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 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
60 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
61 side, to decode a TLS 1.3 packet capture.
63 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
64 Previously the default library behaviour applied, sending two, each in
67 JH/03 Debug output for ACL now gives the config file name and line number for
70 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
72 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
74 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
75 buffer overrun for (non-chunking) other transports.
77 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
78 TLS1.3, means that a server rejecting a client certificate is not visible
79 to the client until the first read of encrypted data (typically the
80 response to EHLO). Add detection for that case and treat it as a failed
81 TLS connection attempt, so that the normal retry-in-clear can work (if
84 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
85 and/or domain. Found and fixed by Jason Betts.
87 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
88 configuration). If a CNAME target was not a wellformed name pattern, a
91 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
92 the OS reports them interleaved with other addresses.
94 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
95 used both for input and for a verify callout, both encrypted, SMTP
96 responses being sent by the server could be lost. This resulted in
97 dropped connections and sometimes bounces generated by a peer sending
100 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
101 string. Previously it was possible to cause undefined behaviour in a
102 library routine (usually a crash). Found by "zerons".
104 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
107 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
108 API was removed, so update to use the newer ones.
110 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
111 any timeout set, is taking a long time. Previously we would hang on to a
112 rotated logfile "forever" if the input was arriving with long gaps
113 (a previous attempt to fix addressed lack, for a long time, of initial
116 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
117 shared (NFS) environment. The length of the tempfile name is now
118 4 + 16 ("hdr.$message_exim_id") which might break on file
119 systems which restrict the file name length to lower values.
122 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
123 shared (NFS) environment.
125 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
126 did for all versions <4.90). Notably -M, -m, --invert, -I may be
129 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
130 on some platforms for bit 31.
132 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
133 to changes apparently associated with TLS1.3 handling some of the APIs
134 previously used were either nonfunctional or inappropriate. Strings
135 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
136 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
137 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
138 This affects log line X= elements, the $tls_{in,out}_cipher variables,
139 and the use of specific cipher names in the encrypted= ACL condition.
141 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
143 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
144 verification result was not updated unless hosts_require_ocsp applied.
146 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
147 queue_list_requires_admin set to false, non-admin users were denied the
150 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
151 directory-of-certs mode. Previously they were advertised despite the
154 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
155 A single TCP connection by a client will now hold a TLS connection open
156 for multiple message deliveries, by default. Previoud the default was to
159 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
160 default. If built with the facility, DANE will be used. The facility
161 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
163 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
164 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
165 must be defined and you must still, unless you define DISABLE_TLS, manage
166 the the include-dir and library-file requirements that go with that
167 choice. Non-TLS builds are still supported.
169 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
172 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
173 default. If the platform supports and has the facility enabled, it will
174 be requested on all coneections.
176 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
177 controlled by the build-time option SUPPORT_PIPE_CONNECT.
179 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
181 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
182 success-DSN messages. Previously the From: header was always the default
183 one for these; the option was ignored.
185 JH/28 Fix the timeout on smtp response to apply to the whole response.
186 Previously it was reset for every read, so a teergrubing peer sending
187 single bytes within the time limit could extend the connection for a
188 long time. Credit to Qualsys Security Advisory Team for the discovery.
190 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
191 delivery address, which leaked information of the results of local
192 forwarding. Change to the original envelope recipient address, per
195 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
196 requested. Previously not bounce was generated and a log entry of
197 error ignored was made.
199 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
201 JH/32 Introduce a general tainting mechanism for values read from the input
202 channel, and values derived from them. Refuse to expand any tainted
203 values, to catch one form of exploit.
205 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
206 was unused and the unexpanded text used for the test. Found and
207 fixed by Ruben Jenster.
209 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
210 an attempt to use a TLS library read routine dereffed a nul pointer,
213 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
214 them in our smtp response.
216 JH/36 Have the general router option retry_use_local_part default to true when
217 any of the restrictive preconditions are set (to anything). Previously it
218 was only for check_local user. The change removes one item of manual
219 configuration which is required for proper retries when a remote router
220 handles a subset of addresses for a domain.
222 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
223 link count into consideration.
225 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
226 caused the extension of big_buffer, the following lines were ignored.
228 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
229 accordance with RFC 2308. Previously there was no expiry, so a longlived
230 receive process (eg. due to ACL delays) versus a short SOA value could
233 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
235 JH/39 Promote DMARC support to mainline.
237 JH/40 Bug 2452: Add a References: header to DSNs.
239 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
240 parameters. The relevant library call is documented as "Deprecated: This
241 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
242 3.6.0, DH parameters are negotiated following RFC7919."
244 HS/06 Change the default of dnssec_request_domains to "*"
246 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
247 carried on and emitted a BDAT command, even when PIPELINING was not
250 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
251 buffer was used for the filename, resulting in a trap when tainted
252 arguments (eg. $domain) were used.
254 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
255 recommended to avoid a possible server-load attack. The feature can be
256 re-enabled via the openssl_options main cofiguration option.
258 JH/45 local_scan API: documented the current smtp_printf() call. This changed
259 for version 4.90 - adding a "more data" boolean to the arguments.
260 Bumped the ABI version number also, this having been missed previously;
261 release versions 4.90 to 4.92.3 inclusive were effectively broken in
262 respect of usage of smtp_printf() by either local_scan code or libraries
263 accessed via the ${dlfunc } expansion item. Both will need coding
264 adjustment for any calls to smtp_printf() to match the new function
265 signature; a FALSE value for the new argument is always safe.
267 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
268 the file-offset (which the Linux syscall does, and exim expects); this
269 resulted in an indefinite loop.
271 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
272 to do ARC verification. The Authentication-Results: header line added
273 by the configuration then had no ARC item.
279 JH/01 Remove code calling the customisable local_scan function, unless a new
280 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
282 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
283 non-signal-safe functions being used.
285 JH/03 Bug 2269: When presented with a received message having a stupidly large
286 number of DKIM-Signature headers, disable DKIM verification to avoid
287 a resource-consumption attack. The limit is set at twenty.
289 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
290 report of oldest_pass in ${authres } in consequence, and separate out
291 some descriptions of reasons for verification fail.
293 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
294 files in the spool were present and unlocked. A queue-runner could spot
295 them, resulting in a duplicate delivery. Fix that by doing the unlock
296 after the unlink. Investigation by Tim Stewart. Take the opportunity to
297 add more error-checking on spoolfile handling while that code is being
300 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
301 No known attacks, no CVE, this is defensive hardening.
303 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
304 a queue-runner could start a delivery while other operations were ongoing.
305 Cutthrough delivery was a common victim, resulting in duplicate delivery.
306 Found and investigated by Tim Stewart. Fix by using the open message data
307 file handle rather than opening another, and not locally closing it (which
308 releases a lock) for that case, while creating the temporary .eml format
309 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
311 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
312 $sender_verify_failure/$recipient_verify_failure to "random".
314 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
317 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
318 Previously this would segfault.
320 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
323 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
324 like zero, since the resolver should be doing this for us, But we need one
325 as a CNAME but no MX presence gets the CNAME returned; we need to check
326 that doesn't point to an MX to declare it "no MX returned" rather than
327 "error, loop". A new main option is added so the older capability of
328 following some limited number of chain links is maintained.
330 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
332 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
333 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
334 now-deprecated earlier definitions used only specified the range up to TLS
335 1.2 (in the older-version library docs).
337 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
339 JH/15 Rework TLS client-side context management. Stop using a global, and
340 explicitly pass a context around. This enables future use of TLS for
341 connections to service-daemons (eg. malware scanning) while a client smtp
342 connection is using TLS; with cutthrough connections this is quite likely.
344 JH/16 Fix ARC verification to do AS checks in reverse order.
346 JH/17 Support a "tls" option on the ${readsocket } expansion item.
348 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
349 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
350 Previously the "utf8" would be re-prepended for every additional message.
352 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
353 Previously thery were accepted, resulting in issues when attempting to
354 forward messages to a non-supporting MTA.
356 PP/02 Let -n work with printing macros too, not just options.
358 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
359 one parent address was copied, and bogus data was used at delivery-logging
360 time. Either a crash (after delivery) or bogus log data could result.
361 Discovery and analysis by Tim Stewart.
363 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
364 Previously if the string ended mid-character, we did not insert the
365 promised '?' replacement.
367 PP/04 Documentation: current string operators work on bytes, not codepoints.
369 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
370 should pack well giving a smaller memory footprint so better caching and
371 therefore performance. Group the declarations where this can't be done so
372 that the byte-sized flag variables are not interspersed among pointer
373 variables, giving a better chance of good packing by the compiler.
375 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
376 non-null, to avoid issues with sites running BATV. Previously reports were
377 sent with an empty envelope sender so looked like bounces.
379 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
380 The ignore_error flag wasn't being returned from the filter subprocess so
381 was not set for later routers. Investigation and fix by Matthias Kurz.
383 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
384 and a msg:complete for the whole, when a message is manually removed using
385 -Mrm. Developement by Matthias Kurz, hacked on by JH.
387 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
388 a "Gnu special" function, asprintf() in the DB utility binary builds; I
389 hope that is portable enough.
391 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
392 requiring a known-CA anchor certificate; make it now rely entirely on the
393 TLSA as an anchor. Checking the name on the leaf cert against the name
394 on the A-record for the host is still done for TA (but not for EE mode).
396 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
397 included in delivery lines for non-proxied connections, when compiled with
398 SUPPORT_SOCKS and running with proxy logging enabled.
400 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
401 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
402 move the existing event to fire before the normal logging of message
403 failure so that custom logging is bracketed by normal logging.
405 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
406 msg:fail:internal event. Developement by Matthias Kurz.
408 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
409 far too small for todays use of crypto signatures stored there. Go all
410 the way to the max DNS message size of 64kB, even though this might be
411 overmuch for IOT constrained device use.
413 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
414 copy a string over itself. The library routine is documented as not
415 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
417 JH/32 For main options check_spool_space and check_inode_space, where the
418 platform supports 64b integers, support more than the previous 2^31 kB
419 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
420 the previous G, M, k.
422 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
423 $authenticated_fail_id variable on authentication failure. Previously
426 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
427 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
428 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
429 GNUTLS_SEC_PARAM_MEDIUM.
431 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
432 side. Previously we would continue as if no SNI had been received.
434 JH/36 Harden the handling of string-lists. When a list consisted of a sole
435 "<" character, which should be a list-separator specification, we walked
436 off past the nul-terimation.
438 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
439 causes) even when the retry time is not yet met. Previously they were
440 not, meaning that when (say) an account was over-quota and temp-rejecting,
441 and multiple senders' messages were queued, only one sender would get
442 notified on each configured delay_warning cycle.
444 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
446 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
447 cipher-suites, an error can be left on the stack even for a succeeding
448 accept; this results in impossible error messages when a later operation
451 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
452 return error codes indicating retry. Under TLS1.3 this becomes required.
454 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
455 it only wrote the new authenticators, resulting in a lack of tracking of
456 peer changes of ESMTP extensions until the next cache flush.
458 JH/41 Fix the loop reading a message header line to check for integer overflow,
459 and more-often against header_maxsize. Previously a crafted message could
460 induce a crash of the recive process; now the message is cleanly rejected.
462 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
463 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
469 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
470 When redis_servers is set to a list of > 1 element, and the Redis servers
471 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
472 case of MOVED into a DEFER case instead, thus moving the query onto the
473 next server in the list. For a cluster of N elements, all N servers must
474 be defined in redis_servers.
476 GF/02 Catch and remove uninitialized value warning in exiqsumm
477 Check for existence of @ARGV before looking at $ARGV[0]
479 JH/01 Replace the store_release() internal interface with store_newblock(),
480 which internalises the check required to safely use the old one, plus
481 the allocate and data copy operations duplicated in both (!) of the
482 extant use locations.
484 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
485 modifier. This matches the restriction on the commandline.
487 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
488 Previously only the last row was returned.
490 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
491 we assumed that tags in the header were well-formed, and parsed the
492 element content after inspecting only the first char of the tag.
493 Assumptions at that stage could crash the receive process on malformed
496 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
497 While running the DKIM ACL we operate on the Permanent memory pool so that
498 variables created with "set" persist to the DATA ACL. Also (at any time)
499 DNS lookups that fail create cache records using the Permanent pool. But
500 expansions release any allocations made on the current pool - so a dnsdb
501 lookup expansion done in the DKIM ACL releases the memory used for the
502 DNS negative-cache, and bad things result. Solution is to switch to the
503 Main pool for expansions.
504 While we're in that code, add checks on the DNS cache during store_reset,
505 active in the testsuite.
506 Problem spotted, and debugging aided, by Wolfgang Breyha.
508 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
509 When none of the hosts presented to a transport match an already-open
510 connection, close it and proceed with the list. Previously we would
511 queue the message. Spotted by Lena with Yahoo, probably involving
514 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
515 Previously a spurious "250 OK id=" response was appended to the proper
518 JH/08 The "support for" informational output now, which built with Content
519 Scanning support, has a line for the malware scanner interfaces compiled
520 in. Interface can be individually included or not at build time.
522 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
523 by the template makefile "src/EDITME". The "STREAM" support for an older
524 ClamAV interface method is removed.
526 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
527 rows affected is given instead).
529 JH/11 The runtime Berkeley DB library version is now additionally output by
530 "exim -d -bV". Previously only the compile-time version was shown.
532 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
533 SMTP connection. Previously, when one had more receipients than the
534 first, an abortive onward connection was made. Move to full support for
535 multiple onward connections in sequence, handling cutthrough connection
536 for all multi-message initiating connections.
538 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
539 routers. Previously, a multi-recipient message would fail to match the
540 onward-connection opened for the first recipient, and cause its closure.
542 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
543 a timeout on read on a GnuTLS initiating connection, resulting in the
544 initiating connection being dropped. This mattered most when the callout
545 was marked defer_ok. Fix to keep the two timeout-detection methods
548 JH/15 Relax results from ACL control request to enable cutthrough, in
549 unsupported situations, from error to silently (except under debug)
550 ignoring. This covers use with PRDR, frozen messages, queue-only and
553 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
555 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
556 metadata, resulting in a crash in free().
558 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
559 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
560 Broken also in d185889f4, with init system revamp.
562 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
563 Previously we abruptly closed the connection after reading a malware-
564 found indication; now we go on to read the "scan ok" response line,
567 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
568 ACL. Previously, a crash would result.
570 JH/19 Speed up macro lookups during configuration file read, by skipping non-
571 macro text after a replacement (previously it was only once per line) and
572 by skipping builtin macros when searching for an uppercase lead character.
574 JH/20 DANE support moved from Experimental to mainline. The Makefile control
575 for the build is renamed.
577 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
578 was allocated for every new TLS startup, meaning one per message. Fix
579 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
581 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
582 reported the original. Fix to report (as far as possible) the ACL
583 result replacing the original.
585 JH/23 Fix memory leak during multi-message connections using STARTTLS under
586 OpenSSL. Certificate information is loaded for every new TLS startup,
587 and the resources needed to be freed.
589 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
591 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
594 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
595 DATA response info to the (existing) per-recipient response info for
596 the "C=" log element. It can have useful tracking info from the
597 destination system. Patch from Simon Arlott.
599 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
600 length value. Previously this would segfault.
602 HS/02 Support Avast multiline protoocol, this allows passing flags to
603 newer versions of the scanner.
605 JH/28 Ensure that variables possibly set during message acceptance are marked
606 dead before release of memory in the daemon loop. This stops complaints
607 about them when the debug_store option is enabled. Discovered specifically
608 for sender_rate_period, but applies to a whole set of variables.
609 Do the same for the queue-runner and queue-list loops, for variables set
610 from spool message files. Do the same for the SMTP per-message loop, for
611 certain variables indirectly set in ACL operations.
613 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
614 as a multi-recipient message from a mailinglist manager). The coding had
615 an arbitrary cutoff number of characters while checking for more input;
616 enforced by writing a NUL into the buffer. This corrupted long / fast
617 input. The problem was exposed more widely when more pipelineing of SMTP
618 responses was introduced, and one Exim system was feeding another.
619 The symptom is log complaints of SMTP syntax error (NUL chars) on the
620 receiving system, and refused recipients seen by the sending system
621 (propating to people being dropped from mailing lists).
622 Discovered and pinpointed by David Carter.
624 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
625 replaced by the ${authresults } expansion.
627 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
629 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
630 allows proper process termination in container environments.
632 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
633 Previously the "final dot" had a newline after it; ensure it is CR,LF.
635 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
636 and "err_perm", deprecated since 4.83 when the RFC-defined words
637 "temperror" and "permerror" were introduced.
639 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
640 transport-filters or DKIM-signing. The restriction was lost in the
641 consolidation of verify-callout and delivery SMTP handling.
642 Extend the restriction to also cover ARC-signing.
644 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
645 in defer=pass mode supply a 450 to the initiator. Previously the message
648 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
649 tls_require_ciphers is used as before.
651 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
652 "pass_unscanned". Only tmpfails from the scanner are written to
653 the paniclog, as they may require admin intervention (permission
654 denied, license issues). Other scanner errors (like decompression
655 bombs) do not cause a paniclog entry.
657 JH/36 Fix reinitialisation of DKIM logging variable between messages.
658 Previously it was possible to log spurious information in receive log
661 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
662 triggered odd behaviour from Outlook Express clients.
664 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
667 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
668 since the IETF WG has not yet settled on that versus the original
669 "bare" representation.
671 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
672 Previously the millisecond value corrupted the output.
673 Fix also for syslog_pid=no and log_selector +pid, for which the pid
674 corrupted the output.
680 JH/01 Rework error string handling in TLS interface so that the caller in
681 more cases is responsible for logging. This permits library-sourced
682 string to be attached to addresses during delivery, and collapses
683 pairs of long lines into single ones.
685 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
686 during configuration. Wildcards are allowed and expanded.
688 JH/02 Rework error string handling in DKIM to pass more info back to callers.
689 This permits better logging.
691 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
692 do not close it down and have the child transport start it up again on
693 the passed-on TCP connection. Instead, proxy the child (and any
694 subsequent ones) for TLS via a unix-domain socket channel. Logging is
695 affected: the continued delivery log lines do not have any DNSSEC, TLS
696 Certificate or OCSP information. TLS cipher information is still logged.
698 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
699 identical IP addresses on different listening ports. Will also affect
702 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
703 add noisy ifdef guards to special-case this sillyness.
