, and I had neglected to close the log afterwards.
5978 30. The text of an SMTP error response that was received during a remote
5979 delivery was being truncated at 512 bytes. This is too short for some of
5980 the long messages that one sometimes sees. I've increased the limit to
5983 31. It is now possible to make retry rules that apply only when a message has a
5984 specific sender, in particular, an empty sender.
5986 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5987 it possible to be selective about when SMTP synchronization is enforced.
5989 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5991 32. Implemented hosts_connection_nolog.
5993 33. Added an ACL for QUIT.
5995 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5998 35. Added mailbox_size and mailbox_filecount to appendfile.
6000 36. Added control = no_multiline_responses to ACLs.
6002 37. There was a bug in the logic of the code that waits for the clock to tick
6003 in the case where the clock went backwards by a substantial amount such
6004 that the microsecond fraction of "now" was more than the microsecond
6005 fraction of "then" (but the whole seconds number was less).
6007 38. Added support for the libradius Radius client library this is found on
6008 FreeBSD (previously only the radiusclient library was supported).
6014 1. When certain lookups returned multiple values in the form name=value, the
6015 quoting of the values was not always being done properly. Specifically:
6016 (a) If the value started with a double quote, but contained no whitespace,
6018 (b) If the value contained whitespace other than a space character (i.e.
6019 tabs or newlines or carriage returns) it was not quoted.
6020 This fix has been applied to the mysql and pgsql lookups by writing a
6021 separate quoting function and calling it from the lookup code. The fix
6022 should probably also be applied to nisplus, ibase and oracle lookups, but
6023 since I cannot test any of those, I have not disturbed their existing code.
6025 2. A hit in the callout cache for a specific address caused a log line with no
6026 reason for rejecting RCPT. Now it says "Previous (cached) callout
6027 verification failure".
6029 3. There was an off-by-one bug in the queryprogram router. An over-long
6030 return line was truncated at 256 instead of 255 characters, thereby
6031 overflowing its buffer with the terminating zero. As well as fixing this, I
6032 have increased the buffer size to 1024 (and made a note to document this).
6034 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6035 when Exim is waiting for an SMTP response from a remote server, Exim
6036 restarts its select() call on the socket, thereby resetting its timeout.
6037 This is not a problem when such interrupts are rare. Somebody set up a cron
6038 job to run exiwhat every 2 minutes, which is less than the normal select()
6039 timeout (5 or 10 minutes). This meant that the select() timeout never
6040 kicked in because it was always reset. I have fixed this by comparing the
6041 time when an interrupt arrives with the time at the start of the first call
6042 to select(). If more time than the timeout has elapsed, the interrupt is
6043 treated as a timeout.
6045 5. Some internal re-factoring in preparation for the addition of Sieve
6046 extensions (by MH). In particular, the "personal" test is moved to a
6047 separate function, and given an option for scanning Cc: and Bcc: (which is
6048 not set for Exim filters).
6050 6. When Exim created an email address using the login of the caller as the
6051 local part (e.g. when creating a From: or Sender: header line), it was not
6052 quoting the local part when it contained special characters such as @.
6054 7. Installed new OpenBSD configuration files.
6056 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6057 try to make them clearer.
6059 9. Callout options, other than the timeout value, were being ignored when
6060 verifying sender addresses in header lines. For example, when using
6062 verify = header_sender/callout=no_cache
6064 the cache was (incorrectly) being used.
6066 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6067 only the Cygwin environment.
6069 11. When return_path_on_delivery was set as a log selector, if different remote
6070 addresses in the same message used different return paths and parallel
6071 remote delivery occurred, the wrong values would sometimes be logged.
6072 (Whenever a remote delivery process finished, the return path value from
6073 the most recently started remote delivery process was logged.)
6075 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6076 lines when AUTH and/or TLS are in use. This is the "received protocol"
6077 field. Exim used to use "asmtp" for authenticated SMTP, without any
6078 indication (in the protocol name) for TLS use. Now it follows the RFC and
6079 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6080 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6081 names appear in log lines as well as in Received: header lines.
6083 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6084 extensions, and comparison tests, and to fix some bugs.
6086 14. Changes to the "personal" filter test:
6088 (1) The test was buggy in that it was just doing the equivalent of
6089 "contains" tests on header lines. For example, if a user's address was
6090 anne@some.where, the "personal" test would incorrectly be true for
6092 To: susanne@some.where
6094 This test is now done by extracting each address from the header in turn,
6095 and checking the entire address. Other tests that are part of "personal"
6096 are now done using regular expressions (for example, to check local parts
6097 of addresses in From: header lines).
6099 (2) The list of non-personal local parts in From: addresses has been
6100 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6101 taken from the Sieve specification recommendations.
6103 (3) If the message contains any header line starting with "List-" it is
6104 treated as non-personal.
6106 (4) The test for "circular" in the Subject: header line has been removed
6107 because it now seems ill-conceived.
6109 15. Minor typos in src/EDITME comments corrected.
6111 16. Installed latest exipick from John Jetmore.
6113 17. If headers_add on a router specified a text string that was too long for
6114 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6115 of string_sprintf() is now avoided.
6117 18. $message_body_size was not set (it was always zero) when running the DATA
6118 ACL and the local_scan() function.
6120 19. For the "mail" command in an Exim filter, no default was being set for
6121 the once_repeat time, causing a random time value to be used if "once" was
6122 specified. (If the value happened to be <= 0, no repeat happened.) The
6123 default is now 0s, meaning "never repeat". The "vacation" command was OK
6124 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6125 (I found it when inspecting the code).
6127 20. There is now an overall timeout for performing a callout verification. It
6128 defaults to 4 times the callout timeout, which applies to individual SMTP
6129 commands during the callout. The overall timeout applies when there is more
6130 than one host that can be tried. The timeout is checked before trying the
6131 next host. This prevents very long delays if there are a large number of
6132 hosts and all are timing out (e.g. when the network connections are timing
6133 out). The value of the overall timeout can be changed by specifying an
6134 additional sub-option for "callout", called "maxwait". For example:
6136 verify = sender/callout=5s,maxwait=20s
6138 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6139 to the end before writing, but this should make it even safer).
6141 22. Exim was forgetting that it had advertised PIPELINING for the second and
6142 subsequent messages on an SMTP connection. It was also not resetting its
6143 memory on STARTTLS and an internal HELO.
6145 23. When Exim logs an SMTP synchronization error within a session, it now
6146 records whether PIPELINING has been advertised or not.
6148 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6149 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6150 rather than long int.
6152 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6154 26. Added the never_mail option to autoreply.
6160 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6161 crash if the getsockname() call failed; this can happen if a connection is
6162 closed very soon after it is established. The problem was simply in the
6163 order in which certain operations were done, causing Exim to try to write
6164 to the SMTP stream before it had set up the file descriptor. The bug has
6165 been fixed by making things happen in the correct order.
6171 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6172 before the connection was closed, thus losing the rejection response.
6174 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6175 some early Solaris releases, but causes trouble in current releases where
6176 socklen_t is defined.
6178 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6181 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6184 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6185 that releasing the top of it at the end releases what was used for sub-
6186 expansions (unless the block got too big). However, discard this block if
6187 the first thing is a variable or header, so that we can use its block when
6188 it is dynamic (useful for very large $message_headers, for example).
6190 6. Lookups now cache *every* query, not just the most recent. A new, separate
6191 store pool is used for this. It can be recovered when all lookup caches are
6192 flushed. Lookups now release memory at the end of their result strings.
6193 This has involved some general refactoring of the lookup sources.
6195 7. Some code has been added to the store_xxx() functions to reduce the amount
6196 of flapping under certain conditions.
6198 8. log_incoming_interface used to affect only the <= reception log lines. Now
6199 it causes the local interface and port to be added to several more SMTP log
6200 lines, for example "SMTP connection from", and rejection lines.
6202 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6204 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6206 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6207 could be overwritten at the end of the current message (or the start of a
6208 new message if it was set in a HELO ACL). The value is now preserved for
6209 the duration of the SMTP connection.
6211 12. If a transport had a headers_rewrite setting, and a matching header line
6212 contained an unqualified address, that address was qualified, even if it
6213 did not match any rewriting rules. The underlying bug was that the values
6214 of the flags that permit the existence of unqualified sender and recipient
6215 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6216 non-local messages, and by -bnq for local messages) were not being
6217 preserved with the message after it was received.