704 Patch from Bernd Kuhls.
706 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
707 than 255 are no longer allowed.
709 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
710 Disable the session-cache too, which might reduce our load. Since we
711 currrectly use a new context for every connection, both as server and
712 client, there is no benefit for these.
713 GnuTLS appears to not support tickets server-side by default (we don't
714 call gnutls_session_ticket_enable_server()) but client side is enabled
715 by default on recent versions (3.1.3 +) unless the PFS priority string
718 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
719 <https://reproducible-builds.org/specs/source-date-epoch/>.
721 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
722 the check for any unsuccessful recipients did not notice the limit, and
723 erroneously found still-pending ones.
725 JH/08 Pipeline CHUNKING command and data together, on kernels that support
726 MSG_MORE. Only in-clear (not on TLS connections).
728 JH/09 Avoid using a temporary file during transport using dkim. Unless a
729 transport-filter is involved we can buffer the headers in memory for
730 creating the signature, and read the spool data file once for the
731 signature and again for transmission.
733 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
734 4.77 as the kernel support then wasn't solid, having issues in 64bit
735 mode. Now, it's been long enough. Add support for FreeBSD also.
737 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
738 case where the routing stage had gathered several addresses to send to
739 a host before calling the transport for the first, we previously failed
740 to close down TLS in the old transport process before passing the TCP
741 connection to the new process. The new one sent a STARTTLS command
742 which naturally failed, giving a failed delivery and bloating the retry
743 database. Investigation and fix prototype from Wolfgang Breyha.
745 JH/12 Fix check on SMTP command input synchronisation. Previously there were
746 false-negatives in the check that the sender had not preempted a response
747 or prompt from Exim (running as a server), due to that code's lack of
748 awareness of the SMTP input buffering.
750 PP/04 Add commandline_checks_require_admin option.
751 Exim drops privileges sanely, various checks such as -be aren't a
752 security problem, as long as you trust local users with access to their
753 own account. When invoked by services which pass untrusted data to
754 Exim, this might be an issue. Set this option in main configuration
755 AND make fixes to the calling application, such as using `--` to stop
758 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
759 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
760 responses to those, into a single TLS record each way (this usually means
761 a single packet). As a side issue, smtp_enforce_sync now works on TLS
764 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
765 affects you only if you're dancing at the edge of the param size limits.
766 If you are, and this message makes sense to you, then: raise the
767 configured limit or use OpenSSL 1.1. Nothing we can do for older
770 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
771 cmdline element to get the documented default one. Previously it was
774 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
775 are unsuable for non-message-specific reasons - eg. connection timeout,
776 banner-time rejection.
778 JH/16 Fix logging of delivery remote port, when specified by router, under
781 PP/06 Repair manualroute's ability to take options in any order, even if one
782 is the name of a transport.
785 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
787 JH/17 Change the list-building routines interface to use the expanding-string
788 triplet model, for better allocation and copying behaviour.
790 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
791 Previously it was constructed the first time a possibly-matching string
792 was met in the configuration file input during startup; now it is done
795 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
796 compatible one, to avoid the (poorly documented) possibility of a config
797 file in the working directory redirecting the DB files, possibly correpting
798 some existing file. CVE-2017-10140 assigned for BDB.
800 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
801 cache-hot. Previously, although the result was properly cached, the
802 initial verify call returned a defer.
804 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
805 the main verify for receipient in uncached-mode.
807 JH/22 Retire historical build files to an "unsupported" subdir. These are
808 defined as "ones for which we have no current evidence of testing".
810 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
811 if present. Previously it was ignored.
813 JH/24 Start using specified-initialisers in C structure init coding. This is
814 a C99 feature (it's 2017, so now considered safe).
816 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
817 if was a fixed-sized field and bitmask ops via macros; it is now more
820 PP/07 GitHub PR 56: Apply MariaDB build fix.
821 Patch provided by Jaroslav Škarvada.
823 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
824 during Coverity cleanups [4.87 JH/47]
825 Diagnosis and fix provided by Michael Fischer v. Mollard.
827 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
828 the right size to place the terminating semicolon on its own folded
829 line, the header hash was calculated to an incorrect value thanks to
830 the (relaxed) space the fold became.
832 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
833 and confused the parent.
835 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
836 which could crash as a result. This could lead to undeliverable messages.
838 JH/28 Logging: "next input sent too soon" now shows where input was truncated
841 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
842 matters on fast-turnover and PID-randomising systems, which were getting
843 out-of-order delivery.
845 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
846 a possibly-overlapping copy. The symptom was that "Remote host closed
847 connection in response to HELO" was logged instead of the actual 4xx
850 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
851 Previously only that bufferd was discarded, resulting in SYMTP command
854 JH/32 DKIM: when a message has multiple signatures matching an identity given
855 in dkim_verify_signers, run the dkim acl once for each. Previously only
856 one run was done. Bug 2189.
858 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
859 "panic the current process" to "deliberately defer". The panic log is
860 still written with the problem list name; the mail and reject logs now
861 get a temp-reject line for the message that was being handled, saying
862 something like "domains check lookup or other defer". The SMTP 451
863 message is still "Temporary local problem".
865 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
866 A crafted sequence of BDAT commands could result in in-use memory beeing
867 freed. CVE-2017-16943.
869 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
870 from SMTP input. Previously it was always done; now only done for DATA
871 and not BDAT commands. CVE-2017-16944.
873 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
874 to the message (such as an overlong header line). Previously this was
875 not done and we did not exit BDAT mode. Followon from the previous item
876 though a different problem.
882 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
883 than -2003 did; needs libidn2 in addition to libidn.
885 JH/02 The path option on a pipe transport is now expanded before use.
887 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
888 Patch provided by "Björn", documentation fix added too.
890 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
891 missing a wire-to-host endian conversion.
893 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
894 close after a BDAT command line could be taken as a following command,
895 giving a synch failure. Fix by only checking for synch immediately
896 before acknowledging the chunk.
898 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
899 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
900 macro. Patches provided by Josh Soref.
902 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
903 Previously we did not; the RFC seems ambiguous and VRFY is not listed
904 by IANA as a service extension. However, John Klensin suggests that we
907 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
908 the dkim code may be unix-mode line endings rather than smtp wire-format
909 CRLF, so prepend a CR to any bare LF.
911 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
912 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
914 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
915 the first were themselves being wrongly included in the feed into dkim
916 processing; with most chunk sizes in use this resulted in an incorrect
917 body hash calculated value.
919 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
920 DKIM signature block, for verification. Although advised against by
921 standards it is specifically not ruled illegal.
923 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
925 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
926 missing a body hash (the bh= tag).
928 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
929 It seems that HAProxy sends the Proxy Protocol information in clear and
930 only then does a TLS startup, so do the same.
932 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
933 TCP connections (such as for Spamd) unless the daemon successfully set
934 Fast Open mode on its listening sockets. This fixes breakage seen on
935 too-old kernels or those not configured for Fast Open, at the cost of
936 requiring both directions being enabled for TFO, and TFO never being used
937 by non-daemon-related Exim processes.
939 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
940 endings, at least on the first header line. Try to canonify any that get
941 past that check, despite the cost.
943 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
944 now limited to an arbitrary five deep, while parsing addresses with the
945 strip_excess_angle_brackets option enabled.
947 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
948 instead leave the unprompted TLS handshake in socket buffer for the
949 TLS library to consume.
951 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
953 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
955 JH/16 Drop variables when they go out of scope. Memory management drops a whole
956 region in one operation, for speed, and this leaves assigned pointers
957 dangling. Add checks run only under the testsuite which checks all
958 variables at a store-reset and panics on a dangling pointer; add code
959 explicitly nulling out all the variables discovered. Fixes one known
960 bug: a transport crash, where a dangling pointer for $sending_ip_address
961 originally assigned in a verify callout, is re-used.
963 PP/06 Drop '.' from @INC in various Perl scripts.
965 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
967 PP/08 Reduce a number of compilation warnings under clang; building with
968 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
969 should be warning-free.
971 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
973 HS/01 Fix portability problems introduced by PP/08 for platforms where
974 realloc(NULL) is not equivalent to malloc() [SunOS et al].
976 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
977 chunk. This allows us to accept broken chunked messages. We need a more
978 general solution here.
980 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
981 already-broken messages in the queue.
983 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
985 JH/19 Fix reference counting bug in routing-generated-address tracking.
991 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
992 supports it and a size is available (ie. the sending peer gave us one).
994 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
995 years of being deprecated). The replacements are the ACLs
996 acl_smtp_mime and acl_not_smtp_mime.
998 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
999 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1000 or even an in-clear connection were permitted. Now, if the host lookup
1001 was dnssec and dane was requested then the host is only used if the
1002 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1003 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1004 if one fails this test.
1005 This means that a poorly-configured remote DNS will make it incommunicado;
1006 but it protects against a DNS-interception attack on it.
1008 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1009 non-noisy when a race steals the message being considered.
1011 JH/05 If main configuration option tls_certificate is unset, generate a
1012 self-signed certificate for inbound TLS connections.
1014 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1015 in rewrites and routers.
1017 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1018 and logged a warning sing 4.83; now they are a configuration file error.
1020 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1021 (lacking @domain). Apply the same qualification processing as RCPT.
1023 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1025 JH/10 Support ${sha256:} applied to a string (as well as the previous
1028 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1029 a cutthrough deliver is pending, as we always want to make a connection.
1030 This also avoids re-routing the message when later placing the cutthrough
1031 connection after a verify cache hit.
1032 Do not update it with the verify result either.
1034 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1035 when routing results in more than one destination address.
1037 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1038 signing (which inhibits the cutthrough capability). Previously only
1039 the presence of an option was tested; now an expansion evaluating as
1040 empty is permissible (obviously it should depend only on data available
1041 when the cutthrough connection is made).
1043 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1044 the relevant preceding SMTP command did not note the pipelining mode.
1046 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1047 Previously they were not counted.
1049 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1050 as one having no matching records. Previously we deferred the message
1051 that needed the lookup.
1053 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1054 distinguished as "(=".
1056 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1057 for missing MX records. Previously it only worked for missing A records.
1059 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1061 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1062 after the data-go-ahead and data-ack. Patch from Jason Betts.
1064 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1065 even for a "none" policy. Patch from Tony Meyer.
1067 JH/22 Fix continued use of a connection for further deliveries. If a port was
1068 specified by a router, it must also match for the delivery to be
1071 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1072 When one of the recipients of a message was unsuitable for the connection
1073 (has no matching addresses), we lost track of needing to mark it
1074 deferred. As a result mail would be lost.
1076 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1078 JH/25 Decoding ACL controls is now done using a binary search; the source code
1079 takes up less space and should be simpler to maintain. Merge the ACL
1080 condition decode tables also, with similar effect.
1082 JH/26 Fix problem with one_time used on a redirect router which returned the
1083 parent address unchanged. A retry would see the parent address marked as
1084 delivered, so not attempt the (identical) child. As a result mail would
1087 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1088 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1089 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1092 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1093 Main config options check_{log,spool}_{inodes,space} are now
1094 100 inodes, 10MB unless set otherwise in the configuration.
1096 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1097 Previously it only applied to the main-section connection policy
1100 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1102 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1103 by me. Added RFC7919 DH primes as an alternative.
1105 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1106 are not in the system include path.
1108 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1109 GnuTLS, when a session startup failed (eg because the client disconnected)
1110 Exim did stdio operations after fclose. This was exposed by a recent
1111 change which nulled out the file handle after the fclose.
1113 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1114 signed directly by the cert-signing cert, rather than an intermediate
1115 OCSP-signing cert. This is the model used by LetsEncrypt.
1117 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1119 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1120 an incoming connection.
1122 HS/02 Bug 1802: Do not half-close the connection after sending a request
1125 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1126 fallback to "prime256v1".
1128 JH/34 SECURITY: Use proper copy of DATA command in error message.
1129 Could leak key material. Remotely exploitable. CVE-2016-9963.
1135 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1136 and 3.4.4 - once the server is enabled to respond to an OCSP request
1137 it does even when not requested, resulting in a stapling non-aware
1138 client dropping the TLS connection.
1140 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1141 support variable-length bit vectors. No functional change.
1143 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1144 The I= interface field on outgoing lines is now after the H= remote
1145 host field, same as incoming lines. There is a separate
1146 outgoing_interface log selector which allows you to disable the
1149 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1150 If not running log_selector +smtp_connection the mainlog would be held
1151 open indefinitely after a "too many connections" event, including to a
1152 deleted file after a log rotate. Leave the per net connection logging
1153 leaving it open for efficiency as that will be quickly detected by the
1154 check on the next write.
1156 HS/01 Bug 1671: Fix post transport crash.
1157 Processing the wait-<transport> messages could crash the delivery
1158 process if the message IDs didn't exist for some reason. When
1159 using 'split_spool_directory=yes' the construction of the spool
1160 file name failed already, exposing the same netto behaviour.
1162 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1163 mime_regex ACL conditions.
1165 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1166 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1167 to HELO, local diagnostic string.
1169 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1170 log line to debug. Even when configured with a tls authenticator many
1171 client connections are expected to not authenticate in this way, so
1172 an authenticate fail is not an error.
1174 HS/02 Add the Exim version string to the process info. This way exiwhat
1175 gives some more detail about the running daemon.
1177 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1178 matter for fast-change records such as DNSBLs.
1180 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1181 retry record, even if constant. There may be multiple transports with
1182 different interface settings and the retry behaviour needs to be kept
1185 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1187 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1189 JH/10 Bug 840: fix log_defer_output option of pipe transport
1191 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1192 hosts_randomize. This matters a lot when combined with mua_wrapper.
1194 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1195 ${quote_pgsql:<string>} operator.
1197 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1199 JH/14 Change header file name for freeradius-client. Relevant if compiling
1200 with Radius support; from the Gentoo tree and checked under Fedora.
1202 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1204 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1205 option transport_filter, meaning no filtering.
1207 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1209 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1210 defaults to "*" (all hosts). The variable is now available when not built
1211 with TLS, default unset, mainly to enable keeping the testsuite sane.
1212 If a server certificate is not supplied (via tls_certificate) an error is
1213 logged, and clients will find TLS connections fail on startup. Presumably
1214 they will retry in-clear.
1215 Packagers of Exim are strongly encouraged to create a server certificate
1216 at installation time.
1218 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1219 with the $config_file variable.
1221 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1222 in transport context, after the attempt, and per-recipient. The latter type
1223 is per host attempted. The event data is the error message, and the errno
1224 information encodes the lookup type (A vs. MX) used for the (first) host,
1225 and the trailing two digits of the smtp 4xx response.
1227 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1228 to write to mainlog (or rejectlog, paniclog) in the window between file
1229 creation and permissions/ownership being changed. Particularly affects
1230 installations where exicyclog is run as root, rather than exim user;
1231 result is that the running daemon panics and dies.
1233 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1235 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1236 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1237 "pri" and "weight". Note that the previous implicit priority given by the
1238 list order is no longer honoured.
1240 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1241 for DKIM processing.
1243 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1244 by defining SUPPORT_SOCKS.
1246 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1247 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1248 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1249 variables are renamed to proxy_{local,external}_{address,port}.
1251 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1252 for a build by defining SUPPORT_I18N
1254 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1255 of the query string, and make ${quote_redis:} do that quoting.
1257 JH/29 Move Events support from Experimental to mainline, enabled by default
1258 and removable for a build by defining DISABLE_EVENT.
1260 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1262 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1263 cached by the daemon.
1265 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1266 by defining LOOKUP_REDIS. The libhiredis library is required.
1268 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1269 keys are given for lookup.
1271 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1272 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1273 only supported when built with TLS support. The PolarSSL SHA routines
1274 are still used when the TLS library is too old for convenient support.
1276 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1277 openssl_options), for security. OpenSSL forces this from version 1.1.0
1278 server-side so match that on older versions.
1280 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1281 allocation for $value could be released as the expansion processing
1282 concluded, but leaving the global pointer active for it.
1284 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1285 and to use the domains and local_parts ACL conditions.
1287 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1288 incorrectly not doubled on cutthrough transmission, hence seen as a
1289 body-termination at the receiving system - resulting in truncated mails.
1290 Commonly the sender saw a TCP-level error, and retransmitted the message
1291 via the normal store-and-forward channel. This could result in duplicates
1292 received - but deduplicating mailstores were liable to retain only the
1293 initial truncated version.
1295 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1297 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1299 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1300 we're in there, support oversigning also; bug 1309.
1302 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1304 HS/04 Add support for keep_environment and add_environment options.
1306 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1307 either intentional arithmetic overflow during PRNG, or testing config-
1310 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1311 delivery resulted in actual delivery. Cancel cutthrough before DATA
1314 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1315 rejects a recipient: pass the reject to the originator.
1317 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1318 Many were false-positives and ignorable, but it's worth fixing the
1321 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1322 for the new environment-manipulation done at startup. Move the routines
1323 from being local to tls.c to being global via the os.c file.
1325 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1326 an extract embedded as result-arg for a map, the first arg for extract
1327 is unavailable so we cannot tell if this is a numbered or keyed
1328 extraction. Accept either.
1334 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1337 JH/02 The smtp transport option "multi_domain" is now expanded.
1339 JH/03 The smtp transport now requests PRDR by default, if the server offers
1342 JH/04 Certificate name checking on server certificates, when exim is a client,
1343 is now done by default. The transport option tls_verify_cert_hostnames
1344 can be used to disable this per-host. The build option
1345 EXPERIMENTAL_CERTNAMES is withdrawn.
1347 JH/05 The value of the tls_verify_certificates smtp transport and main options
1348 default to the word "system" to access the system default CA bundle.
1349 For GnuTLS, only version 3.0.20 or later.
1351 JH/06 Verification of the server certificate for a TLS connection is now tried
1352 (but not required) by default. The verification status is now logged by
1353 default, for both outbound TLS and client-certificate supplying inbound
1356 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1359 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1360 Status Notification (bounce) messages are now MIME format per RFC 3464.
1361 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1362 under the control of the dsn_advertise_hosts option, and routers may
1363 have a dsn_lasthop option.
1365 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1366 default, modifiable by a malware= option. The list separator for
1367 the options can now be changed in the usual way. Bug 68.
1369 JH/10 The smtp_receive_timeout main option is now expanded before use.
1371 JH/11 The incoming_interface log option now also enables logging of the
1372 local interface on delivery outgoing connections.
1374 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1375 if the interface and destination host and port all match.