6219 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6220 "next input=" as part of the text comprising the host identity instead of
6221 the correct text. The code was using the same buffer for two different
6222 strings. However, depending on which order the printing function evaluated
6223 its arguments, the bug did not always show up. Under Linux, for example, my
6224 test suite worked just fine.
6226 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6227 doesn't work with some older versions of Perl. It has been changed to "my",
6228 which in any case is probably the better facility to use.
6230 15. A really picky compiler found some instances of statements for creating
6231 error messages that either had too many or two few arguments for the format
6234 16. The size of the buffer for calls to the DNS resolver has been increased
6235 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6236 for addresses that have a lot of PTR records. This alleviates a problem; it
6237 does not fully solve it.
6239 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6240 buffer now truncates the list and logs the incident, which is the same
6241 action as happens when Exim is looking up a host name and its aliases.
6242 Previously in this situation something unpredictable would happen;
6243 sometimes it was "internal error: store_reset failed".
6245 18. If a server dropped the connection unexpectedly when an Exim client was
6246 using GnuTLS and trying to read a response, the client delivery process
6247 crashed while trying to generate an error log message.
6249 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6250 string, for example:
6252 warn message = H1: something\nH2: something
6254 the text was added as a single header line from Exim's point of view
6255 though it ended up OK in the delivered message. However, searching for the
6256 second and subsequent header lines using $h_h2: did not work. This has been
6257 fixed. Similarly, if a system filter added multiple headers in this way,
6258 the routers could not see them.
6260 20. Expanded the error message when iplsearch is called with an invalid key to
6261 suggest using net-iplsearch in a host list.
6263 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6264 ACLs are no longer actually imposed (and a message to that effect is
6267 22. If a "gecos" field in a passwd entry contained escaped characters, in
6268 particular, if it contained a \" sequence, Exim got it wrong when building
6269 a From: or a Sender: header from that name. A second bug also caused
6270 incorrect handling when an unquoted " was present following a character
6271 that needed quoting.
6273 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6274 was not being matched caselessly.
6276 24. Arranged for all hyphens in the exim.8 source to be escaped with
6279 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6280 properly. Recipient callout cache records were still being keyed to include
6281 the sender, even when use_sender was set false. This led to far more
6282 callouts that were necessary. The sender is no longer included in the key
6283 when use_sender is false.
6285 26. Added "control = submission" modifier to ACLs.
6287 27. Added the ${base62d: operator to decode base 62 numbers.
6289 28. dnsdb lookups can now access SRV records.
6291 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6292 the configuration file.
6294 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6295 (-v) mode. This makes the output for a verbose queue run more intelligible.
6297 31. Added a use_postmaster feature to recipient callouts.
6299 32. Added the $body_zerocount variable, containing the number of binary zero
6300 bytes in the message body.
6302 33. The time of last modification of the "new" subdirectory is now used as the
6303 "mailbox time last read" when there is a quota error for a maildir
6306 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6308 35. Added +ignore_unknown as a special item in host lists.
6310 36. Code for decoding IPv6 addresses in host lists is now included, even if
6311 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6312 address was recognized as an IP address, but was then not correctly decoded
6313 into binary, causing unexpected and incorrect effects when compared with
6320 1. Very minor rewording of debugging text in manualroute to say "list of
6321 hosts" instead of "hostlist".
6323 2. If verify=header_syntax was set, and a header line with an unqualified
6324 address (no domain) and a large number of spaces between the end of the
6325 name and the colon was received, the reception process suffered a buffer
6326 overflow, and (when I tested it) crashed. This was caused by some obsolete
6327 code that should have been removed. The fix is to remove it!
6329 3. When running in the test harness, delay a bit after writing a bounce
6330 message to get a bit more predictability in the log output.
6332 4. Added a call to search_tidyup() just before forking a reception process. In
6333 theory, someone could use a lookup in the expansion of smtp_accept_max_
6334 per_host which, without the tidyup, could leave open a database connection.
6336 5. Added the variables $recipient_data and $sender_data which get set from a
6337 lookup success in an ACL "recipients" or "senders" condition, or a router
6338 "senders" option, similar to $domain_data and $local_part_data.
6340 6. Moved the writing of debug_print from before to after the "senders" test
6343 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6344 problems for message scanning, either using a data ACL, or using
6345 local_scan() because the Received: header was not generated till after they
6346 were called (in order to set the time as the time of reception completion).
6347 I have revised the way this works. The header is now generated after the
6348 body is received, but before the ACL or local_scan() are called. After they
6349 are run, the timestamp in the header is updated.
6355 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6356 before starting a queue runner without re-exec. This happened only when
6357 deliver_drop_privilege was set or when the Exim user was set to root. The
6358 effect of the bug was that timeouts during subsequent deliveries caused
6359 crashes instead of being properly handled. The handler is now left at its
6360 default (and expected) setting.
6362 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6363 message, again when deliver_drop_privilege is set or Exim is run as root.
6364 The bug described in (1) was not present in this case, but the tidying up
6365 of the other signals was missing. I have made the two cases consistent.
6367 3. The ignore_target_hosts setting on a manualroute router was being ignored
6368 for hosts that were looked up using the /MX notation.
6370 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6373 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6374 operated on the sender address. After changing the $sender_address to <>
6375 for the sender address verify, Exim was re-instated it as the original
6376 (before rewriting) address, but remembering that it had rewritten it, so it
6377 wasn't rewriting it again. This bug also had the effect of breaking the
6378 sender address verification caching when the sender address was rewritten.
6380 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6381 This has been changed so that if the ip literal address matches
6382 ignore_target_hosts, the router declines.
6384 7. Added expansion conditions match_domain, match_address, and match_local_
6385 part (NOT match_host).
6387 8. The placeholder for the Received: header didn't have a length field set.
6389 9. Added code to Exim itself and to exim_lock to test for a specific race
6390 condition that could lead to file corruption when using MBX delivery. The
6391 issue is with the lockfile that is created in /tmp. If this file is removed
6392 after a process has opened it but before that process has acquired a lock,
6393 there is the potential for a second process to recreate the file and also
6394 acquire a lock. This could lead to two Exim processes writing to the file
6395 at the same time. The added code performs the same test as UW imapd; it
6396 checks after acquiring the lock that its file descriptor still refers to
6397 the same named file.
6399 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6400 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6403 11. Added the smtp_active_hostname option. If used, this will typically be made
6404 to depend on the incoming interface address. Because $interface_address is
6405 not set up until the daemon has forked a reception process, error responses
6406 that can happen earlier (such as "too many connections") no longer contain
6409 12. If an expansion in a condition on a "warn" statement fails because a lookup
6410 defers, the "warn" statement is abandoned, and the next ACL statement is
6411 processed. Previously this caused the whole ACL to be aborted.
6413 13. Added the iplsearch lookup type.
6415 14. Added ident_timeout as a log selector.
6417 15. Added tls_certificate_verified as a log selector.
6419 16. Added a global option tls_require_ciphers (compare the smtp transport
6420 option of the same name). This controls incoming TLS connections.
6422 17. I finally figured out how to make tls_require_ciphers do a similar thing
6423 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6424 before starting the TLS session.
6426 18. Tabs are now shown as \t in -bP output.
6428 19. If the log selector return_path_on_delivery was set, Exim crashed when
6429 bouncing a message because it had too many Received: header lines.
6431 20. If two routers both had headers_remove settings, and the first one included
6432 a superfluous trailing colon, the final name in the first list and the
6433 first name in the second list were incorrectly joined into one item (with a
6434 colon in the middle).
6440 1. Added -C and -D options to the exinext utility, mainly to make it easier
6441 to include in the automated testing, but these could be helpful when
6442 multiple configurations are in use.
6444 2. The exinext utility was not formatting the output nicely when there was
6445 an alternate port involved in the retry record key, nor when there was a
6446 message id as well (for retries that were specific to a specific message
6447 and a specific host). It was also confused by IPv6 addresses, because of
6448 the additional colons they contain. I have fixed the IPv4 problem, and
6449 patched it up to do a reasonable job for IPv6.
6451 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6452 delivery, the log line now contains "pipelined" if PIPELINING was used.
6454 4. An SMTP transport process used to panic and die if the bind() call to set
6455 an explicit outgoing interface failed. This has been changed; it is now
6456 treated in the same way as a connect() failure.
6458 5. A reference to $sender_host_name in the part of a conditional expansion
6459 that was being skipped was still causing a DNS lookup. This no longer
6462 6. The def: expansion condition was not recognizing references to header lines
6463 that used bh_ and bheader_.