1377 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1380 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1381 Patch from Andrew Lewis.
1383 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1384 now supports optional time-restrictions, weighting, and priority
1385 modifiers per server. Patch originally by <rommer@active.by>.
1387 JH/16 The spamd_address main option now supports a mixed list of local
1388 and remote servers. Remote servers can be IPv6 addresses, and
1389 specify a port-range.
1391 JH/17 Bug 68: The spamd_address main option now supports an optional
1392 timeout value per server.
1394 JH/18 Bug 1581: Router and transport options headers_add/remove can
1395 now have the list separator specified.
1397 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1400 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1403 JH/21 Support for the A6 type of dns record is withdrawn.
1405 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1406 rather than the verbs used.
1408 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1409 from 255 to 1024 chars.
1411 JH/24 Verification callouts now attempt to use TLS by default.
1413 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1414 are generic router options now. The defaults didn't change.
1416 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1417 Original patch from Alexander Shikoff, worked over by JH.
1419 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1420 files if ZCAT_COMMAND is not executable.
1422 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1424 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1426 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1427 Normally benign, it bites when the pair was led to by a CNAME;
1428 modern usage is to not canonicalize the domain to a CNAME target
1429 (and we were inconsistent anyway for A-only vs AAAA+A).
1431 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1433 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1434 when evaluating $sender_host_dnssec.
1436 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1437 $sender_helo_dnssec variable.
1439 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1441 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1443 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1445 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1446 documented as working, but never had. Support all but $spam_report.
1448 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1449 added for tls authenticator.
1451 HS/03 Add perl_taintmode main config option
1457 TL/01 When running the test suite, the README says that variables such as
1458 no_msglog_check are global and can be placed anywhere in a specific
1459 test's script, however it was observed that placement needed to be near
1460 the beginning for it to behave that way. Changed the runtest perl
1461 script to read through the entire script once to detect and set these
1462 variables, reset to the beginning of the script, and then run through
1463 the script parsing/test process like normal.
1465 TL/02 The BSD's have an arc4random API. One of the functions to induce
1466 adding randomness was arc4random_stir(), but it has been removed in
1467 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1468 function when detected.
1470 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1471 cause callback expansion.
1473 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1474 syntax errors in an expansion can be treated as a string instead of
1475 logging or causing an error, due to the internal use of bool_lax
1476 instead of bool when processing it.
1478 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1479 server certificates when making smtp deliveries.
1481 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1483 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1485 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1487 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1488 Merged patch from Sebastian Wiedenroth.
1490 JH/05 Fix results-pipe from transport process. Several recipients, combined
1491 with certificate use, exposed issues where response data items split
1492 over buffer boundaries were not parsed properly. This eventually
1493 resulted in duplicates being sent. This issue only became common enough
1494 to notice due to the introduction of connection certificate information,
1495 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1497 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1498 size buffer was used, resulting in syntax errors when an expansion
1501 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1502 version 3.3.6 or later.
1504 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1505 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1506 both become "event_action", the variables become $event_name, $event_data
1507 and $event_defer_errno. There is a new variable $verify_mode, usable in
1508 routers, transports and related events. The tls:cert event is now also
1509 raised for inbound connections, if the main configuration event_action
1512 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1513 early OCSP support, but no stapling (appears to be less than 1.0.0).
1515 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1516 server certificate names available under the smtp transport option
1517 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1520 JH/10 Time-related extraction expansions from certificates now use the main
1521 option "timezone" setting for output formatting, and are consistent
1522 between OpenSSL and GnuTLS compilations. Bug 1541.
1524 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1525 encoded parameter in the incoming message. Bug 1558.
1527 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1528 include certificate info, eximon was claiming there were spoolfile
1531 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1533 JH/14 Log delivery-related information more consistently, using the sequence
1534 "H=<name> [<ip>]" wherever possible.
1536 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1537 are problematic for Debian distribution, omit them from the release
1540 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1542 JH/16 Fix string representation of time values on 64bit time_t architectures.
1545 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1546 output list separator was used.
1551 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1552 checkers that were complaining about end of non-void function with no
1555 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1556 This was a regression introduced in 4.83 by another bugfix.
1558 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1560 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1561 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1567 TF/01 Correctly close the server side of TLS when forking for delivery.
1569 When a message was received over SMTP with TLS, Exim failed to clear up
1570 the incoming connection properly after forking off the child process to
1571 deliver the message. In some situations the subsequent outgoing
1572 delivery connection happened to have the same fd number as the incoming
1573 connection previously had. Exim would try to use TLS and fail, logging
1574 a "Bad file descriptor" error.
1576 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1577 utilities have not been installed.
1579 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1580 temporary space as the ACL may create new global variables.
1582 TL/01 LDAP support uses per connection or global context settings, depending
1583 upon the detected version of the libraries at build time.
1585 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1586 to extract and use the src ip:port in logging and expansions as if it
1587 were a direct connection from the outside internet. PPv2 support was
1588 updated based on HAProxy spec change in May 2014.
1590 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1592 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1593 Properly escape header and check for NULL return.
1595 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1598 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1600 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1601 characters in header names, implemented as a verify condition.
1602 Contributed by Michael Fischer v. Mollard.
1604 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1605 results permerror and temperror. Previous values are deprecated but
1606 still accepted. In a future release, err_perm and err_temp will be
1607 completely removed, which will be a backward incompatibility if the
1608 ACL tests for either of these two old results. Patch contributed by
1609 user bes-internal on the mailing list.
1611 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1613 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1614 selectors, in both main and reject logs.
1616 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1619 JH/07 Add malware type "sock" for talking to simple daemon.
1621 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1623 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1624 routers/transports under cutthrough routing.
1626 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1627 numbers. Touch up "bool" conditional to keep the same definition.
1629 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1631 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1633 JH/12 Expand items in router/transport headers_add or headers_remove lists
1634 individually rather than the list as a whole. Bug 1452.
1636 Required for reasonable handling of multiple headers_ options when
1637 they may be empty; requires that headers_remove items with embedded
1638 colons must have them doubled (or the list-separator changed).
1640 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1641 view the policy declared in the DMARC record. Currently, $dmarc_status
1642 is a combined value of both the record presence and the result of the
1645 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1647 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1648 dnslookup router and the smtp transport (applying to the forward
1651 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1652 of ldap servers used for a specific lookup. Patch provided by Heiko
1655 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1656 New variable $lookup_dnssec_authenticated for observability.
1658 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1659 Patch submitted by Lars Timman.
1661 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1663 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1664 Requires trusted mode and valid format message id, aborts otherwise.
1665 Patch contributed by Heiko Schlichting.
1667 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1668 certextract with support for various fields. Bug 1358.
1670 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1671 is requested by default, modifiable by smtp transport option
1674 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1675 operate on certificate variables to give certificate fingerprints
1676 Also new ${sha256:cert_variable}.
1678 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1680 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1681 Christian Aistleitner.
1683 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1685 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1686 file. Patch from Wolfgang Breyha.
1688 JH/25 Expand the coverage of the delivery $host and $host_address to
1689 client authenticators run in verify callout. Bug 1476.
1691 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1692 align with daemon_smtp_ports. Bug 72.
1694 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1695 support and error reporting did not work properly.
1697 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1698 and is readable. Patch from Andrew Colin Kissa.
1700 TL/14 Enhance documentation of ${run expansion and how it parses the
1701 commandline after expansion, particularly in the case when an
1702 unquoted variable expansion results in an empty value.
1704 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1706 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1707 by renaming away from it. Addresses GH issue 15, reported by
1710 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1712 TL/15 SECURITY: prevent double expansion in math comparison functions
1713 (can expand unsanitized data). Not remotely exploitable.
1720 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1722 PP/02 Make -n do something, by making it not do something.
1723 When combined with -bP, the name of an option is not output.
1725 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1728 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1729 $sender_host_name and config options to manage this, and basic check
1732 PP/05 DSCP support for outbound connections and control modifier for inbound.
1734 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1735 (Only plugin which currently uses this is kerberos4, which nobody should
1736 be using, but we should make it available and other future plugins might
1737 conceivably use it, even though it would break NAT; stuff *should* be
1738 using channel bindings instead).
1740 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1741 name; added for Sendmail compatibility; requires admin caller.
1742 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1743 just ignore it); requires trusted caller.
1744 Also parse but ignore: -Ac -Am -X<logfile>
1747 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1749 TL/02 Add +smtp_confirmation as a default logging option.
1751 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1752 Patch by Magnus Holmgren from 2007-02-20.
1754 TL/04 Bugzilla 1281 - Spec typo.
1755 Bugzilla 1283 - Spec typo.
1756 Bugzilla 1290 - Spec grammar fixes.
1758 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1760 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1762 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1763 mailing list by by Dmitry Isaikin.
1765 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1767 JH/02 Support "G" suffix to numbers in ${if comparisons.
1769 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1771 NM/01 Bugzilla 1197 - Spec typo
1772 Bugzilla 1196 - Spec examples corrections
1774 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1776 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1777 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1780 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1781 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1783 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1784 "acl {{name}{arg}...}", and optional args on acl condition
1787 JH/05 Permit multiple router/transport headers_add/remove lines.
1789 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1791 JH/07 Avoid using a waiting database for a single-message-only transport.
1792 Performance patch from Paul Fisher. Bugzilla 1262.
1794 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1797 JH/09 Add $headers_added variable, with content from use of ACL modifier
1798 add_header (but not yet added to the message). Bugzilla 199.
1800 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1801 Pulled from Bugzilla 817 by Wolfgang Breyha.
1803 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1805 (nb: this is the same fix as in Exim 4.80.1)
1807 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1810 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1813 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1815 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1816 advertise SMTP AUTH mechanism to us, instead of a generic
1817 protocol violation error. Also, make Exim more robust to bad
1818 data from the Dovecot auth socket.
1820 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1822 When a queue runner is handling a message, Exim first routes the
1823 recipient addresses, during which it prunes them based on the retry
1824 hints database. After that it attempts to deliver the message to
1825 any remaining recipients. It then updates the hints database using
1828 So if a recipient address works intermittently, it can get repeatedly
1829 deferred at routing time. The retry hints record remains fresh so the
1830 address never reaches the final cutoff time.
1832 This is a fairly common occurrence when a user is bumping up against
1833 their storage quota. Exim had some logic in its local delivery code
1834 to deal with this. However it did not apply to per-recipient defers
1835 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1837 This change adds a proper retry rule check during routing so that the
1838 final cutoff time is checked against the message's age. We only do
1839 this check if there is an address retry record and there is not a
1840 domain retry record; this implies that previous attempts to handle
1841 the address had the retry_use_local_parts option turned on. We use
1842 this as an approximation for the destination being like a local
1843 delivery, as in LMTP.
1845 I suspect this new check makes the old local delivery cutoff check
1846 redundant, but I have not verified this so I left the code in place.
1848 TF/02 Correct gecos expansion when From: is a prefix of the username.
1850 Test 0254 submits a message to Exim with the header
1854 When I ran the test suite under the user fanf2, Exim expanded
1855 the header to contain my full name, whereas it should have added
1856 a Resent-Sender: header. It erroneously treats any prefix of the
1857 username as equal to the username.
1859 This change corrects that bug.
1861 GF/01 DCC debug and logging tidyup
1862 Error conditions log to paniclog rather than rejectlog.
1863 Debug lines prefixed by "DCC: " to remove any ambiguity.
1865 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1867 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1868 Bug spotted by Jeremy Harris; was flawed since initial commit.
1869 Would have resulted in OCSP responses post-SNI triggering an Exim
1870 NULL dereference and crash.
1872 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1874 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1875 Bug detection, analysis and fix by Samuel Thibault.
1876 Bugzilla 1331, Debian bug #698092.
1878 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1880 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1881 Server implementation by Todd Lyons, client by JH.
1882 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1883 config variable "prdr_enable" controls whether the server
1884 advertises the facility. If the client requests PRDR a new
1885 acl_data_smtp_prdr ACL is called once for each recipient, after
1886 the body content is received and before the acl_smtp_data ACL.
1887 The client is controlled by both of: a hosts_try_prdr option
1888 on the smtp transport, and the server advertisement.
1889 Default client logging of deliveries and rejections involving
1890 PRDR are flagged with the string "PRDR".
1892 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1893 fclose(). Diagnosis by Todd Lyons.
1895 PP/17 Update configure.default to handle IPv6 localhost better.
1896 Patch by Alain Williams (plus minor tweaks).
1899 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1900 This is now consistent with GnuTLS, and is now documented: the
1901 previous undocumented portable approach to treating the option as
1902 unset was to force an expansion failure. That still works, and
1903 an empty string is now equivalent.
1905 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1906 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1907 not performing validation itself.
1909 PP/20 Added force_command boolean option to pipe transport.
1910 Patch from Nick Koston, of cPanel Inc.
1912 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1915 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1917 PP/21 Fix eximon continuous updating with timestamped log-files.
1918 Broken in a format-string cleanup in 4.80, missed when I repaired the
1919 other false fix of the same issue.
1920 Report and fix from Heiko Schlichting.
1923 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1924 Report from Prashanth Katuri.
1926 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1927 It's SecureTransport, so affects any MacOS clients which use the
1928 system-integrated TLS libraries, including email clients.
1930 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1931 using a MIME ACL for non-SMTP local injection.
1932 Report and assistance in diagnosis by Warren Baker.
1934 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1936 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1938 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1939 last id that failed so it may be referenced in subsequent ACL's.
1941 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1944 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1945 ldap library initialization, allowing self-signed CA's to be
1946 used. Also properly sets require_cert option later in code by
1947 using NULL (global ldap config) instead of ldap handle (per
1948 session). Bug diagnosis and testing by alxgomz.
1950 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1951 the src/util/ subdirectory.
1953 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1954 renamed to Transport Post Delivery Action by Jeremy Harris, as
1957 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1958 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1959 redis_servers = needs to be configured which will be used by the redis
1960 lookup. Patch from Warren Baker, of The Packet Hub.
1962 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1964 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1965 hostname or reverse DNS when processing a host list. Used suggestions
1966 from multiple comments on this bug.
1968 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1970 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1971 Missed a few lines, added it to make the runtest require no keyboard
1974 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1975 contains upper case chars. Make router use caseful_local_part.
1977 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1978 support when GnuTLS has been built with p11-kit.
1984 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1986 This, or similar/improved, will also be change PP/11 of 4.82.
1992 PP/01 Handle short writes when writing local log-files.
1993 In practice, only affects FreeBSD (8 onwards).
1994 Bugzilla 1053, with thanks to Dmitry Isaikin.
1996 NM/01 Bugzilla 949 - Documentation tweak
1998 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2001 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2003 PP/02 Implemented gsasl authenticator.
2005 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2007 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2008 `pkg-config foo` for cflags/libs.
2010 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2011 with rest of GSASL and with heimdal_gssapi.
2013 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2014 `pkg-config foo` for cflags/libs for the TLS implementation.
2016 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2017 properties get this fed in as external SSF. A number of robustness
2018 and debugging improvements to the cyrus_sasl authenticator.
2020 PP/08 cyrus_sasl server now expands the server_realm option.
2022 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2023 Patch by Jeremy Harris.
2025 PP/10 Added dbmjz lookup type.
2027 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2029 PP/12 MAIL args handles TAB as well as SP, for better interop with
2030 non-compliant senders.
2031 Analysis and variant patch by Todd Lyons.
2033 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2034 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2035 Patch from Dirk Mueller <dmueller@suse.com>
2037 PP/13 tls_peerdn now print-escaped for spool files.
2038 Observed some $tls_peerdn in wild which contained \n, which resulted
2039 in spool file corruption.
2041 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2042 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2043 or write after TLS renegotiation, which otherwise led to messages
2046 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2047 as a tracking header (ie: a signed header comes before the signature).
2048 Patch from Wolfgang Breyha.
2050 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2051 comma-sep list; embedded commas doubled.
2053 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2055 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2057 Report and patch from Dmitry Banschikov.
2059 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2060 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2061 failures. We appear to now support TLS1.1+ with Exim.
2063 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2064 lets Exim select keys and certificates based upon TLS SNI from client.
2065 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2066 before an outbound SMTP session. New log_selector, +tls_sni.
2068 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2069 NULL dereference. Report and patch from Alun Jones.
2071 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2072 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2073 Not seeing resolver debug output on NetBSD, but suspect this is a
2074 resolver implementation change.
2076 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2077 Left warnings. Added "eximon gdb" invocation mode.
2079 PP/21 Defaulting "accept_8bitmime" to true, not false.
2081 PP/22 Added -bw for inetd wait mode support.
2083 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2084 locate the relevant includes and libraries. Made this the default.
2086 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2087 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2089 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2090 This may cause build issues on older platforms.
2092 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2093 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2094 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2095 Added SNI support via GnuTLS too.
2096 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2098 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2100 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2101 Applied second patch from Janne, implementing suggestion to default
2102 multiple-strings-in-record handling to match SPF spec.
2104 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2106 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2107 read-only, out of scope).
2108 Patch from Wolfgang Breyha, report from Stuart Northfield.
2110 PP/29 Fix three issues highlighted by clang analyser static analysis.
2111 Only crash-plausible issue would require the Cambridge-specific
2112 iplookup router and a misconfiguration.
2113 Report from Marcin Mirosław.
2115 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2117 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2118 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2119 As part of this, removing so much warning spew let me fix some minor
2120 real issues in debug logging.
2122 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2123 assignment on my part. Fixed.
2125 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2126 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2127 Janne Snabb (who went above and beyond: thank you).
2129 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2130 string otherwise requires a connection and a bunch more work and it's
2131 relatively easy to get wrong. Should also expose TLS library linkage
2134 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2135 64-bit ${eval} (JH/03).
2137 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2138 GNU libc to support some of the 64-bit stuff, should not lead to
2139 conflicts. Defined before os.h is pulled in, so if a given platform
2140 needs to override this, it can.
2142 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2143 protection layer was required, which is not implemented.
2144 Bugzilla 1254, patch from Wolfgang Breyha.
2146 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2147 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2148 tls_dhparam take prime identifiers. Also unbreak combination of
2149 OpenSSL+DH_params+TLSSNI.
2151 PP/39 Disable SSLv2 by default in OpenSSL support.
2157 PP/01 Solaris build fix for Oracle's LDAP libraries.
2158 Bugzilla 1109, patch from Stephen Usher.
2160 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2162 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2165 TF/02 Fix a couple more cases where we did not log the error message
2166 when unlink() failed. See also change 4.74-TF/03.