6465 7. Added the _cache feature to named lists.
6467 8. The code for checking quota_filecount in the appendfile transport was
6468 allowing one more file than it should have been.
6470 9. For compatibility with Sendmail, the command line option
6478 and sets the incoming protocol and host name (for trusted callers). The
6479 host name and its colon can be omitted when only the protocol is to be set.
6480 Note the Exim already has two private options, -pd and -ps, that refer to
6481 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6482 "s", but I don't think that's a major issue.
6484 10. A number of refactoring changes to the code, none of which should affect
6487 (a) The number of logging options was getting close to filling up the
6488 32-bit word that was used as a bit map. I have split them into two classes:
6489 those that are passed in the argument to log_write(), and those that are
6490 only ever tested independently outside of that function. These are now in
6491 separate 32-bit words, so there is plenty of room for expansion again.
6492 There is no change in the user interface or the logging behaviour.
6494 (b) When building, for example, log lines, the code previously used a
6495 macro that called string_cat() twice, in order to add two strings. This is
6496 not really sufficiently general. Furthermore, there was one instance where
6497 it was actually wrong because one of the argument was used twice, and in
6498 one call a function was used. (As it happened, calling the function twice
6499 did not affect the overall behaviour.) The macro has been replaced by a
6500 function that can join an arbitrary number of extra strings onto a growing
6503 (c) The code for expansion conditions now uses a table and a binary chop
6504 instead of a serial search (which was left over from when there were very
6505 few conditions). Also, it now recognizes conditions like "pam" even when
6506 the relevant support is not compiled in: a suitably worded error message is
6507 given if an attempt is made to use such a condition.
6509 11. Added ${time_interval:xxxxx}.
6511 12. A bug was causing one of the ddress fields not to be passed back correctly
6512 from remote delivery subprocesses. The field in question was not being
6513 subsequently used, so this caused to problems in practice.
6515 13. Added new log selectors queue_time and deliver_time.
6517 14. Might have fixed a bug in maildirsizefile handling that threw up
6518 "unexpected character" debug warnings, and recalculated the data
6519 unnecessarily. In any case, I expanded the warning message to give more
6522 15. Added the message "Restricted characters in address" to the statements in
6523 the default ACL that block characters like @ and % in local parts.
6525 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6526 Three changes have been made:
6528 (a) There was a serious bug; a negative response to MAIL caused the whole
6529 recipient domain to be cached as invalid, thereby blocking all messages
6530 to all local parts at the same domain, from all senders. This bug has
6531 been fixed. The domain is no longer cached after a negative response to
6532 MAIL if the sender used is not empty.
6534 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6537 (c) A new callout option, "use_sender" has been added for people who want
6538 the modified behaviour.
6544 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6547 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6548 indeed breaks things for older releases.
6550 3. Added additional logging to the case where there is a problem reading data
6551 from a filter that is running in a subprocess using a pipe, in order to
6552 try to track down a specific problem.
6554 4. Testing facility fudge: when running in the test harness and attempting
6555 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6556 getting "No route to host". Convert this to a timeout.
6558 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6561 6. Some OS don't have socklen_t but use size_t instead. This affects the
6562 fifth argument of getsockopt() amongst other things. This is now
6563 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6564 can be set for individual OS. I have set it for SunOS5, OSF1, and
6565 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6566 some earlier ones do not.
6568 7. Change 4.30/15 was not doing the test caselessly.
6570 8. The standard form for an IPv6 address literal was being rejected by address
6571 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6572 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6573 this, as well as the form without the "IPv6" on the front (but only when
6574 address literals are enabled, of course).
6576 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6578 10. Exim crashed if a message with an empty sender address specified by -f
6579 encountered a router with an errors_to setting. This could be provoked only
6580 by a command such as
6584 where an empty string was supplied; "<>" did not hit this bug.
6586 11. Installed PCRE release 4.5.
6588 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6589 remained set. It is now erased.
6591 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6592 times from message ids (which are base 36 rather than the normal 62).
6594 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6595 were being counted as actual protocol errors, and logged if the log
6596 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6597 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6598 and DATA following a set of rejected RCPTs do not count as protocol errors.
6599 In other words, Exim assumes they were pipelined, though this may not
6600 actually be the case. Of course, in all cases the client gets an
6601 appropriate error code.
6603 15. If a lookup fails in an ACL condition, a message about the failure may
6604 be available; it is used if testing the ACL cannot continue, because most
6605 such messages specify what the cause of the deferral is. However, some
6606 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6607 that caused an old message to be retained and used if a later statement
6608 caused a defer, replacing the real cause of the deferral.
6610 16. If an IP address had so many PTR records that the DNS lookup buffer
6611 was not large enough to hold them, Exim could crash while trying to process
6612 the truncated data. It now detects and logs this case.
6614 17. Further to 4.21/58, another change has been made: if (and only if) the
6615 first line of a message (the first header line) ends with CRLF, a bare LF
6616 in a subsequent header line has a space inserted after it, so as not to
6617 terminate the header.
6619 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6620 unnecessarily, used atoi() instead of strtol(), and didn't check the
6621 termination when getting file sizes from file names by regex.
6623 19. Completely re-implemented the support for maildirsize files, in the light
6624 of a number of problems with the previous contributed implementation
6625 (4.30/29). In particular:
6627 . If the quota is zero, the maildirsize file is maintained, but no quota is
6630 . If the maildir directory does not exist, it is created before any attempt
6631 to write a maildirsize file.
6633 . The quota value in the file is just a cache; if the quota is changed in
6634 the transport, the new value overrides.
6636 . A regular expression is available for excluding directories from the
6639 20. The autoreply transport checks the characters in options that define the
6640 message's headers; it allows continued headers, but it was checking with
6641 isspace() after an embedded newline instead of explicitly looking for a
6644 21. If all the "regular" hosts to which an address was routed had passed their
6645 expiry times, and had not reached their retry times, the address was
6646 bounced, even if fallback hosts were defined. Now Exim should go on to try
6649 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6650 equivalent code in the SMTP transport. Some hosts send humungous responses
6651 to HELO/EHLO, more than 1024 it seems.
6653 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6654 gives compiler warnings in some environments. I've now done it "properly",
6657 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6658 (because of problems with the built-in one) was declared to return uschar *
6659 instead of char *, causing compiler failure.
6661 25. Fixed a file descriptor leak when processing alias/forward files.
6663 26. Fixed a minor format string issue in dbfn.c.
6665 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6667 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6668 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6671 29. When a sender address is verified, it is cached, to save repeating the test
6672 when there is more than one recipient in a message. However, when the
6673 verification involves a callout, it is possible for different callout
6674 options to be set for different recipients. It is too complicated to keep
6675 track of this in the cache, so now Exim always runs a verification when a
6676 callout is required, relying on the callout cache for the optimization.
6677 The overhead is duplication of the address routing, but this should not be
6680 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6681 to be verified with callout=postmaster, and the main callout worked but the
6682 postmaster check failed, the verification correctly failed. However, if a
6683 subsequent RCPT command asked for sender verification *without* the
6684 postmaster check, incorrect caching caused this verification also to fail,
6687 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6688 it was not caching the DNS options (qualify_single, search_parents) that
6689 were used when the lookup failed. A subsequent lookup with different
6690 options therefore always gave the same answer, though there were cases
6691 where it should not have. (Example: a "domains = !$mx_any" option on a
6692 dnslookup router: the "domains" option is always processed without any
6693 widening, but the router might have qualify_single set.) Now Exim uses the
6694 cached value only when the same options are set.
6696 32. Added John Jetmore's "exipick" utility to the distribution.
6698 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6699 than a timeout (e.g. a certificate is required, and is not provided), an
6700 Exim server now closes the connection immediately. Previously it waited for
6701 the client to close - but if the client is SSL, it seems that they each
6702 wait for each other, leading to a delay before one of them times out.
6704 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6705 maintained 0.8.x compatibility because I don't think many are using it, and
6706 it is clearly obsolete.
6708 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6711 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6712 client certificate was expired. A simple patch fixes this, though I don't
6713 understand the full logic of why the verify callback is called multiple
6716 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6717 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6718 which causes problems with some clients (such as the Certicom SSL Plus
6719 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6720 disables the coutermeasure allowing Eudora to connect."