2168 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2169 lock up or crash if it happened to be inside a call to libc when it
2170 got a SIGUSR1 from exiwhat.
2172 The SIGUSR1 handler appends the current process status to the process
2173 log which is later printed by exiwhat. It used to use the general
2174 purpose logging code to do this, but several functions it calls are
2175 not safe for signals.
2177 The new output code in the SIGUSR1 handler is specific to the process
2178 log, and simple enough that it's easy to inspect for signal safety.
2179 Removing some special cases also simplifies the general logging code.
2180 Removing the spurious timestamps from the process log simplifies
2183 TF/04 Improved ratelimit ACL condition.
2185 The /noupdate option has been deprecated in favour of /readonly which
2186 has clearer semantics. The /leaky, /strict, and /readonly update modes
2187 are mutually exclusive. The update mode is no longer included in the
2188 database key; it just determines when the database is updated. (This
2189 means that when you upgrade Exim will forget old rate measurements.)
2191 Exim now checks that the per_* options are used with an update mode that
2192 makes sense for the current ACL. For example, when Exim is processing a
2193 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2194 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2195 must specify per_mail/readonly. If you omit the update mode it defaults to
2196 /leaky where that makes sense (as before) or /readonly where required.
2198 The /noupdate option is now undocumented but still supported for
2199 backwards compatibility. It is equivalent to /readonly except that in
2200 ACLs where /readonly is required you may specify /leaky/noupdate or
2201 /strict/noupdate which are treated the same as /readonly.
2203 A useful new feature is the /count= option. This is a generalization
2204 of the per_byte option, so that you can measure the throughput of other
2205 aggregate values. For example, the per_byte option is now equivalent
2206 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2208 The per_rcpt option has been generalized using the /count= mechanism
2209 (though it's more complicated than the per_byte equivalence). When it is
2210 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2211 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2212 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2213 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2214 Note that using per_rcpt with a non-readonly update mode in more than
2215 one ACL will cause the recipients to be double-counted. (The per_mail
2216 and per_byte options don't have this problem.)
2218 The handling of very low rates has changed slightly. If the computed rate
2219 is less than the event's count (usually one) then this event is the first
2220 after a long gap. In this case the rate is set to the same as this event's
2221 count, so that the first message of a spam run is counted properly.
2223 The major new feature is a mechanism for counting the rate of unique
2224 events. The new per_addr option counts the number of different
2225 recipients that someone has sent messages to in the last time period. It
2226 behaves like per_rcpt if all the recipient addresses are different, but
2227 duplicate recipient addresses do not increase the measured rate. Like
2228 the /count= option this is a general mechanism, so the per_addr option
2229 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2230 example, measure the rate that a client uses different sender addresses
2231 with the options per_mail/unique=$sender_address. There are further
2232 details in the main documentation.
2234 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2236 TF/06 Removed a few PCRE remnants.
2238 TF/07 Automatically extract Exim's version number from tags in the git
2239 repository when doing development or release builds.
2241 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2242 Bugzilla 879. Patch from Paul Fisher.
2244 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2245 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2248 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2250 PP/05 Variable $av_failed, true if the AV scanner deferred.
2251 Bugzilla 1078. Patch from John Horne.
2253 PP/06 Stop make process more reliably on build failure.
2254 Bugzilla 1087. Patch from Heiko Schlittermann.
2256 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2257 Bugzilla 1089. Patch from Heiko Schlittermann.
2259 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2260 Bugzilla 1131. Patch from Holger Weiß.
2262 PP/09 Handle IPv6 addresses with SPF.
2263 Bugzilla 860. Patch from Wolfgang Breyha.
2265 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2267 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2270 PP/11 match_* no longer expand right-hand-side by default.
2271 New compile-time build option, EXPAND_LISTMATCH_RHS.
2272 New expansion conditions, "inlist", "inlisti".
2274 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2276 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2278 PP/14 fix log_write() format string regression from TF/03.
2279 Bugzilla 1152. Patch from Dmitry Isaikin.
2285 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2287 PP/02 Harmonised TLS library version reporting; only show if debugging.
2288 Layout now matches that introduced for other libraries in 4.74 PP/03.
2290 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2292 PP/04 New "dns_use_edns0" global option.
2294 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2297 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2298 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2300 TK/01 Updated PolarSSL code to 0.14.2.
2301 Bugzilla 1097. Patch from Andreas Metzler.
2303 PP/07 Catch divide-by-zero in ${eval:...}.
2304 Fixes bugzilla 1102.
2306 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2309 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2310 format-string attack -- SECURITY: remote arbitrary code execution.
2312 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2313 time unintentionally subject to list matching rules, letting the header
2314 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2315 arbitrary string expansion). This allowed for information disclosure.
2317 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2318 INT_MIN/-1 -- value coerced to INT_MAX.
2324 NM/01 Workaround for PCRE version dependency in version reporting
2327 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2328 This fixes portability to compilers other than gcc, notably
2329 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2331 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2332 makefiles for portability to HP-UX and POSIX correctness.
2334 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2335 Also via indented variable definition in the Makefile.
2336 (Debugging by Oliver Heesakkers).
2338 PP/02 Restore caching of spamd results with expanded spamd_address.
2339 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2341 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2342 Improves build reliability. Fix from: Frank Elsner
2344 NM/02 Fix wide character breakage in the rfc2047 coding
2345 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2347 NM/03 Allow underscore in dnslist lookups
2348 Fixes bug 1026. Patch from Graeme Fowler
2350 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2351 Code patches from Adam Ciarcinski of NetBSD.
2353 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2356 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2357 is logged, to avoid truncation. Patch from John Horne.
2359 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2360 Patch from Jakob Hirsch.
2362 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2363 SQL string expansion failure details.
2364 Patch from Andrey Oktyabrski.
2366 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2367 Patch from Simon Arlott.
2369 PP/09 New lookups functionality failed to compile on old gcc which rejects
2370 extern declarations in function scope.
2371 Patch from Oliver Fleischmann
2373 PP/10 Use sig_atomic_t for flags set from signal handlers.
2374 Check getgroups() return and improve debugging.
2375 Fixed developed for diagnosis in bug 927 (which turned out to be
2378 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2379 Patch from Mark Zealey.
2381 PP/12 Bugzilla 1056: Improved spamd server selection.
2382 Patch from Mark Zealey.
2384 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2385 Based on patch from Heiko Schlittermann.
2387 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2388 Patch from Uwe Doering, sign-off by Michael Haardt.
2390 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2391 variable declaration deep within a block. Bug and patch from
2394 PP/15 lookups-Makefile IRIX compatibility coercion.
2396 PP/16 Make DISABLE_DKIM build knob functional.
2398 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2399 Patch by Simon Arlott
2401 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2402 variable argument macros. Our copy now differs from upstream.
2408 TF/01 Failure to get a lock on a hints database can have serious
2409 consequences so log it to the panic log.
2411 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2412 controlled using the smtp_confirmation log selector.
2414 TF/03 Include the error message when we fail to unlink a spool file.
2416 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2417 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2418 for maintaining out-of-tree patches for some time.
2420 PP/01 Bugzilla 139: Documentation and portability issues.
2421 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2422 Handle per-OS dynamic-module compilation flags.
2424 PP/02 Let /dev/null have normal permissions.
2425 The 4.73 fixes were a little too stringent and complained about the
2426 permissions on /dev/null. Exempt it from some checks.
2427 Reported by Andreas M. Kirchwitz.
2429 PP/03 Report version information for many libraries, including
2430 Exim version information for dynamically loaded libraries. Created
2431 version.h, now support a version extension string for distributors
2432 who patch heavily. Dynamic module ABI change.
2434 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2435 privilege escalation vulnerability whereby the Exim run-time user
2436 can cause root to append content of the attacker's choosing to
2439 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2442 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2443 If dropping privileges for untrusted macros, we disabled normal logging
2444 on the basis that it would fail; for the Exim run-time user, this is not
2445 the case, and it resulted in successful deliveries going unlogged.
2446 Fixed. Reported by Andreas Metzler.
2452 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2453 only prepend for the Resent-* case. Fixes regression introduced in
2454 Exim 4.70 by NM/22 for Bugzilla 607.
2456 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2457 increasing numbers of administrators be bitten by this.
2459 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2461 PP/03 Bugzilla 994: added openssl_options main configuration option.
2463 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2465 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2467 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2469 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2470 without a peer certificate, leading to a segfault because of an
2471 assumption that peers always have certificates. Be a little more
2472 paranoid. Problem reported by Martin Tscholak.
2474 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2475 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2476 NB: ClamAV planning to remove STREAM in "middle of 2010".
2477 CL also introduces -bmalware, various -d+acl logging additions and
2478 more caution in buffer sizes.
2480 PP/09 Implemented reverse_ip expansion operator.
2482 PP/10 Bugzilla 937: provide a "debug" ACL control.
2484 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2486 PP/12 Bugzilla 973: Implement --version.
2488 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2490 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2492 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2494 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2495 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2496 ignore trailing whitespace.
2498 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2500 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2503 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2504 Notification from Dr Andrew Aitchison.
2506 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2507 ExtendedDetectionInfo response format.
2508 Notification from John Horne.
2510 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2513 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2514 XSL and documented dependency on system catalogs, with examples of how
2517 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2520 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2521 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2522 configuration file which is writeable by the Exim user or group.
2524 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2525 of configuration files to cover files specified with the -C option if
2526 they are going to be used with root privileges, not just the default
2529 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2530 option (effectively making it always true).
2532 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2533 files to be used while preserving root privileges.
2535 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2536 that rogue child processes cannot use them.
2538 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2539 run-time user, instead of root.
2541 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2542 Exim run-time user without dropping privileges.
2544 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2545 result string, instead of calling string_vformat() twice with the same
2548 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2549 for other users. Others should always drop root privileges if they use
2550 -C on the command line, even for a whitelisted configure file.
2552 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2554 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2560 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2561 $data_path, and $header_path variables; fixed documentation bugs and
2564 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2565 exipick to access non-standard spools, including the "frozen" queue
2568 NM/01 Bugzilla 965: Support mysql stored procedures.
2569 Patch from Alain Williams
2571 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2573 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2574 Patch from Andreas Metzler
2576 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2577 Patch from Kirill Miazine
2579 NM/05 Bugzilla 671: Added umask to procmail example.
2581 JJ/03 installed exipick 20100323.0, fixing doc bug
2583 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2584 directory. Notification and patch from Dan Rosenberg.
2586 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2588 TK/02 Improve log output when DKIM signing operation fails.
2590 MH/01 Treat the transport option dkim_domain as a colon separated
2591 list, not as a single string, and sign the message with each element,
2592 omitting multiple occurences of the same signer.
2594 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2595 Bugzilla 985, 986. Patch by Simon Arlott
2597 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2598 Patch by Simon Arlott
2600 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2601 MBX locking. Notification from Dan Rosenberg.
2607 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2609 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2611 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2613 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2615 NM/04 Bugzilla 915: Flag broken perl installation during build.
2621 TK/01 Added patch by Johannes Berg that expands the main option
2622 "spamd_address" if it starts with a dollar sign.
2624 TK/02 Write list of recipients to X-Envelope-Sender header when building
2625 the mbox-format spool file for content scanning (suggested by Jakob
2628 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2629 (http://www.dcc-servers.net/) support via dccifd. Activated by
2630 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2632 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2633 by Mark Daniel Reidel <mr@df.eu>.
2635 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2636 When building exim an external PCRE library is now needed -
2637 PCRE is a system library on the majority of modern systems.
2638 See entry on PCRE_LIBS in EDITME file.
2640 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2641 conversation. Added nologin parameter to request.
2642 Patch contributed by Kirill Miazine.
2644 TF/01 Do not log submission mode rewrites if they do not change the address.
2646 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2648 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2649 log files in place. Contributed by Roberto Lima.
2651 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2653 TF/03 Bugzilla 615: When checking the local_parts router precondition
2654 after a local_part_suffix or local_part_prefix option, Exim now
2655 does not use the address's named list lookup cache, since this
2656 contains cached lookups for the whole local part.
2658 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2659 Robert Millan. Documentation is in experimental-spec.txt.
2661 TF/04 Bugzilla 668: Fix parallel build (make -j).
2663 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2665 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2666 Patch provided by Jan Srzednicki.
2668 TF/05 Leading white space used to be stripped from $spam_report which
2669 wrecked the formatting. Now it is preserved.
2671 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2672 that they are available at delivery time.
2674 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2676 TF/08 TLS error reporting now respects the incoming_interface and
2677 incoming_port log selectors.
2679 TF/09 Produce a more useful error message if an SMTP transport's hosts
2680 setting expands to an empty string.
2682 NM/06 Bugzilla 744: EXPN did not work under TLS.
2683 Patch provided by Phil Pennock.
2685 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2686 Patch provided by Richard Godbee.
2688 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2689 acl_smtp_notquit, added index entry.
2691 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2692 Patch provided by Eugene Bujak.
2694 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2695 accept(). Patch provided by Maxim Dounin.
2697 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2698 Patch provided by Phil Pennock.
2700 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2702 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2703 Patch provided by Brad "anomie" Jorsch.
2705 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2706 Patch provided by Dean Brooks.
2708 TK/05 Add native DKIM support (does not depend on external libraries).
2710 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2711 Patch provided by Graeme Fowler.
2713 NM/16 Bugzilla 851: Documentation example syntax fix.
2715 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2717 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2720 NM/19 Bugzilla 745: TLS version reporting.
2721 Patch provided by Phil Pennock.
2723 NM/20 Bugzilla 167: bool: condition support.
2724 Patch provided by Phil Pennock.
2726 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2727 clients. Patch provided by Phil Pennock.
2729 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2730 Patch provided by Brad "anomie" Jorsch.
2732 NM/23 Bugzilla 687: Fix misparses in eximstats.
2733 Patch provided by Heiko Schlittermann.
2735 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2736 Patch provided by Heiko Schlittermann.
2738 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2739 plus update to original patch.
2741 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2743 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2744 Patch provided by David Brownlee.
2746 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2748 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2750 NM/30 Bugzilla 888: TLS documentation bugfixes.
2752 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2754 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2755 Unlike the original bugzilla I have changed all shell scripts in src tree.
2757 NM/33 Bugzilla 898: Transport filter timeout fix.
2758 Patch by Todd Rinaldo.
2760 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2761 Patch by Serge Demonchaux.
2763 NM/35 Bugzilla 39: Base64 decode bug fixes.
2764 Patch by Jakob Hirsch.
2766 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2768 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2770 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2772 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2778 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2779 ALT-N's libdkim that I have put here:
2780 http://duncanthrax.net/exim-experimental/
2782 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2783 were called 'true' and it seems that C99 defines that as a reserved
2784 keyword to be used with 'bool' variable types. That means you could
2785 not include C99-style headers which use bools without triggering
2786 build errors in sieve.c.
2788 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2789 as mailq or other aliases. Changed the --help handling significantly
2790 to do whats expected. exim_usage() emits usage/help information.
2792 SC/01 Added the -bylocaldomain option to eximstats.
2794 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2796 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2798 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2804 PH/01 Another patch from the Sieve maintainer.
2806 PH/02 When an IPv6 address is converted to a string for single-key lookup
2807 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2808 dots are used instead of colons so that keys in lsearch files need not
2809 contain colons. This was done some time before quoting was made available
2810 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2811 (notated using the quote facility) so as to distinguish them from IPv4
2812 keys. This meant that lookups for IP addresses in host lists did not work
2813 for iplsearch lookups.
2815 This has been fixed by arranging for IPv6 addresses to be expressed with
2816 colons if the lookup type is iplsearch. This is not incompatible, because
2817 previously such lookups could never work.
2819 The situation is now rather anomalous, since one *can* have colons in
2820 ordinary lsearch keys. However, making the change in all cases is
2821 incompatible and would probably break a number of configurations.
2823 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2826 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2827 conversion specification without a maximum field width, thereby enabling
2828 a rogue spamd server to cause a buffer overflow. While nobody in their
2829 right mind would setup Exim to query an untrusted spamd server, an
2830 attacker that gains access to a server running spamd could potentially
2831 exploit this vulnerability to run arbitrary code as the Exim user.
2833 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2834 $primary_hostname instead of what libspf2 thinks the hosts name is.
2836 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2837 a directory entry by the name of the lookup key. Previously, if a
2838 symlink pointed to a non-existing file or a file in a directory that
2839 Exim lacked permissions to read, a lookup for a key matching that
2840 symlink would fail. Now it is enough that a matching directory entry
2841 exists, symlink or not. (Bugzilla 503.)
2843 PH/03 The body_linecount and body_zerocount variables are now exported in the
2846 PH/04 Added the $dnslist_matched variable.
2848 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2849 This means they are set thereafter only if the connection becomes
2852 PH/06 Added the client_condition to authenticators so that some can be skipped
2853 by clients under certain conditions.
2855 PH/07 The error message for a badly-placed control=no_multiline_responses left
2856 "_responses" off the end of the name.
2858 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2860 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2861 (without spaces) instead of just copying the configuration text.
2863 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2865 PH/11 Added $max_received_linelength.
2867 PH/12 Added +ignore_defer and +include_defer to host lists.
2869 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2870 way in which PCRE > 7.0 is built.
2872 PH/14 Implemented queue_only_load_latch.
2874 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2875 MAIL command. The effect was to mangle the value on 64-bit systems.
2877 PH/16 Another patch from the Sieve maintainer.
2879 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2881 PH/18 If a system quota error occurred while trying to create the file for
2882 a maildir delivery, the message "Mailbox is full" was not appended to the
2883 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2884 only to a quota excession during the actual writing of the file.
2886 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2887 characters?) which causes problems in log lines. The DN values are now
2888 passed through string_printing() before being added to log lines.
2890 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2891 and InterBase are left for another time.)
2893 PH/21 Added message_body_newlines option.
2895 PH/22 Guard against possible overflow in moan_check_errorcopy().
2897 PH/23 POSIX allows open() to be a macro; guard against that.
2899 PH/24 If the recipient of an error message contained an @ in the local part
2900 (suitably quoted, of course), incorrect values were put in $domain and
2901 $local_part during the evaluation of errors_copy.
2907 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2908 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2911 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2912 issue a MAIL command.
2914 PH/02 In an ACL statement such as
2916 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2918 if a client was not listed at all, or was listed with a value other than
2919 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2920 the condition was not true (as it should be), so access was not denied.
2921 The bug was that the ! inversion was incorrectly passed on to the second
2922 item. This has been fixed.
2924 PH/03 Added additional dnslists conditions == and =& which are different from
2925 = and & when the dns lookup returns more than one IP address.
2927 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2928 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2930 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2931 FSYNC, which compiles an option called disable_fsync that allows for
2932 bypassing fsync(). The documentation is heavily laced with warnings.
2934 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2936 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2937 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2938 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2939 including adding "make clean"; (3) Added -fPIC when compiling the test
2940 dynamically loaded module, to get rid of a warning.
2942 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2943 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2944 = 0s. The bug is otherwise harmless.
2946 PH/07 There was a bug in the dovecot authenticator such that the value of
2947 $auth1 could be overwritten, and so not correctly preserved, after a
2948 successful authentication. This usually meant that the value preserved by
2949 the server_setid option was incorrect.
2951 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2953 PH/09 Installed PCRE release 7.0.
2955 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2956 run for batched SMTP input. It is now run at the start of every message
2957 in the batch. While fixing this I discovered that the process information
2958 (output by running exiwhat) was not always getting set for -bs and -bS
2959 input. This is fixed, and it now also says "batched" for BSMTP.
2961 PH/11 Added control=no_pipelining.
2963 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2964 patch, slightly modified), and move the expansion of helo_data till after
2965 the connection is made in the smtp transport (so it can use these
2968 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2970 PH/14 Added log_selector = +pid.
2972 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2974 PH/16 Add ${if forany and ${if forall.
2976 PH/17 Added dsn_from option to vary the From: line in DSNs.
2978 PH/18 Flush SMTP output before performing a callout, unless control =
2979 no_callout_flush is set.
2981 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2982 was true (the default) a successful delivery failed to delete the retry
2983 item, thus causing premature timeout of the address. The bug is now
2986 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2988 PH/21 Long custom messages for fakedefer and fakereject are now split up
2989 into multiline responses in the same way that messages for "deny" and
2990 other ACL rejections are.
2992 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2993 with slight modification.
2995 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2996 draft, changing the syntax and factoring some duplicate code".
2998 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2999 for deliveries of the second and subsequent messages over the same SMTP
3002 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3003 ${reduce, with only minor "tidies".
3005 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3007 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3008 expansion side effects.
3010 PH/27 When a message times out after an over-quota error from an Exim-imposed
3011 quota, the bounce message says "mailbox is full". This message was not
3012 being given when it was a system quota that was exceeded. It now should
3015 MH/03 Made $recipients available in local_scan(). local_scan() already has
3016 better access to the recipient list through recipients_list[], but
3017 $recipients can be useful in postmaster-provided expansion strings.
3019 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3020 in the case of a MAIL command with additional options following the
3021 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3022 were accidentally chopped off.
3024 PH/29 SMTP synchronization checks are implemented when a command is read -
3025 there is a check that no more input is waiting when there shouldn't be
3026 any. However, for some commands, a delay in an ACL can mean that it is
3027 some time before the response is written. In this time, more input might
3028 arrive, invalidly. So now there are extra checks after an ACL has run for
3029 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3030 pipelining has not been advertised.
3032 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3034 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3035 returned in the SMTP response when smtp_return_error_details was false.
3036 This has been fixed.
3038 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3039 instead of the C I/O that was originally supplied, because problems were
3040 reported on Solaris.
3042 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3043 Exim which did not show up earlier: it was assuming that a call to
3044 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3045 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3046 was a macro that became an assignment, so it seemed to work. This has
3047 changed to a proper function call with a void return, hence the compile
3048 error. Exim's code has been fixed.
3050 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3053 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3055 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3057 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3058 because it turns out to be too big or there is a timeout) there is no
3059 "Completed" line in the log. When some messages of this type were
3060 selected by exigrep, they were listed as "not completed". Others were
3061 picked up by some special patterns. I have improved the selection
3062 criteria to be more general.
3064 PH/38 The host_find_failed option in the manualroute router can now be set
3065 to "ignore", to completely ignore a host whose IP address cannot be
3066 found. If all hosts are ignored, the behaviour is controlled by the new
3067 host_all_ignored option.
3069 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3070 homing or because of multiple MX records with /mx) generated more than
3071 one IP address, and the following item turned out to be the local host,
3072 all the secondary addresses of the first item were incorrectly removed
3073 from the list, along with the local host and any following hosts (which
3074 is what is supposed to happen).
3076 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3077 whoever called Exim into the -H file. In the case of the daemon it was
3078 behaving confusingly. When first started, it used values for whoever
3079 started the daemon, but after a SIGHUP it used the Exim user (because it
3080 calls itself on a restart). I have changed the code so that it now always
3083 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3084 message are rejected with the same error (e.g. no authentication or bad
3085 sender address), and a DATA command is nevertheless sent (as can happen
3086 with PIPELINING or a stupid MUA), the error message that was given to the
3087 RCPT commands is included in the rejection of the DATA command. This is
3088 intended to be helpful for MUAs that show only the final error to their
3091 PH/42 Another patch from the Sieve maintainer.
3093 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3094 Eximstats - Fixed some broken HTML links and added missing column headers
3096 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3097 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3099 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3101 PH/43 Yet another patch from the Sieve maintainer.
3103 PH/44 I found a way to check for a TCP/IP connection going away before sending
3104 the response to the final '.' that terminates a message, but only in the
3105 case where the client has not sent further data following the '.'
3106 (unfortunately, this is allowed). However, in many cases there won't be
3107 any further data because there won't be any more messages to send. A call
3108 to select() can be used: if it shows that the input is "ready", there is
3109 either input waiting, or the socket has been closed. An attempt to read
3110 the next input character can distinguish the two cases. Previously, Exim
3111 would have sent an OK response which the client would never have see.
3112 This could lead to message repetition. This fix should cure that, at
3113 least in a lot of common cases.
3115 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3116 advertised in response to EHLO.
3122 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3123 fixed by 4.65/MH/01 (is this a record?) are fixed:
3125 (i) An empty string was always treated as zero by the numeric comparison
3126 operators. This behaviour has been restored.
3128 (ii) It is documented that the numeric comparison operators always treat
3129 their arguments as decimal numbers. This was broken in that numbers
3130 starting with 0 were being interpreted as octal.
3132 While fixing these problems I realized that there was another issue that
3133 hadn't been noticed. Values of message_size_limit (both the global option
3134 and the transport option) were treated as octal if they started with 0.
3135 The documentation was vague. These values are now always treated as
3136 decimal, and I will make that clear in the documentation.
3142 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3143 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3146 MH/01 Don't check that the operands of numeric comparison operators are
3147 integers when their expansion is in "skipping" mode (fixes bug
3148 introduced by 4.64-PH/07).
3150 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3151 child addresses, Exim now panics and dies. Previously, because the count
3152 is held in a short int, deliveries were likely to be lost. As such a
3153 large number of recipients for a single message is ridiculous
3154 (performance will be very, very poor), I have chosen to impose a limit
3155 rather than extend the field.
3161 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3162 leftover -K file (the existence of which was triggered by #402).
3163 While we were at it, introduced process PID as part of the -K
3164 filename. This should rule out race conditions when creating
3167 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3168 processing considerably. Previous code took too long for large mails,
3169 triggering a timeout which in turn triggers #401.
3171 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3172 in the DK code in transports.c. sendfile() is not really portable,
3173 hence the _LINUX specificness.
3175 TF/01 In the add_headers option to the mail command in an Exim filter,
3176 there was a bug that Exim would claim a syntax error in any
3177 header after the first one which had an odd number of characters
3180 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3181 callout verification, Exim cached a "reject" for the entire domain. This
3182 is correct for most verifications, but it is not correct for a recipient
3183 verification with use_sender or use_postmaster set, because in that case
3184 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3185 case of MAIL FROM:<> rejection from other early rejections (e.g.
3186 rejection of HELO). When verifying a recipient using a non-null MAIL
3187 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3188 Whatever the result of the callout, the value of the domain cache is
3189 left unchanged (for any other kind of callout, getting as far as trying
3190 RCPT means that the domain itself is ok).
3192 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3195 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3196 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3197 session, the daemon ignores ECONNECT errors and logs others; it now
3198 ignores EPIPE as well.
3200 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3201 (quoted-printable decoding).
3203 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3204 later the small subsequent patch to fix an introduced bug.
3206 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3208 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3210 PH/08 An error is now given if message_size_limit is specified negative.
3212 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3213 to be given (somewhat) arbitrary names.
3215 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3218 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3219 miscellaneous code fixes
3221 PH/10 Added the log_reject_target ACL modifier to specify where to log
3224 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3225 hostname. This is wrong, because it relates to the incoming message (and
3226 probably the interface on which it is arriving) and not to the outgoing
3227 callout (which could be using a different interface). This has been
3228 changed to use the value of the helo_data option from the smtp transport
3229 instead - this is what is used when a message is actually being sent. If
3230 there is no remote transport (possible with a router that sets up host
3231 addresses), $smtp_active_hostname is used.
3233 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3234 tweaks were necessary in order to get it to work (see also 21 below):
3235 (a) The code assumed that strncpy() returns a negative number on buffer
3236 overflow, which isn't the case. Replaced with Exim's string_format()
3238 (b) There were several signed/unsigned issues. I just did the minimum
3239 hacking in of casts. There is scope for a larger refactoring.
3240 (c) The code used strcasecmp() which is not a standard C function.
3241 Replaced with Exim's strcmpic() function.
3242 (d) The code set only $1; it now sets $auth1 as well.
3243 (e) A simple test gave the error "authentication client didn't specify
3244 service in request". It would seem that Dovecot has changed its
3245 interface. Fortunately there's a specification; I followed it and
3246 changed what the client sends and it appears to be working now.
3248 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3251 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3252 address A is aliased to B and C, where B exists and C does not. Without
3253 -v the output is "A verified" because verification stops after a
3254 successful redirection if more than one address is generated. However,
3255 with -v the child addresses are also verified. Exim was outputting "A
3256 failed to verify" and then showing the successful verification for C,
3257 with its parentage. It now outputs "B failed to verify", showing B's
3258 parentage before showing the successful verification of C.
3260 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3261 look up a TXT record in a specific list after matching in a combined
3264 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3265 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3266 they consult the DNS. I had assumed they would set it the way they
3267 wanted; and indeed my experiments on Linux seem to show that in some
3268 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3269 To be on the safe side, however, I have now made the interface to
3270 host_find_byname() similar to host_find_bydns(), with an argument
3271 containing the DNS resolver options. The host_find_byname() function now
3272 sets these options at its start, just as host_find_bydns() does. The smtp
3273 transport options dns_qualify_single and dns_search_parents are passed to
3274 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3275 of host_find_byname() use the default settings of RES_DEFNAMES
3276 (qualify_single) but not RES_DNSRCH (search_parents).
3278 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3279 spool_read_header() do less string testing, by means of a preliminary
3280 switch on the second character of optional "-foo" lines. (This is
3281 overdue, caused by the large number of possibilities that now exist.
3282 Originally there were few.) While I was there, I also converted the
3283 str(n)cmp tests so they don't re-test the leading "-" and the first
3284 character, in the hope this might squeeze out yet more improvement.
3286 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3287 flag allowing group syntax was set by the header_syntax check but not
3288 turned off, possible causing trouble later; (2) The flag was not being
3289 set at all for the header_verify test, causing "group"-style headers to
3290 be rejected. I have now set it in this case, and also caused header_
3291 verify to ignore an empty address taken from a group. While doing this, I
3292 came across some other cases where the code for allowing group syntax
3293 while scanning a header line wasn't quite right (mostly, not resetting
3294 the flag correctly in the right place). These bugs could have caused
3295 trouble for malformed header lines. I hope it is now all correct.
3297 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3298 with the "reply" argument non-NULL. The code, however (which originally
3299 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3300 but it didn't always do it. This confused somebody who was copying the
3301 code for some other use. I have removed all the tests.
3303 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3304 feature that was used to support insecure browsers during the U.S. crypto
3305 embargo. It requires special client support, and Exim is probably the
3306 only MTA that supported it -- and would never use it because real RSA is
3307 always available. This code has been removed, because it had the bad
3308 effect of slowing Exim down by computing (never used) parameters for the
3309 RSA_EXPORT functionality.
3311 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3312 authenticator to fail if there's a tab character in the incoming data
3313 (there should never be unless someone is messing about, as it's supposed
3314 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3315 if the connection is using TLS or if the remote IP is the same as the
3316 local IP, and the "valid-client-cert option" if a client certificate has
3319 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3320 authenticators. This can be used for authorization after authentication
3321 succeeds. (In the case of plaintext, it servers for both authentication
3324 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3325 if any retry times were supplied.
3327 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3328 connection, where there is no client IP address to check. In this
3329 situation, the verify now always succeeds.
3331 PH/25 Applied John Jetmore's -Mset patch.
3333 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3335 PH/27 In a string expansion for a processed (not raw) header when multiple
3336 headers of the same name were present, leading whitespace was being
3337 removed from all of them, but trailing whitespace was being removed only
3338 from the last one. Now trailing whitespace is removed from each header
3339 before concatenation. Completely empty headers in a concatenation (as
3340 before) are ignored.
3342 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3343 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3345 PH/29 [Removed. This was a change that I later backed out, and forgot to
3346 correct the ChangeLog entry (that I had efficiently created) before
3347 committing the later change.]
3349 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3350 address errors (4xx response to RCPT) over the same connection as other
3351 messages routed to the same hosts. Such deliveries are always "forced",
3352 so retry times are not inspected. This resulted in far too many retries
3353 for the affected addresses. The effect occurred only when there were more
3354 hosts than the hosts_max_try setting in the smtp transport when it had
3355 the 4xx errors. Those hosts that it had tried were not added to the list
3356 of hosts for which the message was waiting, so if all were tried, there
3357 was no problem. Two fixes have been applied:
3359 (i) If there are any address or message errors in an SMTP delivery, none
3360 of the hosts (tried or untried) are now added to the list of hosts
3361 for which the message is waiting, so the message should not be a
3362 candidate for sending over the same connection that was used for a
3363 successful delivery of some other message. This seems entirely
3364 reasonable: after all the message is NOT "waiting for some host".
3365 This is so "obvious" that I'm not sure why it wasn't done
3366 previously. Hope I haven't missed anything, but it can't do any
3367 harm, as the worst effect is to miss an optimization.
3369 (ii) If, despite (i), such a delivery is accidentally attempted, the
3370 routing retry time is respected, so at least it doesn't keep
3371 hammering the server.
3373 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3374 in ${readsocket because some servers need this prod.
3376 PH/32 Added some extra debug output when updating a wait-xxx database.
3378 PH/33 The hint "could be header name not terminated by colon", which has been
3379 given for certain expansion errors for a long time, was not being given
3380 for the ${if def:h_colon_omitted{... case.
3382 PH/34 The spec says: "With one important exception, whenever a domain list is
3383 being scanned, $domain contains the subject domain." There was at least
3384 one case where this was not true.
3386 PH/35 The error "getsockname() failed: connection reset by peer" was being
3387 written to the panic log as well as the main log, but it isn't really
3388 panic-worthy as it just means the connection died rather early on. I have
3389 removed the panic log writing for the ECONNRESET error when getsockname()
3392 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3393 runs only) independently of the message's sender address. This meant
3394 that, if the 4xx error was in fact related to the sender, a different
3395 message to the same recipient with a different sender could confuse
3396 things. In particular, this can happen when sending to a greylisting
3397 server, but other circumstances could also provoke similar problems.
3398 I have changed the default so that the retry time for these errors is now
3399 based a combination of the sender and recipient addresses. This change
3400 can be overridden by setting address_retry_include_sender=false in the
3403 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3404 remote server are returned as part of bounce messages. This was not
3405 happening for LMTP over a pipe (the lmtp transport), but now it is the
3406 same for both kinds of LMTP.
3408 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3409 in header lines that were in fact CNAMEs. This is no longer the case.
3411 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3412 and queue runs started by the daemon processed all messages. This has
3413 been fixed so that -R and -S can now usefully be given with -q<time>.
3415 PH/40 Import PCRE release 6.7 (fixes some bugs).
3417 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3419 PH/42 Give an error if -q is specified more than once.
3421 PH/43 Renamed the variables $interface_address and $interface_port as
3422 $received_ip_address and $received_port, to make it clear that these
3423 values apply to message reception, and not to the outgoing interface when
3424 a message is delivered. (The old names remain recognized, of course.)
3426 PH/44 There was no timeout on the connect() call when using a Unix domain
3427 socket in the ${readsocket expansion. There now is.
3429 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3430 be meaningful with "accept".
3432 SC/01 Eximstats V1.43
3433 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3435 SC/02 Eximstats V1.44
3436 Use a glob alias rather than an array ref in the generated
3437 parser. This improves both readability and performance.
3439 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3440 Collect SpamAssassin and rejection statistics.
3441 Don't display local sender or destination tables unless
3442 there is data to show.
3443 Added average volumes into the top table text output.
3445 SC/04 Eximstats V1.46
3446 Collect data on the number of addresses (recipients)
3447 as well as the number of messages.
3449 SC/05 Eximstats V1.47
3450 Added 'Message too big' to the list of mail rejection
3451 reasons (thanks to Marco Gaiarin).
3453 SC/06 Eximstats V1.48
3454 Mainlog lines which have GMT offsets and are too short to
3455 have a flag are now skipped.
3457 SC/07 Eximstats V1.49 (Alain Williams)
3458 Added the -emptyok flag.
3460 SC/08 Eximstats V1.50
3461 Fixes for obtaining the IP address from reject messages.
3463 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3464 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3465 whitespace changes from 4.64-PH/27
3467 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3470 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3471 are found, allow negative numbers in numeric criteria)
3473 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3475 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3476 to match changes made in 4.64-PH/43
3478 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3480 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3481 "domain" is actually an IP address, and force a failure. This locks out
3482 those revolvers/nameservers that support "A-for-A" lookups, in
3483 contravention of the specifications.
3485 PH/48 When a host name was looked up from an IP address, and the subsequent
3486 forward lookup of the name timed out, the host name was left in
3487 $sender_host_name, contrary to the specification.
3489 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3490 restricted to single-key lookups, Exim was not diagnosing an error if
3491 * or *@ was used with a query-style lookup.