6722 38. Exim was not checking that a write() to a log file succeeded. This could
6723 lead to Bad Things if a log got too big, in particular if it hit a file
6724 size limit. Exim now panics and dies if it cannot write to a log file, just
6725 as it does if it cannot open a log file.
6727 39. Modified OS/Makefile-Linux so that it now contains
6729 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6731 The two -D definitions ensure that Exim is compiled with large file
6732 support, which makes it possible to handle log files that are bigger than
6735 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6736 instance) a domain was checked against a named list that involved a lookup,
6737 causing $domain_data to be set, then another domain was checked against the
6738 same list, then the first domain was re-checked, the value of $domain_data
6739 after the final check could be wrong. In particular, if the second check
6740 failed, it could be set empty. This bug probably also applied to
6743 41. The strip_trailing_dot option was not being applied to the address given
6744 with the -f command-line option.
6746 42. The code for reading a message's header from the spool was incrementing
6747 $received_count, but never initializing it. This meant that the value was
6748 incorrect (doubled) while delivering a message in the same process in which
6749 it was received. In the most common configuration of Exim, this never
6750 happens - a fresh exec is done - but it can happen when
6751 deliver_drop_privilege is set.
6753 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6754 it now includes up to 150 characters of the unexpected data in the log
6757 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6758 and building data strings. The size of both of these buffers was 10 000
6759 bytes - far larger than anybody would *ever* want, thought I. Needless to
6760 say, somebody hit the limit. I have increased the maximum line length to
6761 20 000 and the maximum data length of concatenated lines to 100 000. I have
6762 also fixed two bugs, because there was no checking on these buffers. Tsk,
6763 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6764 buffer is too small.
6766 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6767 lsearch lookups. Now it does.
6769 46. When parsing a route_list item in a manualroute router, a fixed-length
6770 buffer was used for the list of hosts. I made this 1024 bytes long,
6771 thinking that nobody would ever have a list of hosts that long. Wrong.
6772 Somebody had a whole pile of complicated expansion conditions, and the
6773 string was silently truncated, leading to an expansion error. It turns out
6774 that it is easier to change to an unlimited length (owing to other changes
6775 that have happened since this code was originally written) than to build
6776 structure for giving a limitation error. The length of the item that
6777 expands into the list of hosts is now unlimited.
6779 47. The lsearch lookup could not handle data where the length of text line was
6780 more than 4095 characters. Such lines were truncated, leading to shortened
6781 data being returned. It should now handle lines of any length.
6783 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6784 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6787 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6788 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6789 no longer happens in comment lines. A list of replacements is now placed
6790 at the head of all of the source files, except those whose only change is
6791 to replace PERL_COMMAND in the very first #! line.
6793 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6794 messages on the queue, with a bottom-up merge sort, using code contributed
6795 by Michael Haardt. This should make operations like -bp somewhat faster on
6796 large queues. It won't affect queue runners, except when queue_run_in_order
6799 51. Installed eximstats 1.31 in the distribution.
6801 52. Added support for SRV lookups to the dnslookup router.
6803 53. If an ACL referred to $message_body or $message_body_end, the value was not
6804 reset for any messages that followed in the same SMTP session.
6806 54. The store-handling optimization for building very long strings was not
6807 differentiating between the different store pools. I don't think this
6808 actually made any difference in practice, but I've tidied it.
6810 55. While running the routers to verify a sender address, $sender_address
6811 was still set to the sender address. This is wrong, because when routing to
6812 send a bounce to the sender, it would be empty. Therefore, I have changed
6813 it so that, while verifying a sender address, $sender_address is set to <>.
6814 (There is no change to what happens when verifying a recipient address.)
6816 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6817 target A or AAAA records (if not already returned) without resetting the
6818 qualify_single or search_parents options of the DNS resolver. These are
6819 inappropriate in this case because the targets of MX and SRV records must
6820 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6821 target that, when qualified, matched something in the local domain. These
6822 two options are now turned off when doing these lookups.
6824 57. It seems that at least some releases of Reiserfs (which does not have the
6825 concept of a fixed number of inodes) returns zero and not -1 for the
6826 number of available inodes. This interacted badly with check_spool_inodes,
6827 which assumed that -1 was the "no such thing" setting. What I have done is
6828 to check that the total number of inodes is greater than zero before doing
6829 the test of how many are available.
6831 58. When a "warn" ACL statement has a log_message modifier, the message is
6832 remembered, and not repeated. This is to avoid a lot of repetition when a
6833 message has many recipients that cause the same warning to be written.
6834 However, Exim was preserving the list of already written lines for an
6835 entire SMTP session, which doesn't seem right. The memory is now reset if a
6836 new message is started.
6838 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6839 debugging output unless log_rewrite was also set.
6841 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6842 of (int)(handle) when we know that handle contains (void *)(-1).
6844 61. The Exim daemon panic-logs an error return when it closes the incoming
6845 connection. However "connection reset by peer" seems to be common, and
6846 isn't really an error worthy of noting specially, so that particular error
6849 62. When Exim is trying to find all the local interfaces, it used to panic and
6850 die if the ioctl to get the interface flags failed. However, it seems that
6851 on at least one OS (Solaris 9) it is possible to have an interface that is
6852 included in the list of interfaces, but for which you get a failure error
6853 for this call. This happens when the interface is not "plumbed" into a
6854 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6855 failure of the "get flags" call assumes that the interface is down.
6857 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6858 makes life easier for people who are doing arithmetic on fields extracted
6859 from dates, where you often get leading zeros that should not be
6860 interpreted as octal.
6862 64. Added qualify_domain to the redirect router, to override the global
6865 65. If a pathologically long header line contained very many addresses (the
6866 report of this problem mentioned 10 000) and each of them was rewritten,
6867 Exim could use up a very large amount of memory. (It kept on making new
6868 copies of the header line as it rewrote, and never released the old ones.)
6869 At the expense of a bit more processing, the header rewriting function has
6870 been changed so that it no longer eats memory in this way.
6872 66. The generation of the Received: header has been moved from the time that a
6873 message starts to be received, to the time that it finishes. The timestamp
6874 in the Received: header should now be very close to that of the <= log
6875 line. There are two side-effects of this change:
6877 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6878 logged header lines no longer include the local Received: line, because
6879 it has not yet been created. The same applies to a copy of the message
6880 that is returned to a non-SMTP sender when a message is rejected.
6882 (b) When a filter file is tested using -bf, no additional Received: header
6883 is added to the test message. After some thought, I decided that this
6886 This change does not affect the value of $received_for. It is still set
6887 after address rewriting, but before local_scan() is called.
6889 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6891 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6892 gave an unhelpful panic error message, and a defer error. I have managed to
6893 change this behaviour so that it now rejects any supplied certificate,
6894 which seems right, as the list of acceptable certificates is empty.
6896 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6897 gave an unhelpful defer error. I have not managed to make this reject any
6898 supplied certificates, but the error message it gives is "no certificate
6899 supplied", which is not helpful.
6901 70. exigrep's output now also includes lines that are not associated with any
6902 message, but which match the given pattern. Implemented by a patch from
6903 Martin Sluka, which also tidied up the Perl a bit.
6905 71. Recipient callout verification, like sender verification, was using <> in
6906 the MAIL FROM command. This isn't really the right thing, since the actual
6907 sender may affect whether the remote host accepts the recipient or not. I
6908 have changed it to use the actual sender in the callout; this means that
6909 the cache record is now keyed on a recipient/sender pair, not just the
6910 recipient address. There doesn't seem to be a real danger of callout loops,
6911 since a callout by the remote host to check the sender would use <>.
6912 [SEE ABOVE: changed after hitting problems.]
6914 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6915 temporary errors. However, in the case of such a code being given after
6916 the end of a data transmission (i.e. after ".") Exim was failing to write
6917 a retry record for the message. (Yes, there was some broken host that was
6918 actually sending 8xx at this point.)
6920 73. An unknown lookup type in a host list could cause Exim to panic-die when
6921 the list was checked. (An example that provoked this was putting <; in the
6922 middle of a list instead of at the start.) If this happened during a DATA
6923 ACL check, a -D file could be left lying around. This kind of configuration
6924 error no longer causes Exim to die; instead it causes a defer error. The
6925 incident is still logged to the main and panic logs.
6927 74. Buglet left over from Exim 3 conversion. The message "too many messages
6928 in one connection" was written to the rejectlog but not the mainlog, except
6929 when address rewriting (yes!) was being logged.
6931 75. Added write_rejectlog option.