3493 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3495 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3496 long ago, but noone interested enough thought of it. Let's just say that
3497 the "1.1" means that there are some new functions that weren't there at
3498 some point in the past.
3500 PH/51 Error processing for expansion failure of helo_data from an smtp
3501 transport during callout processing was broken.
3503 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3504 tested/used via the -bh/-bhc/-bs options.
3506 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3507 bug, fixed in subsequent PCRE releases).
3509 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3510 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3512 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3518 SC/01 Use a glob alias rather than an array ref in eximstats generated
3519 parser. This improves both readability and performance.
3521 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3522 Don't display local sender or destination tables in eximstats unless
3523 there is data to show.
3524 Added average volumes into the eximstats top table text output.
3526 SC/03 Collect data on the number of addresses (recipients) as well
3527 as the number of messages in eximstats.
3529 TF/01 Correct an error in the documentation for the redirect router. Exim
3530 does (usually) call initgroups() when daemonizing.
3532 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3533 with consistent privilege compared to when running as a daemon.
3535 TF/03 Note in the spec that $authenticated_id is not set for local
3536 submissions from trusted users.
3538 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3539 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3541 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3542 by adding some example configuration directives to the default
3543 configuration file. A little bit of work is required to uncomment the
3544 directives and define how usernames and passwords are checked, but
3545 there is now a framework to start from.
3547 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3548 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3549 without this. I don't know how relevant this is to other LDAP libraries.
3551 PH/02 Add the verb name to the "unknown ACL verb" error.
3553 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3555 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3557 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3558 directory not expanded when it should be if an expanded home directory
3559 was set for the address (which is overridden by the transport).
3561 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3564 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3565 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3566 because it is too late at that time, and has no effect.
3568 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3569 security issue with \' (bugzilla #107). I could not use the
3570 PQescapeStringConn() function, because it needs a PGconn value as one of
3573 PH/08 When testing addresses using -bt, indicate those final addresses that
3574 are duplicates that would not cause an additional delivery. At least one
3575 person was confused, thinking that -bt output corresponded to deliveries.
3576 (Suppressing duplicates isn't a good idea as you lose the information
3577 about possibly different redirections that led to the duplicates.)
3579 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3580 systems where poll() doesn't work, in particular OS X.
3582 PH/10 Added more information to debugging output for retry time not reached.
3584 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3585 operations in malware.c.
3587 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3590 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3591 syslog_duplication set false, log lines that would normally be written
3592 both the the main log and to the reject log were not written to syslog at
3595 PH/14 In the default configuration, change the use of "message" in ACL warn
3596 statements to "add_header".
3598 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3599 not followed by a command (e.g. "seen endif").
3601 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3602 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3605 PH/17 Added extra conditions to the default value of delay_warning_condition
3609 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3610 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3611 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3614 The Auto-Submitted: and various List- headers are standardised, whereas I
3615 don't think Precedence: ever was.
3617 PH/18 Refactored debugging code in route_finduser() to show more information,
3618 in particular, the error code if getpwnam() issues one.
3620 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3621 This is apparently needed in addition to the PH/07 change above to avoid
3622 any possible encoding problems.
3624 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3625 but not after initializing Perl.
3627 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3628 output them only if debugging. By default they are written stderr,
3629 apparently, which is not desirable.
3631 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3634 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3637 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3639 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3640 authenticated or an ident call has been made. Suppress the default
3641 values for $authenticated_id and $authenticated_sender (but permit -oMai
3642 and -oMas) when testing with -bh.
3644 PH/24 Re-jigged the order of the tests in the default configuration so that the
3645 tests for valid domains and recipients precede the DNS black list and CSA
3646 tests, on the grounds that those ones are more expensive.
3648 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3649 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3650 This bug exists in every version of Exim that I still have, right back to
3653 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3654 However, an attempt to turn on case-sensitivity in a regex key by
3655 including (?-i) didn't work because the subject string was already
3656 lowercased, and the effects were non-intuitive. It turns out that a
3657 one-line patch can be used to allow (?-i) to work as expected.
3663 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3664 other effects) broke the use of negated acl sub-conditions.
3666 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3669 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3670 "Deny" causes Exim to reject the incoming connection with a 554 error.
3671 Unfortunately, if there is a major crisis, such as a disk failure,
3672 tcp-wrappers gives "deny", whereas what one would like would be some
3673 kind of temporary error. A kludge has been added to help with this.
3674 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3675 554 error is used if errno is still zero or contains ENOENT (which occurs
3676 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3679 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3681 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3682 errors. Otherwise a message that provokes a temporary error (when other
3683 messages do not) can cause a whole host to time out.
3685 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3686 addresses were routed directly to files or pipes from a redirect router.
3687 File deliveries just didn't batch; pipe deliveries might have suffered
3690 PH/06 A failure to get a lock for a hints database would erroneously always say
3691 "Failed to get write lock", even when it was really a read lock.
3693 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3694 of 0600. This has been changed to use the value of the lockfile_mode
3695 option (which defaults to 0600).
3697 PH/08 Applied small patch from the Sieve maintainer.
3699 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3700 folder from quota calculations, a direct delivery into this folder messed
3701 up the contents of the maildirsize file. This was because the regex was
3702 used only to exclude .Trash (or whatever) when the size of the mailbox
3703 was calculated. There was no check that a delivery was happening into an
3704 excluded directory. This bug has been fixed by ignoring all quota
3705 processing for deliveries into excluded directories.
3707 PH/10 Added the maildirfolder_create_regex option to appendfile.
3713 PH/01 The code for finding all the local interface addresses on a FreeBSD
3714 system running IPv6 was broken. This may well have applied to all BSD
3715 systems, as well as to others that have similar system calls. The broken
3716 code found IPv4 interfaces correctly, but gave incorrect values for the
3717 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3718 that it would not match correctly against @[] and not recognize the IPv6
3721 PH/02 The ipliteral router was not recognizing addresses of the form user@
3722 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3724 PH/03 Added disable_ipv6.
3726 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3727 decoded form, because it is most often used to construct To: headers
3728 lines in autoreplies, and the decoded form may well be syntactically
3729 invalid. However, $reply_address has leading white space removed, and all
3730 newlines turned into spaces so that the autoreply transport does not
3733 PH/05 If group was specified without a user on a router, and no group or user
3734 was specified on a transport, the group from the router was ignored.
3736 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3737 for visible compile-time settings that can be used to change these
3738 numbers, for those that want even more. Backwards compatibility with old
3739 spool files has been maintained. However, going back to a previous Exim
3740 release will lost any variables that are in spool files.
3742 PH/07 Two small changes when running in the test harness: increase delay when
3743 passing a TCP/IP connection to a new process, in case the original
3744 process has to generate a bounce, and remove special handling of
3745 127.0.0.2 (sic), which is no longer necessary.
3747 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3748 be the same on different OS.
3750 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3753 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3754 whether --show-vars was specified or not
3756 JJ/02 exipick: Added support for new ACL variable spool format introduced
3759 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3760 syntactically invalid From: or Reply-to: line, and a filter used this to
3761 generate an autoreply, and therefore failed to obtain an address for the
3762 autoreply, Exim could try to deliver to a non-existent relative file
3763 name, causing unrelated and misleading errors. What now happens is that
3764 it logs this as a hard delivery error, but does not attempt to create a
3767 PH/11 The exinext utility has a -C option for testing purposes, but although
3768 the given file was scanned by exinext itself; it wasn't being passed on
3769 when Exim was called.
3771 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3772 an end-of-file indication when reading a command response.
3774 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3775 compiled. In many other places in Exim, IPv6 addresses are always
3776 recognized, so I have changed this. It also means that IPv4 domain
3777 literals of the form [IPV4:n.n.n.n] are now always recognized.
3779 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3780 used if the router is not running as root, for example, when verifying at
3781 ACL time, or when using -bh. The debugging output from this situation was
3782 non-existent - all you got was a failure to exec. I have made two
3785 (a) Failures to set uid/gid, the current directory, or a process leader
3786 in a subprocess such as that created by queryprogram now generate
3787 suitable debugging output when -d is set.
3789 (b) The queryprogram router detects when it is not running as root,
3790 outputs suitable debugging information if -d is set, and then runs
3791 the subprocess without attempting to change uid/gid.
3793 PH/15 Minor change to Makefile for building test_host (undocumented testing
3796 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3797 additional section of a DNS packet that returns MX or SRV records.
3798 Instead, it always explicitly searches for A/AAAA records. This avoids
3799 major problems that occur when a DNS server includes only records of one
3800 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3801 fixed another bug: if SRV records were looked up and the corresponding
3802 address records were *not* found in the additional section, the port
3803 values from the SRV records were lost.
3805 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3806 using the correct key (the original address) when searching the retry
3807 rules in order to find which one to use for generating the retry hint.
3809 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3810 adding the default one. Similarly, if it contains a Reply-To: header, the
3811 errors_reply_to option, if set, is not used.
3813 PH/19 When calculating a retry time, Exim used to measure the "time since
3814 failure" by looking at the "first failed" field in the retry record. Now
3815 it does not use this if it is later than than the arrival time of the
3816 message. Instead it uses the arrival time. This makes for better
3817 behaviour in cases where some deliveries succeed, thus re-setting the
3818 "first failed" field. An example is a quota failure for a huge message
3819 when small messages continue to be delivered. Without this change, the
3820 "time since failure" will always be short, possible causing more frequent
3821 delivery attempts for the huge message than are intended.
3822 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3824 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3825 $1, $2, $3) because the numerical variables can be reset during some
3826 expansion items (e.g. "match"), thereby losing the authentication data.
3828 PH/21 Make -bV show the size of off_t variables so that the test suite can
3829 decide whether to run tests for quotas > 2G.
3831 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3832 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3833 filecount value is greater than 2G or if a quota value is greater than 2G
3834 on a system where the size of off_t is not greater than 4, a panic error
3837 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3838 never match. The debug and -bh output now contains an explicit error
3839 message indicating a malformed IPv4 address or mask.
3841 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3842 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3843 PH/23 above applies.
3845 PH/25 Do not write to syslog when running in the test harness. The only
3846 occasion when this arises is a failure to open the main or panic logs
3847 (for which there is an explicit test).
3849 PH/26 Added the /no_tell option to "control=freeze".
3851 PH/27 If a host name lookup failed very early in a connection, for example, if
3852 the IP address matched host_lookup and the reverse lookup yielded a name
3853 that did not have a forward lookup, an error message of the form "no IP
3854 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3855 could be logged. Now it outputs the IP address instead of "NULL".
3857 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3858 allows the sender and the authenticated sender to be set when
3859 submitting a message from within Exim. Since child_open_exim() is
3860 documented for local_scan(), the new function should be too.
3862 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3863 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3864 results in an empty string is now treated as unset.
3866 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3868 PH/31 Added sender_verify_fail logging option.
3870 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3871 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3872 tidied the source and removed it altogether.
3874 PH/33 When a queue run was abandoned because the load average was too high, a
3875 log line was always written; now it is written only if the queue_run log
3876 selector is set. In addition, the log line for abandonment now contains
3877 information about the queue run such as the pid. This is always present
3878 in "start" and "stop" lines but was omitted from the "abandon" line.
3880 PH/34 Omit spaces between a header name and the colon in the error message that
3881 is given when verify = headers_syntax fails (if there are lots of them,
3882 the message gets confusing).
3884 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3885 names, as there are now some PTR records that contain slashes. This check
3886 is only to protect against broken name servers that fall over on strange
3887 characters, so the fact that it applies to all lookups doesn't matter.
3889 PH/36 Now that the new test suite is complete, we can remove some of the
3890 special code in Exim that was needed for the old test suite. For example,
3891 sorting DNS records because real resolvers return them in an arbitrary
3892 order. The new test suite's fake resolver always returns records in the
3895 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3896 bounces) except when queue_only is set, to avoid logging races between
3897 the different processes.
3899 PH/38 Panic-die if .include specifies a non-absolute path.
3901 PH/39 A tweak to the "H" retry rule from its user.
3903 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3904 a label. They prevented compilation on older perls.
3906 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3907 a warning to be raised on newish perls.
3909 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3910 on queue. Changes to match documented behaviour of showing count of
3911 messages matching specified criteria.
3913 PH/40 Changed the default ident timeout from 30s to 5s.
3915 PH/41 Added support for the use of login_cap features, on those BSD systems
3916 that have them, for controlling the resources used by pipe deliveries.
3918 PH/42 The content-scanning code uses fopen() to create files in which to put
3919 message data. Previously it was not paying any attention to the mode of
3920 the files. Exim runs with umask(0) because the rest of the code creates
3921 files with open(), and sets the required mode explicitly. Thus, these
3922 files were ending up world-writeable. This was not a big issue, because,
3923 being within the spool directory, they were not world-accessible. I have
3924 created a function called modefopen, which takes an additional mode
3925 argument. It sets umask(777), creates the file, chmods it to the required
3926 mode, then resets the umask. All the relevant calls to fopen() in the
3927 content scanning code have been changed to use this function.
3929 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3930 to 24 hours. This avoids potential overflow problems when processing G
3931 and H retry rules. I suspect nobody ever tinkers with this value.
3933 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3935 PH/45 When the plaintext authenticator is running as a client, the server's
3936 challenges are checked to ensure they are valid base64 strings. By
3937 default, the authentication attempt is cancelled if an invalid string is
3938 received. Setting client_ignore_invalid_base64 true ignores these errors.
3939 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3940 they are received. Thus, the responses can be made to depend on the
3941 challenges. If an invalid string is ignored, an empty string is placed in
3944 PH/46 Messages that are created by the autoreply transport now contains a
3945 References: header, in accordance with RFCs 2822 and 3834.
3947 PH/47 Added authenticated_sender_force to the smtp transport.
3949 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3951 PH/49 Installed latest patch from the Sieve maintainer.
3953 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3954 was also set, the appendfile transport was unnecessarily scanning a
3955 directory of message files (e.g. for maildir delivery) to find the count
3956 of files (along with the size), even though it did not need this
3957 information. It now does the scan only if it needs to find either the
3958 size of the count of files.
3960 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3962 PH/52 Two bugs concerned with error handling when the smtp transport is
3965 (i) Exim was not creating retry information for temporary errors given
3966 for individual recipients after the DATA command when the smtp transport
3967 was used in LMTP mode. This meant that they could be retried too
3968 frequently, and not timed out correctly.
3970 (ii) Exim was setting the flag that allows error details to be returned
3971 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3972 recipients that were returned after the DATA command.
3974 PH/53 This is related to PH/52, but is more general: for any failing address,
3975 when detailed error information was permitted to be returned to the
3976 sender, but the error was temporary, then after the final timeout, only
3977 "retry timeout exceeded" was returned. Now it returns the full error as
3978 well as "retry timeout exceeded".
3980 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3981 do this, and (what is worse) MTAs that accept it.
3983 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3984 will now be deprecated.
3986 PH/56 New os.c-cygwin from the Cygwin maintainer.
3988 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3989 formats (previously only available in exim formats via -bpr, -bpru,
3990 and -bpra. Now also available in native and exiqgrep formats)
3992 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3993 with very large, slow to parse queues
3995 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3997 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3999 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4000 responses to authentication challenges, though it was showing the
4001 challenges; (ii) I've removed the CR characters from the debug output for
4004 PH/58 Allow for the insertion of a newline as well as a space when a string
4005 is turned into more than one encoded-word during RFC 2047 encoding. The
4006 Sieve code now uses this.
4008 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4009 data_4xx, lost_connection, tls_required.
4011 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4012 message was being sent as an SMTP response.
4014 PH/61 Add -l and -k options to exicyclog.
4016 PH/62 When verifying, if an address was redirected to one new address, so that
4017 verification continued, and the new address failed or deferred after
4018 having set something in $address_data, the value of $address_data was not
4019 passed back to the ACL. This was different to the case when no
4020 redirection occurred. The value is now passed back in both cases.
4022 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4023 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4024 use login_cap.h, so on its own it isn't the distinguishing feature. The
4025 new name refers directly to the setclassresources() function.
4027 PH/65 Added configuration files for NetBSD3.
4029 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4031 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4032 is preferred over IPv4.
4034 PH/68 The bounce_return_message and bounce_return_body options were not being
4035 honoured for bounces generated during the reception of non-SMTP messages.
4036 In particular, this applied to messages rejected by the ACL. This bug has
4037 been fixed. However, if bounce_return_message is true and bounce_return_
4038 body is false, the headers that are returned for a non-SMTP message
4039 include only those that have been read before the error was detected.
4040 (In the case of an ACL rejection, they have all been read.)
4042 PH/69 The HTML version of the specification is now built in a directory called
4043 spec_html instead of spec.html, because the latter looks like a path with
4044 a MIME-type, and this confuses some software.
4046 PH/70 Catch two compiler warnings in sieve.c.
4048 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4049 function verify_get_ident() calls ip_connect() to connect a socket, but
4050 if the "connect()" function timed out, ip_connect() used to close the
4051 socket. However, verify_get_ident() also closes the socket later, and in
4052 between Exim writes to the log, which may get opened at this point. When
4053 the socket was closed in ip_connect(), the log could get the same file
4054 descriptor number as the socket. This naturally causes chaos. The fix is
4055 not to close the socket in ip_connect(); the socket should be closed by
4056 the function that creates it. There was only one place in the code where
4057 this was missing, in the iplookup router, which I don't think anybody now
4058 uses, but I've fixed it anyway.
4060 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4061 well as to direct DNS lookups. Otherwise the handling of names in host
4062 lists is inconsistent and therefore confusing.
4068 PH/01 Two changes to the default runtime configuration:
4070 (1) Move the checks for relay_from_hosts and authenticated clients from
4071 after to before the (commented out) DNS black list checks.
4073 (2) Add control=submission to the relay_from_hosts and authenticated
4074 clients checks, on the grounds that messages accepted by these
4075 statements are most likely to be submissions.
4077 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4079 (1) Generate an error if the third argument for the ${prvs expansion is
4082 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4085 (3) Reset the variables that are obtained from the first argument of
4086 ${prvscheck and used in the second argument before leaving the code,
4087 because their memory is reclaimed, so using them afterwards may do
4090 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4091 one (it's much easier than Tom thought :-).
4093 (5) Because of (4), we can now allow for the use of $prvscheck_result
4094 inside the third argument.
4096 PH/03 For some reason, the default setting of PATH when running a command from
4097 a pipe transport was just "/usr/bin". I have changed it to
4100 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4101 anything to be listed in the output from -bV.
4103 PH/05 When a filter generated an autoreply, the entire To: header line was
4104 quoted in the delivery log line, like this:
4106 => >A.N.Other <ano@some.domain> <original@ddress> ...
4108 This has been changed so that it extracts the operative address. There
4109 may be more than one such address. If so, they are comma-separated, like
4112 => >ano@some.domain,ona@other.domain <original@ddress> ...
4114 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4115 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4116 not being looked up in the rDNS to get a host name, Exim was showing the
4117 IP address twice in Received: lines, even though the IP addresses were
4118 identical. For example:
4120 Received: from [1.2.3.4] (helo=[1.2.3.4])
4122 However, if the real host name was known, it was omitting the HELO data
4123 if it matched the actual IP address. This has been tidied up so that it
4124 doesn't show the same IP address twice.
4126 PH/07 When both +timestamp and +memory debugging was on, the value given by
4127 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4128 called by the debug printing, thereby overwriting the timestamp buffer.
4129 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4131 PH/08 When the original message was included in an autoreply transport, it
4132 always said "this is a copy of the message, including all the headers",
4133 even if body_only or headers_only was set. It now gives an appropriate
4136 PH/09 Applied a patch from the Sieve maintainer which:
4138 o fixes some comments
4139 o adds the (disabled) notify extension core
4140 o adds some debug output for the result of if/elsif tests
4141 o points to the current vacation draft in the documentation
4142 and documents the missing references header update
4146 o fixes a bug in processing the envelope test (when testing
4147 multiple envelope elements, the last element determined the
4150 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4151 Electronic Mail") by including:
4153 Auto-submitted: auto-generated
4155 in the messages that it generates (bounce messages and others, such as
4156 warnings). In the case of bounce messages for non-SMTP messages, there was
4157 also a typo: it was using "Auto_submitted" (underscore instead of
4158 hyphen). Since every message generated by Exim is necessarily in response
4159 to another message, thes have all been changed to:
4161 Auto-Submitted: auto-replied
4163 in accordance with these statements in the RFC:
4165 The auto-replied keyword:
4167 - SHOULD be used on messages sent in direct response to another
4168 message by an automatic process,
4170 - MUST NOT be used on manually-generated messages,
4172 - MAY be used on Delivery Status Notifications (DSNs) and Message
4173 Disposition Notifications (MDNs),
4175 - MUST NOT be used on messages generated by automatic or periodic
4176 processes, except for messages which are automatic responses to
4179 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4180 to the default Received: header definition.
4182 PH/12 Added log selector acl_warn_skipped (default on).
4184 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4185 variables because (a) they are in the wrong storage pool and (b) even if
4186 they were copied, it wouldn't work properly because of the caching.
4188 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4189 checking when decoding. Apparently there are clients that generate
4190 overlong encoded strings. Why am I not surprised?
4192 PH/15 If the first argument of "${if match_address" was not empty, but did not
4193 contain an "@" character, Exim crashed. Now it writes a panic log message
4194 and treats the condition as false.
4196 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4198 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4199 extension "envelope-auth". The code is finished and in agreement with
4200 other implementations, but there is no documentation so far and in fact,
4201 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4202 not changing the active code.
4204 Print executed "if" and "elsif" statements when debugging is used. This
4205 helps a great deal to understand what a filter does.
4207 Document more things not specified clearly in RFC3028. I had all this
4208 sorted out, when out of a sudden new issues came to my mind. Oops."
4210 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4213 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4214 canonical form (as documented). However, after a host name lookup from
4215 the IP address, check_host() was doing a simple string comparison with
4216 addresses acquired from the DNS when checking that the found name did
4217 have the original IP as one of its addresses. Since any found IPv6
4218 addresses are likely to be in abbreviated form, the comparison could
4219 fail. Luckily, there already exists a function for doing the comparison
4220 by converting both addresses to binary, so now that is used instead of
4221 the text comparison.
4223 PH/20 There was another similar case to PH/19, when a complete host name was
4224 given in a host list; looking up its IP address could give an abbreviated
4225 form, whereas the current host's name might or might not be abbreviated.
4226 The same fix has been applied.
4232 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4233 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4236 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4237 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4239 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4241 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4242 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4243 Also, for "Auto-Submitted", treat anything other than "no" as
4244 non-personal, in accordance with RFC 3834. (Previously it treated
4245 anything starting "auto-" as non-personal.)
4247 TF/01 The control=submission/name=... option had a problem with syntax
4248 errors if the name included a slash character. The /name= option
4249 now slurps the rest of the string, so it can include any characters
4250 but it must come last in the list of options (after /sender_retain
4253 PH/05 Some modifications to the interface to the fake nameserver for the new
4261 TK/01 Added the "success_on_redirect" address verification option. See
4262 NewStuff for rationale and an example.
4264 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4266 PH/02 Patch to exigrep to allow it to work on syslog lines.
4268 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4269 fread() to skip over the body file's header line, because in Cygwin the
4270 header line is locked and is inaccessible.
4272 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4273 co-exist for some time) to make it clear that it is the Exim ID that is
4274 referenced, not the Message-ID: header line.
4276 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4277 string_format() function, because snprintf() does not exist on all
4280 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4281 ${stat: expansion item.
4283 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4284 protocol synchronization error", to keep the pedants happy.
4286 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4287 well as for IRIX systems, when gcc is being used. See the host.c source
4290 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4292 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4295 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4296 transport and to the smtp transport in LMTP mode.
4298 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4300 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4301 run a filter in a subprocess. This could lead to confusion in subsequent
4302 lookups in the parent process. There should also be a search_tidyup() at
4303 the end of the subprocess.
4305 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4306 only if the host matched helo_try_verify_hosts, which caused the
4307 verification to occur when the EHLO/HELO command was issued. The ACL just
4308 tested the remembered result. Now, if a previous verification attempt has
4309 not happened, "verify = helo" does it there and then.
4311 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4313 TK/03 Fix log output including CR from clamd.
4315 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4316 exist provoked a memory error which could cause a segfault.
4318 PH/15 Installed PCRE 6.2
4320 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4322 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4323 of the problem. Specifically, suggested +O2 rather than +O1 for the
4326 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4328 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4329 record, the port was not correctly logged when the outgoing_port log
4330 selector was set (it logged the transort's default port).
4332 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4333 fallback_hosts, and "hosts" in the smtp transport.
4335 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4336 host errors such as "Connection refused".
4338 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4339 authentication with radiusclient 0.4.9:
4341 - Error returned from rc_read_config was caught wrongly
4342 - Username/password not passed on to radius server due to wrong length.
4344 The presumption is that some radiusclient API changes for 4.51/PH/17
4345 were not taken care of correctly. The code is still untested by me (my
4346 Linux distribution still has 0.3.2 of radiusclient), but it was
4347 contributed by a Radius user.
4349 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4350 expanding the "port" option of the smtp transport.
4352 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4353 while reading a MIME header. Thanks to Tom Hughes for a patch.
4355 PH/24 Include config.h inside local_scan.h so that configuration settings are
4358 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4359 that in an ACL for RCPT (for example), you can examine exactly what was
4362 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4363 commands, but it was not correctly comparing the address with the actual
4364 client host address. Thus, it would show the EHLO address in Received:
4365 header lines when this was not necessary.
4367 PH/27 Added the % operator to ${eval:}.
4369 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4370 it should be ignoring failures (because with -C, for example, it has lost
4371 privilege). It wasn't ignoring creation failures other than "already
4374 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4377 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4378 because an input line was too long, either on its own, or by virtue of
4379 too many continuations, the temporary file was not being removed, and the
4380 return code was incorrect.
4382 PH/31 Missing "BOOL" in function definition in filtertest.c.
4384 PH/32 Applied Sieve patches from the maintainer.
4386 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4388 PH/33 Added "verify = not_blind".
4390 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4391 Local/Makefile (with some defaults set). These are used in built scripts
4392 such as exicyclog, but they have never been used in the exim_install
4393 script (though there are many overriding facilities there). I have
4394 arranged that the exim_install script now takes note of these two
4397 PH/35 Installed configuration files for Dragonfly.
4399 PH/36 When a locally submitted message by a trusted user did not contain a
4400 From: header, and the sender address was obtained from -f or from an SMTP
4401 MAIL command, and the trusted user did not use -F to supply a sender
4402 name, $originator_name was incorrectly used when constructing a From:
4403 header. Furthermore, $originator_name was used for submission mode
4404 messages from external hosts without From: headers in a similar way,
4405 which is clearly wrong.
4407 PH/37 Added control=suppress_local_fixups.
4409 PH/38 When log_selector = +received_sender was set, and the addition of the
4410 sender made the log line's construction buffer exactly full, or one byte
4411 less than full, an overflow happened when the terminating "\n" was
4414 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4415 when the result of a list match is failure because a DNS lookup failed.
4417 PH/40 RM_COMMAND is now used in the building process.
4419 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4420 the "build-* directories that it finds.
4422 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4423 address in a domain literal was a prefix of an interface address.
4425 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4426 when verifying a sender address, unless rewrite_headers is false.
4428 PH/44 Wrote a long comment about why errors_to addresses are verified as
4429 recipients, not senders.
4431 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4432 the ratelimit ACL was added.
4434 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4436 PH/46 Added extra information about PostgreSQL errors to the error string.
4438 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4439 avoiding the need to install special zones in a real server. This is
4440 backwards compatible; if it can't find the fake resolver, it drops back.
4441 Thus, both old and new test suites can be run.
4443 TF/02 Added util/ratelimit.pl
4445 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4446 clock is set back in time.
4448 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4449 Candler <B.Candler@pobox.com>.
4451 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4452 OK for addresses that are the result of redirections.
4454 PH/48 A number of further additions for the benefit of the new test suite,
4455 including a fake gethostbyname() that interfaces to the fake DNS resolver
4458 TF/06 The fix for widen_domains has also been applied to qualify_single and
4459 search_parents which are the other dnslookup options that can cause
4462 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4465 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4467 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4468 service" if it encountered an MX record with an empty target hostname.
4469 The message is now "an MX or SRV record indicated no SMTP service".
4471 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4472 if the DNS of the sending site is misconfigured. This is quite a
4473 common situation. This change restores the behaviour of treating a
4474 helo verification defer as a failure.
4476 PH/51 If self=fail was set on a router, the bounce message did not include the
4477 actual error message.
4483 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4485 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4486 command itself ended in error, the underlying message about the transport
4487 filter timeout was being overwritten with the pipe command error. Now the
4488 underlying error message should be appended to the second error message.
4490 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4492 PH/02 Reduce the amount of output that "make" produces by default. Full output
4493 can still be requested.
4495 PH/03 The warning log line about a condition test deferring for a "warn" verb
4496 was being output only once per connection, rather than after each
4497 occurrence (because it was using the same function as for successful
4498 "warn" verbs). This seems wrong, so I have changed it.
4500 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4501 it should not have, which might have caused a crash in the right
4502 circumstances, but probably never did.
4504 PH/04 Installed a modified version of Tony Finch's patch to make submission
4505 mode fix the return path as well as the Sender: header line, and to
4506 add a /name= option so that you can make the user's friendly name appear
4509 TF/03 Added the control = fakedefer ACL modifier.
4511 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4512 Mark Lowes for thorough testing.
4514 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4516 TK/03 Merged latest SRS patch from Miles Wilton.
4518 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4519 with the definition in sysexits.h (which is #included earlier).
4520 Fortunately, Exim does not actually use EX_OK. The code used to try to
4521 preserve the sysexits.h value, by assuming that macro definitions were
4522 scanned for macro replacements. I have been disabused of this notion,
4523 so now the code just undefines EX_OK before #including unistd.h.
4525 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4526 in the smtp transport. When a block could not be written in a single
4527 write() function, the timeout was being re-applied to each part-write.
4528 This seems wrong - if the receiver was accepting one byte at a time it
4529 would take for ever. The timeout is now adjusted when this happens. It
4530 doesn't have to be particularly precise.
4532 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4533 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4535 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4536 without a domain if the check to <postmaster@domain> fails.
4538 SC/01 Eximstats: added -xls and the ability to specify output files
4539 (patch written by Frank Heydlauf).
4541 SC/02 Eximstats: use FileHandles for outputting results.
4543 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4545 SC/04 Eximstats: fixed display of large numbers with -nvr option
4547 SC/05 Eximstats: fixed merging of reports with empty tables.
4549 SC/06 Eximstats: added the -include_original_destination flag
4551 SC/07 Eximstats: removed tabs and trailing whitespace.
4553 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4555 TK/06 MBOX spool code: Add real "From " MBOX separator line
4556 so the .eml file is really in mbox format (even though
4557 most programs do not really care). Patch from Alex Miller.
4559 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4560 The latter is generated from $received_to and is only set if the
4561 message has one envelope recipient. SA can use these headers,
4562 obviously out-of-the-box. Patch from Alex Miller.
4564 PH/08 The ${def test on a variable was returning false if the variable's
4565 value was "0", contrary to what the specification has always said!
4566 The result should be true unless the variable is empty.
4568 PH/09 The syntax error of a character other than { following "${if
4569 def:variable_name" (after optional whitespace) was not being diagnosed.
4570 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4571 accidental colon was present, for example, could give incorrect results.
4573 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4574 result is used (not including appendfile, where other changes are about
4577 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4578 This involved changing a lot of size variables from int to off_t. It
4579 should work with maildirs and everything.
4581 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4582 spamd dying while we are connected to it.
4584 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4587 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4588 function for BDB 4.3.
4590 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4592 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4593 the sources for PCRE have been re-arranged and more files are now
4596 PH/15 The code I had for printing potentially long long variables in PH/11
4597 above was not the best (it lost precision). The length of off_t variables
4598 is now inspected at build time, and an appropriate printing format (%ld
4599 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4600 to be "long long int" or "long int". This is needed for the internal
4601 formatting function string_vformat().
4603 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4604 the configuration file to be ":syslog", then the script "guesses" where
4605 the logs files are, rather than using the compiled in default. In our
4606 case the guess is not the same as the compiled default, so the script
4607 suddenly stopped working when I started to use syslog. The patch checks
4608 to see if log_file_path is "". If so, it attempts to read it from exim
4609 with no configuration file to get the compiled in version, before it
4610 falls back to the previous guessing code."
4612 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4613 implementing BATV in an Exim configuration. See NewStuff for the gory
4616 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4617 Makefile that are specific to HP-UX.
4619 PH/18 If the "use_postmaster" option was set for a recipient callout together
4620 with the "random" option, the postmaster address was used as the MAIL
4621 FROM address for the random test, but not for the subsequent recipient
4622 test. It is now used for both.
4624 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4625 patch removes a few documentation additions to RFC 3028, because the
4626 latest draft now contains them. It adds the new en;ascii-case comparator
4627 and a new error check for 8bit text in MIME parts. Comparator and
4628 require names are now matched exactly. I enabled the subaddress
4629 extension, but it is not well tested yet (read: it works for me)."
4631 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4632 rework some of the code of TK/09 above to avoid the hardwired use of
4633 "%lld" and "long long". Replaced the call to snprintf() with a call to
4636 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4637 records point to non-existent hosts", "retry timeout exceeded", and
4638 "retry time not reached for any host after a long failure period".
4640 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4641 experimental DomainKeys support:
4643 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4644 (2) On an error such as an illegally used "control", the wrong name for
4645 the control was given.
4647 These problems did NOT occur unless DomainKeys support was compiled.
4649 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4651 PH/24 Added ${if match_ip condition.
4653 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4654 databases so that it will be absolutely obvious if a crash occurs in the
4655 DB library. This is a regular occurrence (often caused by mis-matched
4658 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4659 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4660 picked up on a user's system that detects such things. There doesn't seem
4661 to be a gcc warning option for this - only an attribute that has to be
4662 put on the function's prototype. It seems that in Fedora Core 4 they have
4663 set this on a number of new functions. No doubt there will be more in due
4666 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4667 specify a transport. However, if an address that was verified by such a
4668 router was the subject of a callout, Exim crashed because it tried to
4669 read the rcpt_include_affixes from the non-existent transport. Now it
4670 just assumes that the setting of that option is false. This bug was
4671 introduced by 4.51/PH/31.
4673 PH/28 Changed -d+all to exclude +memory, because that information is very
4674 rarely of interest, but it makes the output a lot bigger. People tend to
4675 do -d+all out of habit.
4677 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4678 code in os-type was giving problems when libc.so lives in lib64, like on
4681 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4682 aren't the modern standard, and it seems that some systems' include files
4683 don't always have them. Exim was already checking for some of the newer
4684 ones like T_AAAA, and defining it itself. I've added checks for all the
4685 record types that Exim uses.
4687 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4688 not automatically generating a new one, as it is supposed to. This
4689 prevented TLS from working. If the file did exist, but contained invalid
4690 data, a new version was generated, as expected. It was only the case of a
4691 non-existent file that was broken.
4693 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4694 with a change in libdomainkeys > 0.64.
4696 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4697 from DNS. If the selector record carries the flag, it now has
4698 precedence over the domain-wide flag.
4700 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4702 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4703 the use of a port name that isn't defined in /etc/services) occurred, the
4704 message was deferred as in a normal delivery, and thus remained on the
4705 spool, instead of being failed because of the mua_wrapper setting. This
4706 is now fixed, and I tidied up some of the mua_wrapper messages at the
4709 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4710 the messages in a hash of arrays rather than using individual hashes.
4711 This is a bit cleaner and results in dramatic memory savings, albeit
4712 at a slight CPU cost.
4714 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4715 as requested by Marc Sherman.
4717 SC/10 Eximstats: added histograms for user specified patterns as requested
4720 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4722 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4723 fopen() in the content-scanning modules that did not already have it.
4729 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4730 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4732 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4734 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4736 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4737 to test Sieve filters that use "vacation".
4739 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4740 that changes the way the GnuTLS parameters are stored in the cache file.
4741 The new format can be generated externally. For backward compatibility,
4742 if the data in the cache doesn't make sense, Exim assumes it has read an
4743 old-format file, and it generates new data and writes a new file. This
4744 means that you can't go back to an older release without removing the
4747 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4748 work if there are any delivery delays because "one_time" forces the
4749 parent to be marked "delivered", so its unseen clone is never tried
4750 again. For this reason, Exim now forbids the simultaneous setting of
4753 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4754 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4755 entry if you want to know the details. The fix, however, neglected to
4756 consider the case when local delivery batching is involved. The test for
4757 "previously delivered" was not happening when checking to see if an
4758 address could be batched with a previous (undelivered) one; under
4759 certain circumstances this could lead to multiple deliveries to the same
4762 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4763 in its include files, and this causes problems building Exim.