6933 76. When a system filter was run not as root (that is, when system_filter_user
6934 was set), the values of the $n variables were not being returned to the
6935 main process; thus, they were not subsequently available in the $sn
6938 77. Added +return_path_on_delivery log selector.
6940 78. A connection timeout was being treated differently from recipients deferred
6941 when testing hosts_max_try with a message that was older than the host's
6942 retry timeout. (The host should not be counted, thus allowing all hosts to
6943 be tried at least once before bouncing.) This may have been the cause of an
6944 occasionally reported bug whereby a message would remain on the queue
6945 longer than the retry timeout, but would be bounced if a delivery was
6946 forced. I say "may" because I never totally pinned down the problem;
6947 setting up timeout/retry tests is difficult. See also the next item.
6949 79. The ultimate address timeout was not being applied to errors that involved
6950 a combination of host plus message (for example, a timeout on a MAIL
6951 command). When an address resolved to a number of possible hosts, and they
6952 were not all tried for each delivery (e.g. because of hosts_max_try), a
6953 message could remain on the queue longer than the retry timeout.
6955 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6958 81. Fixed an obscure SMTP outgoing bug which required at least the following
6959 conditions: (a) there was another message waiting for the same server;
6960 (b) the server returned 5xx to all RCPT commands in the first message so
6961 that the message was not completed; (c) the server dropped the connection
6962 or gave a negative response to the RSET that Exim sends to abort the
6963 transaction. The observed case was a dropped connection after DATA that had
6964 been sent in pipelining mode. That is, the server had advertised PIPELINING
6965 but was not implementing it correctly. The effect of the bug was incorrect
6966 behaviour, such as trying another host, and this could lead to a crash.
6972 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6973 and daemon.c were passed as pointers to ints; they should have been
6974 pointers to socklen_t variables (which are typically unsigned ints).
6976 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6979 3. Fixed a really odd bug that affected only the testing scheme; patching a
6980 certain fixed string in the binary changed the value of another string that
6981 happened to be identical to the end of the original first string.
6983 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6984 name", it returns that address as the IP address. On some operating
6985 systems (e.g. Solaris), it also passes back the IP address string as the
6986 "host name". However, on others (e.g. Linux), it passes back an empty
6987 string. Exim wasn't checking for this, and was changing the host name to an
6988 empty string, assuming it had been canonicalized.
6990 5. Although rare, it is permitted to have more than one PTR record for a given
6991 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6992 all the names associated with an address, because they do in Solaris.
6993 However, it seems that they do not in Linux for data that comes from the
6994 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6995 I found this out when I moved to a new Linux workstation and tried to run
6996 the Exim test suite.
6998 To get round this problem I have changed the code so that it now does its
6999 own call to the DNS to look up PTR records when searching for a host name.
7000 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7001 addresses that are only in /etc/hosts are still found.
7003 This behaviour is, however, controlled by an option called host_lookup_
7004 order, which defaults to "bydns:byaddr". If people want to use the other
7005 order, or indeed, just use one or the other means of lookup, they can
7006 specify it in this variable.
7008 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7009 some operating systems, this comes back from gethostbyaddr() as an empty
7010 string, and this is what Exim used to test for. However, it seems that in
7011 other systems, "." is yielded. Exim now tests for this case too.
7013 7. The values of check_spool_space and check_log_space are now held internally
7014 as a number of kilobytes instead of an absolute number of bytes. If a
7015 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7016 kilobyte. This means that much larger values can be stored.
7018 8. Exim monitor: an attempt to get the action menu when not actually pointing
7019 at a message produces an empty menu entitled "No message selected". This
7020 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7021 no entries in it ("Shell widget menu has zero width and/or height"). So I
7022 have added a single, blank menu entry in this case.
7024 9. Added ${quote_local_part.
7026 10. MIME decoding is now applied to the contents of Subject: header lines when
7029 11. Now that a reference to $sender_host_address automatically causes a reverse
7030 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7031 host lookup before query-style lookups in lists that might use this
7032 variable. This has therefore been abolished, and the "net-" prefix is no
7033 longer necessary for query-style lookups.
7035 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7036 have been a typo for LFLAGS, so it has been changed.
7038 13. The install script calls Exim with "-C /dev/null" in order to find the
7039 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7040 to be output. However, since Exim outputs its version number before the
7041 error, it didn't break the script. It just looked ugly. I fixed this by
7042 always allowing "-C /dev/null" if the caller is root.
7044 14. Ignore overlarge ACL variable number when reading spool file - insurance
7045 against a later release with more variables having written the file.
7047 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7048 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7049 this, as well as the form without the "IPv6" on the front.
7051 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7052 OS/Makefile-Darwin file.
7054 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7055 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7057 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7058 variable. However, some OS use ints for inodes. I've added cast to long int
7059 to get rid of the compiler warning.
7061 19. I had forgotten to lock out "/../" in configuration file names when
7062 ALT_CONFIG_PREFIX was set.
7064 20. Routers used for verification do not need to specify transports. However,
7065 if such a router generated a host list, and callout was configured, Exim
7066 crashed, because it could not find a port number from the (non-existent)
7067 transport. It now assumes port 25 in this circumstance.
7069 21. Added the -t option to exigrep.
7071 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7072 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7073 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7074 them is set, LOOKUP_LSEARCH is forced.
7076 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7078 24. Added sender and host information to the "rejected by local_scan()" log
7079 line; previously there was no indication of these.
7081 25. Added .include_if_exists.
7083 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7084 Linux. It turns out that not all file systems support this. Apparently some
7085 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7086 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7087 not supported on the file descriptor, is now ignored when Exim is trying to
7088 sync a directory. This applies only to Linux.
7090 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7092 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7093 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7096 29. Added support for maildirsize files from supplied patch (modified a bit).
7098 30. The use of :fail: followed by an empty string could lead Exim to respond to
7099 sender verification failures with (e.g.):
7101 550 Verification failed for <xxx>
7102 550 Sender verify failed
7104 where the first response line was missing the '-' that indicates it is not
7105 the final line of the response.
7107 31. The loop for finding the name of the user that called Exim had a hardwired
7108 limit of 10; it now uses the value of finduser_retries, which is used for
7109 all other user lookups.
7111 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7114 33. Exim was neglecting to zero errno before one call of strtol() when
7115 expanding a string and expecting an integer value. On some systems this
7116 resulted in spurious "integer overflow" errors. Also, it was casting the
7117 result into an int without checking.
7119 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7120 did not work. The code looks as if it has *never* worked, though it appears
7121 to have been documented since at least release 1.62. I have made it work.
7123 35. The "timeout_DNS" error in retry rules, also documented since at least
7124 1.62, also never worked. As it isn't clear exactly what this means, and
7125 clearly it isn't a major issue, I have abolished the feature by treating it
7126 as "timeout", and writing a warning to the main and panic logs.
7128 36. The display of retry rules for -brt wasn't always showing the error code
7131 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7132 timeout_connect_A, timeout_connect_MX.
7134 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7135 to the empty sender.
7137 39. The daemon was not analysing the content of -oX till after it had closed
7138 stderr and disconnected from the controlling terminal. This meant that any
7139 syntax errors were only noted on the panic log, and the return code from
7140 the command was 0. By re-arranging the code a little, I've made the
7141 decoding happen first, so such errors now appear on stderr, and the return
7142 code is 1. However, the actual setting up of the sockets still happens in
7143 the disconnected process, so errors there are still only recorded on the
7146 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7147 connections (as happens on some IP stacks) was logged at start up time as
7148 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7149 it from "IPv6 and IPv4", which means that two separate sockets are being
7152 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7153 says whether AF_INET or AF_INET6 was passed as an argument.
7155 42. Exiwhat output was messed up when time zones were included in log
7158 43. Exiwhat now gives more information about the daemon's listening ports,
7159 and whether -tls-on-connect was used.
7161 44. The "port" option of the smtp transport is now expanded.
7163 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7164 silently ignored. Now an error message is written to the main and panic
7167 46. There's a new ACL modifier called "logwrite" which writes to a log file
7168 as soon as it is encountered.
7170 47. Added $local_user_uid and $local_user_gid at routing time.
7172 48. Exim crashed when trying to verify a sender address that was being
7175 49. Exim was recognizing only a space character after ".include". It now also
7176 recognizes a tab character.
7178 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7179 extracting the relevant information from the specification. The man page no
7180 longer contains scrambled data for the -d option, and I've added a section
7181 at the front about calling Exim under different names.