4765 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4766 header_syntax) but Exim was just ignoring anything given after a slash.
4767 In particular, this caused confusion with an attempt to use "verify =
4768 reverse_host_lookup/defer_ok". An error is now given when options are
4769 supplied for verify items that do not have them. (Maybe reverse_host_
4770 lookup should have a defer_ok option, but that's a different point.)
4772 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4773 defined by RFC 821) to 2048, because there were problems with some AUTH
4774 commands, and RFC 1869 says the size should be increased for extended
4775 SMTP commands that take arguments.
4777 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4780 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4781 "unknown" error; now it says that the functionality isn't in the binary.
4783 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4784 an address' error message when a string expansion fails (syntax or
4785 whatever). Otherwise the password may appear in the log. Following change
4786 PH/42 below, there is no longer a chance of it appearing in a bounce
4789 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4791 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4792 of its addresses was ever tried. (Bugzilla bug #2.)
4794 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4795 the result incorrectly in the debug output. (It correctly added a newline
4796 to what was transported.)
4798 TF/01 Added $received_time.
4800 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4801 commented out examples of how to interface to a virus scanner and to
4802 SpamAssassin. Also added commented examples of av_scanner and
4803 spamd_address settings.
4805 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4806 and controls are allowed in which ACLs. There were a couple of minor
4807 errors. Some of the entries in the conditions table (which is a table of
4808 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4809 negation of where the condition IS allowed.
4811 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4813 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4814 header file does not have a version number, so I've had to invent a new
4815 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4816 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4817 radiusclient), but it was contributed by a Radius user.
4819 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4820 files or directories, for OpenSSL.
4822 PH/19 When an Exim process that is running as root has to create an Exim log
4823 file, it does so in a subprocess that runs as exim:exim so as to get the
4824 ownership right at creation (otherwise, other Exim processes might see
4825 the file with the wrong ownership). There was no test for failure of this
4826 fork() call, which would lead to the process getting stuck as it waited
4827 for a non-existent subprocess. Forks do occasionally fail when resources
4828 run out. I reviewed all the other calls to fork(); they all seem to check
4831 PH/20 When checking for unexpected SMTP input at connect time (before writing
4832 the banner), Exim was not dealing correctly with a non-positive return
4833 from the read() function. If the client had disconnected by this time,
4834 the result was a log entry for a synchronization error with an empty
4835 string after "input=" when read() returned zero. If read() returned -1
4836 (an event I could not check), uninitialized data bytes were printed.
4837 There were reports of junk text (parts of files, etc) appearing after
4840 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4842 PH/22 Added support for macro redefinition, and (re)definition in between
4843 driver and ACL definitions.
4845 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4846 forgetting to use the resulting value; it was using the unexpanded value.
4848 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4849 hadn't been configured. The fix is from Juergen Kreileder, who
4850 understands it better than I do:
4852 "Here's what I see happening with three configured cyrus_sasl
4853 authenticators configured (plain, login, cram-md5):
4855 On startup auth_cyrus_sasl_init() gets called for each of these.
4856 This means three calls to sasl_listmech() without a specified mech_list.
4857 => SASL tests which mechs of all available mechs actually work
4858 => three warnings about OTP not working
4859 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4861 With the patch, sasl_listmech() also gets called three times. But now
4862 SASL's mech_list option is set to the server_mech specified in the the
4863 authenticator. Or in other words, the answer from sasl_listmech()
4864 gets limited to just the mech you're testing for (which is different
4866 => the return list contains just 'plain' or 'login', 'cram-md5' or
4867 nothing depending on the value of ob->server_mech.
4869 I've just tested the patch: Authentication still works fine,
4870 unavailable mechs specified in the exim configuration are still
4871 caught, and the auth.log warnings about OTP are gone."
4873 PH/25 When debugging is enabled, the contents of the command line are added
4874 to the debugging output, even when log_selector=+arguments is not
4877 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4878 answer is "GNU", and only if the return is "GNU/something" is the answer
4881 PH/27 $acl_verify_message is now set immediately after the failure of a
4882 verification in an ACL, and so is available in subsequent modifiers. In
4883 particular, the message can be preserved by coding like this:
4885 warn !verify = sender
4886 set acl_m0 = $acl_verify_message
4888 Previously, $acl_verify_message was set only while expanding "message"
4889 and "log_message" when a very denied access.
4891 PH/28 Modified OS/os.c-Linux with
4893 -#ifndef OS_LOAD_AVERAGE
4894 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4896 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4897 nomenclature these days.)
4899 PH/29 Installed patch from the Sieve maintainer that adds the options
4900 sieve_useraddress and sieve_subaddress to the redirect router.
4902 PH/30 In these circumstances:
4903 . Two addresses routed to the same list of hosts;
4904 . First host does not offer TLS;
4905 . First host accepts first address;
4906 . First host gives temporary error to second address;
4907 . Second host offers TLS and a TLS session is established;
4908 . Second host accepts second address.
4909 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4910 and peerdn, if requested) that were in fact used only for the second
4913 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4914 attention to any local part prefix or suffix that was matched by the
4915 router that accepted the address. It now behaves in the same way as it
4916 does for delivery: the affixes are removed from the local part unless
4917 rcpt_include_affixes is set on the transport.
4919 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4920 timeout during the DATA phase of an incoming message.
4922 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4923 applied a patch sanctioned by the Sieve maintainer.
4925 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4926 the uid or gid is negative. A case of a negative gid caused this to be
4927 noticed. The fix allows for either to be negative.
4929 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4930 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4933 PH/36 The change PH/12 above was broken. Fixed it.
4935 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4936 the grounds that routing the same address twice would always produce the
4937 same answer. This might have been true once, but it is certainly no
4938 longer true now. Routing a child address may depend on the previous
4939 routing that produced that child. Some complicated redirection strategies
4940 went wrong when messages had multiple recipients, and made Exim's
4941 behaviour dependent on the order in which the addresses were given.
4943 I have moved the duplicate checking until after the routing is complete.
4944 Exim scans the addresses that are assigned to local and remote
4945 transports, and removes any duplicates. This means that more work will be
4946 done, as duplicates will always all be routed, but duplicates are
4947 presumably rare, so I don't expect this is of any significance.
4949 For deliveries to pipes, files, and autoreplies, the duplicate checking
4950 still happens during the routing process, since they are not going to be
4953 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4954 It corrects a timeout issue with spamd. This is Ian's comment: "The
4955 background is that sometimes spamd either never reads data from a
4956 connection it has accepted, or it never writes response data. The exiscan
4957 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4958 blindly assumes that writes won't block so it may never time out."
4960 PH/39 Allow G after quota size as well as K and M.
4962 PH/40 The value set for $authenticated_id in an authenticator may not contain
4963 binary zeroes or newlines because the value is written to log lines and
4964 to spool files. There was no check on this. Now the value is run through
4965 the string_printing() function so that such characters are converted to
4966 printable escape sequences.
4968 PH/41 $message_linecount is a new variable that contains the total number of
4969 lines in the message. Compare $body_linecount, which is the count for the
4972 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4973 bounce and delay warning messages, except in certain special cases, which
4976 (a) An SMTP error message from a remote host;
4977 (b) A message specified in a :fail: redirection;
4978 (c) A message specified in a "fail" command in a system filter;
4979 (d) A message specified in a FAIL return from the queryprogram router;
4980 (e) A message specified by the cannot_route_message router option.
4982 In these cases only, Exim does include the error details in bounce and
4983 warning messages. There are also a few cases where bland messages such
4984 as "unrouteable address" or "local delivery error" are given.
4986 PH/43 $value is now also set for the "else" part of a ${run expansion.
4988 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4989 being worked on, but at least Exim now implements the latest version to
4992 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4993 process to complete was treated as a delivery failure, a timeout while
4994 writing the message to the pipe was logged, but erroneously treated as a
4995 successful delivery. Such timeouts include transport filter timeouts. For
4996 consistency with the overall process timeout, these timeouts are now
4997 treated as errors, giving rise to delivery failures by default. However,
4998 there is now a new Boolean option for the pipe transport called
4999 timeout_defer, which, if set TRUE, converts the failures into defers for
5000 both kinds of timeout. A transport filter timeout is now identified in
5003 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5004 systems where "make" and "gmake" are different, calling "gmake" at top
5005 level broke things. I've arranged for the value of $(MAKE) to be passed
5006 from the Makefile to this script so that it can call the same version of
5010 A note about Exim versions 4.44 and 4.50
5011 ----------------------------------------
5013 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5014 changes of various kinds. As a consequence, a big documentation update was
5015 needed. This delayed the release for rather longer than seemed good, especially
5016 in the light of a couple of (minor) security issues. Therefore, the changes
5017 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5018 release. So 4.44 and 4.50 are in effect two different branches that both start
5021 I have left the 4.50 change log unchanged; it contains all the changes since
5022 4.43. The change log for 4.44 is below; many of its items are identical to
5023 those for 4.50. This seems to be the most sensible way to preserve the
5024 historical information.
5030 1. Minor wording change to the doc/README.SIEVE file.
5032 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5033 computation of the current number of files was incorrect.
5035 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5036 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5039 4. Give more explanation in the error message when the command for a transport
5040 filter fails to execute.
5042 5. There are several places where Exim runs a non-Exim command in a
5043 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5044 was being done only for the command run by the queryprogram router. It is
5045 now done for all such subprocesses. The other cases are: ${run, transport
5046 filters, and the commands run by the lmtp and pipe transports.
5048 6. Added CONFIGURE_GROUP build-time option.
5050 7. Some older OS have a limit of 256 on the maximum number of file
5051 descriptors. Exim was using setrlimit() to set 1000 as a large value
5052 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5053 systems. I've change it so that if it can't get 1000, it tries for 256.
5055 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5056 was an oversight, and furthermore, ever since the addition of extra
5057 controls (e.g. 4.43/32), the checks on when to allow different forms of
5058 "control" were broken. There should now be diagnostics for all cases when a
5059 control that does not make sense is encountered.
5061 9. Added the /retain_sender option to "control=submission".
5063 10. $recipients is now available in the predata ACL (oversight).
5065 11. Tidy the search cache before the fork to do a delivery from a message
5066 received from the command line. Otherwise the child will trigger a lookup
5067 failure and thereby defer the delivery if it tries to use (for example) a
5068 cached ldap connection that the parent has called unbind on.
5070 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5071 of $address_data from the recipient verification was clobbered by the
5072 sender verification.
5074 13. The value of address_data from a sender verification is now available in
5075 $sender_address_data in subsequent conditions in the ACL statement.
5077 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5079 15. Added a new option "connect=<time>" to callout options, to set a different
5082 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5083 was its contents. (It was OK if the option was not defined at all.)
5085 17. A "Completed" log line is now written for messages that are removed from
5086 the spool by the -Mrm option.
5088 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5089 information about exactly what failed.
5091 19. Added -dd to debug only the daemon process.
5093 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5094 handles timeouts, both on the server side and network timeouts. Renamed the
5095 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5097 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5098 It is now set to "smtps".
5100 22. $host_address is now set to the target address during the checking of
5101 ignore_target_hosts.
5103 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5104 being passed; this would have caused $sender_host_name to have been used if
5105 matching the list had actually called for a host name (not very likely,
5106 since this list is usually IP addresses). A host name is now passed as
5109 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5110 code that specifies a non-restarting handler (typically sigaction() in
5111 modern systems) in an attempt to fix a rare and obscure crash bug.
5113 25. Narrowed the window for a race in the daemon that could cause it to ignore
5114 SIGCHLD signals. This is not a major problem, because they are used only to
5115 wake it up if nothing else does.
5117 26. A malformed maildirsize file could cause Exim to calculate negative values
5118 for the mailbox size or file count. Odd effects could occur as a result.
5119 The maildirsize information is now recalculated if the size or filecount
5122 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5123 support for a long time. Removed HAVE_SYS_VFS_H.
5125 28. Installed the latest version of exipick from John Jetmore.
5127 29. In an address list, if the pattern was not a regular expression, an empty
5128 subject address (from a bounce message) matched only if the pattern was an
5129 empty string. Non-empty patterns were not even tested. This was the wrong
5130 because it is perfectly reasonable to use an empty address as part of a
5131 database query. An empty address is now tested by patterns that are
5132 lookups. However, all the other forms of pattern expect the subject to
5133 contain a local part and a domain, and therefore, for them, an empty
5134 address still always fails if the pattern is not itself empty.
5136 30. Exim went into a mad DNS loop when attempting to do a callout where the
5137 host was specified on an smtp transport, and looking it up yielded more
5138 than one IP address.
5140 31. Re-factored the code for checking spool and log partition space into a
5141 function that finds that data and another that does the check. The former
5142 is then used to implement four new variables: $spool_space, $log_space,
5143 $spool_inodes, and $log_inodes.
5145 32. The RFC2047 encoding function was originally intended for short strings
5146 such as real names; it was not keeping to the 75-character limit for
5147 encoded words that the RFC imposes. It now respects the limit, and
5148 generates multiple encoded words if necessary. To be on the safe side, I
5149 have increased the buffer size for the ${rfc2047: expansion operator from
5152 33. It is now permitted to omit both strings after an "if" condition; if the
5153 condition is true, the result is "true". As before, when the second string
5154 is omitted, a false condition yields an empty string. This makes it less
5155 cumbersome to write custom ACL and router conditions.
5157 34. Failure to deliver a bounce message always caused it to be frozen, even if
5158 there was an errors_to setting on the router. The errors_to setting is now
5161 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5162 canonical form (fully expanded) before being placed in
5163 $sender_host_address.
5165 36. The table in the code that translates DNS record types into text (T_A to
5166 "A" for instance) was missing entries for NS and CNAME. It is just possible
5167 that this could have caused confusion if both these types were looked up
5168 for the same domain, because the text type is used as part of Exim's
5169 per-process caching. But the chance of anyone hitting this buglet seems
5172 37. The dnsdb lookup has been extended in a number of ways.
5174 (1) There is a new type, "zns", which walks up the domain tree until it
5175 finds some nameserver records. It should be used with care.
5177 (2) There is a new type, "mxh", which is like "mx" except that it returns
5178 just the host names, not the priorities.
5180 (3) It is now possible to give a list of domains (or IP addresses) to be
5181 looked up. The behaviour when one of the lookups defers can be
5182 controlled by a keyword.
5184 (4) It is now possible to specify the separator character for use when
5185 multiple records are returned.
5187 38. The dnslists ACL condition has been extended: it is now possible to supply
5188 a list of IP addresses and/or domains to be looked up in a particular DNS
5191 39. Added log_selector=+queue_time_overall.
5193 40. When running the queue in the test harness, wait just a tad after forking a
5194 delivery process, to get repeatability of debugging output.
5196 41. Include certificate and key file names in error message when GnuTLS fails
5197 to set them up, because the GnuTLS error message doesn't include the name
5198 of the failing file when there is a problem reading it.
5200 42. Allow both -bf and -bF in the same test run.
5202 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5204 44. The "Exiscan patch" is now merged into the mainline Exim source.
5206 45. Sometimes the final signoff response after QUIT could fail to get
5207 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5208 < 0 before doing a fflush(). This bug looks as though it goes back to the
5209 introduction of TLS in release 3.20, but "sometimes" must have been rare
5210 because the tests only now provoked it.
5212 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5213 (this can affect the format of dates).
5215 47. exim_tidydb, when checking for the continued existence of a message for
5216 which it has found a message-specific retry record, was not finding
5217 messages that were in split spool directories. Consequently, it was
5218 deleting retry records that should have stayed in existence.
5220 48. Steve fixed some bugs in eximstats.
5222 49. The SPA authentication driver was not abandoning authentication and moving
5223 on to the next authenticator when an expansion was forced to fail,
5224 contradicting the general specification for all authenticators. Instead it
5225 was generating a temporary error. It now behaves as specified.
5227 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5228 (the order specifies the preference for clients). The order is now AES256,
5229 AES128, 3DES, ARCFOUR128.
5231 51. Small patch to Sieve code - explicitly set From: when generating an
5234 52. Exim crashed if a remote delivery caused a very long error message to be
5235 recorded - for instance if somebody sent an entire SpamAssassin report back
5236 as a large number of 550 error lines. This bug was coincidentally fixed by
5237 increasing the size of one of Exim's internal buffers (big_buffer) that
5238 happened as part of the Exiscan merge. However, to be on the safe side, I
5239 have made the code more robust (and fixed the comments that describe what
5242 53. Now that there can be additional text after "Completed" in log lines (if
5243 the queue_time_overall log selector is set), a one-byte patch to exigrep
5244 was needed to allow it to recognize "Completed" as not the last thing in
5247 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5248 patch that reportedly fixes this has been added. I am not expert enough to
5249 create a test for it. This is what the patch creator wrote:
5251 "I found a little strange behaviour of ldap code when working with
5252 Windows 2003 AD Domain, where users was placed in more than one
5253 Organization Units. When I tried to give exim partial DN, the exit code
5254 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5255 But simultaneously result of request was absolutely normal ldap result,
5256 so I produce this patch..."
5258 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5259 REFERENCE, so I have modified the code to exclude the patch when that macro
5262 55. Some experimental protocols are using DNS PTR records for new purposes. The
5263 keys for these records are domain names, not reversed IP addresses. The
5264 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5265 leaves it alone. Component reversal etc. now happens only for IP addresses.
5268 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5270 57. Double the size of the debug message buffer (to 2048) so that more of very
5271 long debug lines gets shown.
5273 58. The exicyclog utility now does better if the number of log files to keep
5274 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5276 59. Two changes related to the smtp_active_hostname option:
5278 (1) $smtp_active_hostname is now available as a variable.
5279 (2) The default for smtp_banner uses $smtp_active_hostname instead
5280 of $primary_hostname.
5282 60. The host_aton() function is supposed to be passed a string that is known
5283 to be a valid IP address. However, in the case of IPv6 addresses, it was
5284 not checking this. This is a hostage to fortune. Exim now panics and dies
5285 if the condition is not met. A case was found where this could be provoked
5286 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5287 components; fortuitously, this particular loophole had already been fixed
5288 by change 4.50/55 above.
5290 If there are any other similar loopholes, the new check in host_aton()
5291 itself should stop them being exploited