7183 51. Added "extra_headers" argument to the "mail" command in filter files.
7185 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7188 53. Installed eximstats 1.29.
7190 54. Added transport_filter_timeout as a generic transport option.
7192 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7193 Cc: header lines. This was required by RFC 822, but it not required by RFC
7196 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7197 incoming messages that did not have them. Now it does so only if the
7198 message originates locally, that is, if there is no associated remote host
7199 address. When Resent- header lines are present, this applies to the Resent-
7200 lines rather than the non-Resent- lines.
7202 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7203 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7205 58. Messages for configuration errors now include the name of the main
7206 configuration file - useful now that there may be more than one file in a
7207 list (.included file names were always shown).
7209 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7210 for those rare installations that do not start the daemon as root or run it
7211 setuid root. I've cut out the call to initgroups() if the daemon is not
7214 60. The Exim user and group can now be bound into the binary as text strings
7215 that are looked up at the start of Exim's processing.
7217 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7219 62. Added $mailstore_basename variable.
7221 63. Installed patch to sieve.c from Michael Haardt.
7223 64. When Exim failed to open the panic log after failing to open the main log,
7224 the original message it was trying to log was written to stderr and debug
7225 output, but if they were not available (the usual case in production), it
7226 was lost. Now it is written to syslog before the two lines that record the
7227 failures to open the logs.
7229 65. Users' Exim filters run in subprocesses under the user's uid. It is
7230 possible for a "deliver" command or an alias in a "personal" command to
7231 provoke an address rewrite. If logging of address rewriting is configured,
7232 this fails because the process is not running as root or exim. There may be
7233 a better way of dealing with this, but for the moment (because 4.30 needs
7234 to be released), I have disabled address rewrite logging when running a
7235 filter in a non-root, non-exim process.
7241 1. The buildconfig auxiliary program wasn't quoting the value set for
7242 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7243 not defined. This bug was masked in 4.22 by the effect that was fixed in
7246 2. Some messages that were rejected after a message id was allocated were
7247 shown as "incomplete" by exigrep. It no longer does this for messages that
7248 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7250 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7251 have allow_domain_literals set, the ID did not get logged in the <= line.
7252 Domain literals are now always recognized in Message-ID: header lines.
7254 4. The first argument for a ${extract expansion item is the key name or field
7255 number. Leading and trailing spaces in this item were not being ignored,
7256 causing some misleading effects.
7258 5. When deliver_drop_privilege was set, single queue runner processes started
7259 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7260 same command in the process it spins off) were not dropping privilege.
7262 6. When the daemon running as "exim" started a queue runner, it always
7263 re-executed Exim in the spun-off process. This is a waste of effort when
7264 deliver_drop_privilege is set. The new process now just calls the
7265 queue-runner function directly.
7271 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7274 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7275 ignored. Though the use of -oP was forcing the writing of a pid file, it
7276 was always written to the default place.
7278 3. If the message "no IP address found for host xxxx" is generated during
7279 incoming verification, it is now followed by identification of the incoming
7280 connection (so you can more easily find what provoked it).
7282 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7284 5. Added some features to "harden" Exim a bit more against certain attacks:
7286 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7287 be put in Local/Makefile. This is like the never_users runtime option,
7288 but it cannot be overridden. The default setting is "root".
7290 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7291 prefix string with which any file named in a -C command line option
7294 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7295 is retained for -C and -D only if the caller of Exim is root. Without
7296 it, the exim user may also use -C and -D and retain privilege.
7298 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7299 command line option is disabled.
7301 6. Macro names set by the -D option must start with an upper case letter, just
7302 like macro names defined in the configuration file.
7304 7. Added "dereference=" facility to LDAP.
7306 8. Two instances of the typo "uknown" in the source files are fixed.
7308 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7309 the Configure-Makefile script screwed up while processing it.
7311 10. Incorporated PCRE 4.4.
7313 11. The SMTP synchronization check was not operating right at the start of an
7314 SMTP session. For example, it could not catch a HELO sent before the client
7315 waited for the greeting. There is now a check for outstanding input at the
7316 point when the greeting is written. Because of the duplex, asynchronous
7317 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7318 way, but not yet received, when the check is performed.
7320 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7321 on TCP/IP sockets, because this apparently causes some broken clients to
7324 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7325 unchanged) from the Cygwin maintainer.
7327 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7328 was supported instead of testing for mbx. Effectively a typo.
7330 15. The spa authenticator server code was not checking that the input it
7331 received was valid base64.
7333 16. The debug output line for the "set" modifier in ACLs was not showing the
7334 name of the variable that was being set.
7336 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7338 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7339 lookup on its own. Something else was needed to trigger the lookup. For
7340 example, a match in host_lookup or the need for a host name in a host list.
7341 Now, if $sender_host_name is referenced and the host name has not yet been
7342 looked up, a lookup is performed. If the lookup fails, the variable remains
7343 empty, and $host_lookup_failed is set to "1".
7345 19. Added "eqi" as a case-independent comparison operator.
7347 20. The saslauthd authentication condition could segfault if neither service
7348 nor realm was specified.
7350 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7351 error message that was logged was misleading, and incoming SMTP
7352 connections were dropped. The message is now more accurate, and temporary
7353 errors are given to SMTP connections.
7355 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7356 (except RSET) with a 421 error, until QUIT is received. However, it was
7357 failing to send a response to QUIT.
7359 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7360 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7361 ACL is now run after the helo_try_verify_hosts code.
7363 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7364 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7365 case-independent, so other case variants are also recognized). Apparently
7366 some systems use these upper case variants.
7368 25. If more than two messages were waiting for the same host, and a transport
7369 filter was specified for the transport, Exim sent two messages over the
7370 same TCP/IP connection, and then failed with "socket operation on non-
7371 socket" when it tried to send the third.
7373 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7375 27. The extern definition of crypt16() in expand.c was not being excluded when
7376 the OS had its own crypt16() function.
7378 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7379 as a preferred synonym for return_size_limit, both as an option and as an
7382 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7384 30. Changed the default configuration ACL to relax the local part checking rule
7385 for addresses that are not in any local domains. For these addresses,
7386 slashes and pipe symbols are allowed within local parts, but the sequence
7387 /../ is explicitly forbidden.
7389 31. SPA server authentication was not clearing the challenge buffer before
7392 32. log_message in a "warn" ACL statement was writing to the reject log as
7393 well as to the main log, which contradicts the documentation and doesn't
7394 seem right (because no rejection is happening). So I have stopped it.
7396 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7397 However, I am unable to do any testing of this.
7399 34. Fixed an infelicity in the appendfile transport. When checking directories
7400 for a mailbox, to see if any needed to be created, it was accidentally
7401 using path names with one or more superfluous leading slashes; tracing
7402 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7404 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7405 contents are added to the log line that is written for every discarded
7406 recipient. (Previously a log_message setting was ignored.)
7408 36. The ${quote: operator now quotes the string if it is empty.
7410 37. The install script runs exim in order to find its version number. If for
7411 some reason other than non-existence or emptiness, which it checks, it
7412 could not run './exim', it was installing it with an empty version number,
7413 i.e. as "exim-". This error state is now caught, and the installation is
7416 38. An argument was missing from the function that creates an error message
7417 when Exim fails to connect to the socket for saslauthd authentication.
7418 This could cause Exim to crash, or give a corrupted message.
7420 39. Added isip, isip4, and isip6 to ${if conditions.
7422 40. The ACL variables $acl_xx are now saved with the message, and can be
7423 accessed later in routers, transports, and filters.
7425 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7426 strings in the file are not string-expanded.
7428 42. If a MAIL command specified a SIZE value that was too large to fit into an
7429 int variable, the check against message_size_limit failed. Such values are
7430 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7431 day this will have to be increased, but I don't think I want to be around
7432 when emails are that large.
7439 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7440 iconv() is not standard in FreeBSD.
7442 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7443 IPv6 enabled. The observed symptom was a segmentation fault on return from
7444 the function os_common_find_running_interfaces() in src/os.c.
7446 3. In the check_special_case() function in daemon.c I had used "errno" as an
7447 argument name, which causes warnings on some systems. This was basically a
7448 typo, since it was named "eno" in the comments!
7450 4. The code that waits for the clock to tick (at a resolution of some fraction
7451 of a second) so as to ensure message-id uniqueness was always waiting for
7452 at least one whole tick, when it could have waited for less. [This is
7453 almost certainly not relevant at current processor speeds, where it is
7454 unlikely to ever wait at all. But we try to future-proof.]
7456 5. The function that sleeps for a time interval that includes fractions of a
7457 second contained a race. It did not block SIGALRM between setting the
7458 timer, and suspending (a couple of lines later). If the interval was short
7459 and the sigsuspend() was delayed until after it had expired, the suspension
7460 never ended. On busy systems this could lead to processes getting stuck for
7463 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7464 process, before it forks any delivery processes. The open lookup caching
7465 mechanism meant that the open file or database connection was passed into
7466 the delivery process. The problem was that delivery processes always tidy
7467 up cached lookup data. This could cause a problem for the next delivery
7468 process started by the queue runner, because the external queue runner
7469 process does not know about the closure. So the next delivery process
7470 still has data in the lookup cache. In the case of a file lookup, there was
7471 no problem because closing a file descriptor in a subprocess doesn't affect
7472 the parent. However, if the lookup was caching a connection to a database,
7473 the connection was closed, and the second delivery process was likely to
7474 see errors such as "PGSQL: query failed: server closed the connection
7475 unexpectedly". The problem has been fixed by closing all cached lookups
7476 in a queue runner before running a delivery process.
7478 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7479 seem to have the "const" qualifier which it has on other OS. I've
7482 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7483 *of the same type* (client or server) with the same public name that an
7484 error should be diagnosed.
7486 9. When Exim looked up a host name for an IP address, but failed to find the
7487 original IP address when looking up the host name (a safety check), it
7488 output the message "<ip address> does not match any IP for NULL", which was
7489 confusing, to say the least. The bug was that the host name should have
7490 appeared instead of "NULL".
7492 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7493 user that is built into the binary, and the -C or -D options is used, root
7494 privilege is dropped before the configuration file is read. In addition,
7495 logging is switched to stderr instead of the normal log files. If the
7496 configuration then re-defines the Exim user, the unprivileged environment
7497 is probably not what is expected, so Exim logs a panic warning message (but
7500 However, if deliver_drop_privilege is set, the unprivileged state may well
7501 be exactly what is intended, so the warning has been cut out in that case,
7502 and Exim is allowed to try to write to its normal log files.
7508 1. smtp_return_error_details was not giving details for temporary sender
7509 or receiver verification errors.
7511 2. Diagnose a configuration error if two authenticators have the same public
7514 3. Exim used not to create the message log file for a message until the first
7515 delivery attempt. This could be confusing when incoming messages were held
7516 for policy or load reasons. The message log file is now created at the time
7517 the message is received, and an initial "Received" line is written to it.
7519 4. The automatically generated man page for command line options had a minor
7520 bug that caused no ill effects; however, a more serious problem was that
7521 the procedure for building the man page automatically didn't always
7522 operate. Consequently, release 4.20 contains an out-of-date version. This
7523 shouldn't happen again.
7525 5. When building Exim with embedded Perl support, the script that builds the
7526 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7527 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7529 6. The freeze_tell option was not being used for messages that were frozen on
7530 arrival, either by an ACL or by local_scan().
7532 7. Added the smtp_incomplete_transaction log selector.
7534 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7535 was accepting AUTH without a new EHLO.
7537 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7538 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7541 10. Logging of TCP/IP connections (when configured) now happens in the main
7542 daemon process instead of the child process, so that the TCP/IP connection
7543 count is more accurate (but it can never be perfect).
7545 11. The use of "drop" in a nested ACL was not being handled correctly in the
7546 outer ACL. Now, if condition failure induced by the nested "drop" causes
7547 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7548 or "require"), the connection is dropped.
7550 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7551 that yield "discard" can now be used with an "accept" or a "discard" verb,
7552 but an error is generated for any others (because I can't see a useful way
7553 to define what should happen).
7555 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7556 are now processed in the same way as lines in the Exim configuration file.
7557 In particular, continuation lines are supported.
7559 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7561 15. Added -ti meaning -t -i.
7563 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7564 domains, and warn by logging if others are found.
7566 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7567 returned from GIFCONF when Exim is trying to find the list of interfaces on
7568 a host. The code in os.c has been modified to copy each ifreq to an aligned
7569 structure in all cases.
7571 Also, in some cases, the returned ifreq's were being copied to a 'struct
7572 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7573 means the last couple of bytes of an IPv6 address could be chopped if the
7574 ifreq contained only a normal sockaddr (14 bytes storage).
7576 18. Named domain lists were not supported in the hosts_treat_as_local option.
7577 An entry such as +xxxx was not recognized, and was treated as a literal
7580 19. Ensure that header lines added by a DATA ACL are included in the reject log
7581 if the ACL subsequently rejects the message.
7583 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7584 MD5 (which is deprecated).
7586 21. When testing a filter file using -bf, Exim was writing a message when it
7587 took the sender from a "From " line in the message, but it was not doing so
7588 when it took $return_path from a Return-Path: header line. It now does.
7590 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7591 with a valid header line field name (a series of printing characters
7592 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7594 23. Changed "disc" in the source to "disk" to conform to the documentation and
7595 the book and for uniformity.
7597 24. Ignore Sendmail's -Ooption=value command line item.
7599 25. When execve() failed while trying to run a command in a pipe transport,
7600 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7601 could be confused with a return value of 69 from the command itself. This
7602 has been changed to 127, the value the shell returns if it is asked to run
7603 a non-existent command. The wording for the related log line suggests a
7604 non-existent command as the problem.
7606 26. If received_header_text expands to an empty string, do not add a Received:
7607 header line to the message. (Well, it adds a token one on the spool, but
7608 marks it "old" so that it doesn't get used or transmitted.)
7610 27. Installed eximstats 1.28 (addition of -nt option).
7612 28. There was no check for failure on the call to getsockname() in the daemon
7613 code. This can fail if there is a shortage of resources on the system, with
7614 ENOMEM, for example. A temporary error is now given on failure.
7616 29. Contrary to the C standard, it seems that in some environments, the
7617 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7618 program. Exim now does this explicitly; it affects the formatting of
7619 timestamps using strftime().
7621 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7622 complaints. I've now initialized all the variables, to avoid this.
7624 32. Header lines added by a system filter were not being "seen" during
7625 transport-time rewrites.
7627 33. The info_callback() function passed to OpenSSL is set up with type void
7628 (*)(SSL *, int, int), as described somewhere. However, when calling the
7629 function (actually a macro) that sets it up, the type void(*)() is
7630 expected. I've put in a cast to prevent warnings from picky compilers.
7632 34. If a DNS black list lookup found a CNAME record, but there were no A
7633 records associated with the domain it pointed at, Exim crashed.
7635 35. If a DNS black list lookup returned more than one A record, Exim ignored
7636 all but the first. It now scans all returned addresses if a particular IP
7637 value is being sought. In this situation, the contents of the
7638 $dnslist_value variable are a list of all the addresses, separated by a
7641 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7642 to accept a host name and all its aliases if the forward lookup for any of
7643 them yielded the IP address of the incoming connection. Now it accepts only
7644 those names whose forward lookup yields the correct IP address. Any other
7645 names are discarded. This closes a loophole whereby a rogue DNS
7646 administrator could create reverse DNS records to break through a
7647 wildcarded host restriction in an ACL.
7649 37. If a user filter or a system filter that ran in a subprocess used any of
7650 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7651 the wrong values were passed to the pipe command ($thisaddress had the
7652 value of $0, $0 had the value of $1, etc). This bug was introduced by
7653 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7655 38. Improved the line breaking for long SMTP error messages from ACLs.
7656 Previously, if there was no break point between 40 and 75 characters, Exim
7657 left the rest of the message alone. Two changes have been made: (a) I've
7658 reduced the minimum length to 35 characters; (b) if it can't find a break
7659 point between 35 and 75 characters, it looks ahead and uses the first one
7660 that it finds. This may give the occasional overlong line, but at least the
7661 remaining text gets split now.
7663 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7664 file descriptors might be low, and that setting 1000 would always raise it.
7665 It turns out that in some environments, the limit is already over 1000 and
7666 that lowering it causes trouble. So now Exim takes care not to decrease it.
7668 40. When delivering a message, the value of $return_path is set to $sender_
7669 address at the start of routing (routers may change the value). By an
7670 oversight, this default was not being set up when an address was tested by
7671 -bt or -bv, which affected the outcome if any router or filter referred to
7674 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7675 log only when "message" or "log_message" are set. However, if one of the
7676 conditions was an address verification, or a call to a nested ACL, the
7677 messages generated by the underlying test were being passed through. This
7678 no longer happens. The underlying message is available in $acl_verify_
7679 message for both "message" and "log_message" expansions, so it can be
7680 passed through if needed.
7682 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7683 new expansion $bh_ to give the encoded byte string without charset
7684 translation. Translation happens only if iconv() is available; HAVE_ICONV
7685 indicates this at build time. HEADERS_CHARSET gives the charset to
7686 translate to; headers_charset can change it in the configuration, and
7687 "headers charset" can change it in an individual filter file.
7689 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7690 that creates RFC 2047 encoded "words" labels them as that charset instead
7691 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7692 expansion operator; (ii) when Exim creates a From: line for a local
7693 message; (iii) when a header line is rewritten to include a "phrase" part.
7695 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7696 buggy, causing it to skip the first lines of messages whose message ID
7697 ended in 'D'. This would not have bitten before Exim release 4.14, because
7698 message IDs were unlikely to end in 'D' before then. The effect was to have
7699 incorrect size information for certain domains.
7701 45. #include "config.h" was missing at the start of the crypt16.c module. This
7702 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7705 46. If there was a timeout during a "random" callout check, Exim treated it as
7706 a failure of the random address, and carried on sending RSET and the real
7707 address. If the delay was just some slowness somewhere, the response to the
7708 original RCPT would be taken as a response to RSET and so on, causing
7709 mayhem of various kinds.
7711 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7712 when I implemented it. It didn't allow for the fact that some option values
7713 may legitimately be negative (e.g. size_addition), and it didn't even do
7714 the right test for positive values.
7716 48. Domain names in DNS records are case-independent. Exim always looks them up
7717 in lower case. Some resolvers return domain names in exactly the case they
7718 appear in the zone file, that is, they may contain uppercase letters. Not
7719 all resolvers do this - some return always lower case. Exim was treating a
7720 change of case by a resolver as a change of domain, similar to a widening
7721 of a domain abbreviation. This triggered its re-routing code and so it was
7722 trying to route what was effectively the same domain again. This normally
7723 caused routing to fail (because the router wouldn't handle the domain
7724 twice). Now Exim checks for this case specially, and just changes the
7725 casing of the domain that it ultimately uses when it transmits the message
7728 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7731 50. If a filter generated a file delivery with a non-absolute name (possible if
7732 no home directory exists for the router), the forbid_file option was not
7735 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7736 the existing equality matching.
7738 52. Exim was using ints instead of ino_t variables in some places where it was
7739 dealing with inode numbers.
7741 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7742 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7743 called TMPDIR, and if it finds it is different, it changes its value.
7745 54. The smtp_printf() function is now made available to local_scan() so
7746 additional output lines can be written before returning. There is also an
7747 smtp_fflush() function to enable the detection of a dropped connection.
7748 The variables smtp_input and smtp_batched_input are exported to
7751 55. Changed the default runtime configuration: the message "Unknown user"
7752 has been removed from the ACL, and instead placed on the localuser router,
7753 using the cannot_route_message feature. This means that any verification
7754 failures that generate their own messages won't get overridden. Similarly,
7755 the "Unrouteable address" message that was in the ACL for unverifiable
7756 relay addresses has also been removed.
7758 56. Added hosts_avoid_esmtp to the smtp transport.
7760 57. The exicyclog script was not checking for the esoteric option
7761 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7762 will work only if exicyclog is run under the appropriate euid.
7764 58. Following a discussion on the list, the rules by which Exim recognises line
7765 endings on incoming messages have been changed. The -dropcr and drop_cr
7766 options are now no-ops, retained only for backwards compatibility. The
7767 following line terminators are recognized: LF CRLF CR. However, special
7768 processing applies to CR:
7770 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7771 nor a local message in the state where . is a terminator.
7773 (ii) If a bare CR is encountered in a header line, an extra space is added
7774 after the line terminator so as not to end the header. The reasoning
7775 behind this is that bare CRs in header lines are most likely either
7776 to be mistakes, or people trying to play silly games.
7778 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7779 was being incorrectly given as 18 bytes larger than it should have been.
7780 This is a VOB (very old bug).
7782 60. This may never have affected anything current, but just in case it has:
7783 When the local host is found other than at the start of a list of hosts,
7784 the local host, those with the same MX, and any that follow, are discarded.
7785 When the list in question was part of a longer list of hosts, the following
7786 hosts (not currently being processed) were also being discarded. This no
7787 longer happens. I'm not sure if this situation could ever has previously
7790 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7793 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7794 header. This is something that is recommended in a new Internet Draft, and
7795 is something that is documented as being done by Sendmail. There are two
7796 possible values. For messages generated by the autoreply transport, Exim
7799 Auto-Submitted: auto-replied
7801 whereas for all other generated messages (e.g. bounces) it adds
7803 Auto-Submitted: auto-generated
7805 63. The "personal" condition in filters now includes a test for the
7806 Auto-Submitted: header. If it contains the string "auto-" the message it
7807 not considered personal.
7809 64. Added rcpt_include_affixes as a generic transport option.
7811 65. Added queue_only_override (default true).
7813 66. Added the syslog_duplication option.
7815 67. If what should have been the first header line of a message consisted of
7816 a space followed by a colon, Exim was mis-interpreting it as a header line.
7817 It isn't of course - it is syntactically invalid and should therefore be
7818 treated as the start of the message body. The misbehaviour could have
7819 caused a number of strange effects, including loss of data in subsequent
7820 header lines, and spool format errors.
7822 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7823 client host had authenticated. This control can now be exercised by an ACL
7824 for more flexibility.
7826 69. By default, callouts do not happen when testing with -bh. There is now a
7827 variant, -bhc, which does actually run the callout code, including
7828 consulting and updating the callout cache.
7830 70. Added support for saslauthd authentication, courtesy of Alexander
7833 71. If statvfs() failed on the spool or log directories while checking their
7834 size for availability, Exim confusingly gave the error "space shortage".
7835 Furthermore, in debugging mode it crashed with a floating point exception.
7836 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7837 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7838 serious problem, Exim now writes to the main and panic logs when this
7839 happens, with details of the failure. It then refuses to accept the
7840 incoming message, giving the message "spool directory problem" or "log
7841 directory problem" with a 421 code for SMTP messages.
7843 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7844 0, 1, and 2 exist, because some OS complain for execs without them (see
7845 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7846 descriptors. However, the code omitted to check that the open succeeded,
7847 causing mysterious errors if for some reason the permissions on /dev/null
7848 got screwed. Now Exim writes a message to the main and panic logs, and
7849 bombs out if it can't open /dev/null.
7851 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7852 interact so that it is all more flexible. It is supposed to remain
7853 backwards compatible. Also added extra_local_interfaces.
7855 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7856 to bomb out with an assertion failure - to the client this appears as a
7857 connection drop. This problem occurs in the part of the code that was taken
7858 from the Samba project. Fortunately, the assertion is in a very simple
7859 function, so I have fixed this by reproducing the function inline in the
7860 one place where it is called, and arranging for authentication to fail
7861 instead of killing the process with assert().
7863 75. The SPA client code was not working when the server requested OEM rather
7864 than Unicode encoding.
7866 76. Added code to make require_files with a specific uid setting more usable in
7867 the case where statting the file as root fails - usually a non-root-mounted
7868 NFS file system. When this happens and the failure is EACCES, Exim now
7869 forks a subprocess and does the per-uid checking as the relevant uid.
7871 77. Added process_log_path.
7873 78. If log_file_path was not explicitly set, a setting of check_log_space or
7874 check_log_inodes was ignored.
7876 79. If a space check for the spool or log partitions fails, the incident is now
7877 logged. Of course, in the latter case the data may get lost...
7879 80. Added the %p formatting code to string_format() so that it can be used to
7880 print addresses in debug_print(). Adjusted all the address printing in the
7881 debugging in store.c to use %p rather than %d.
7883 81. There was a concern that a line of code in smtp_in.c could overflow a
7884 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7885 initially expressed, the concern was not well-founded, because trailing
7886 spaces are removed early. However, if the trailing spaces were followed by
7887 a NULL, they did not get removed, so the overflow was possible. Two fixes
7890 (a) I re-wrote the offending code in a cleaner fashion.
7891 (b) If an incoming SMTP command contains a NULL character, it is rejected
7894 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7895 runs initgroups(), so that if the Exim user is in any additional groups,
7896 they will be used during message reception.
7902 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob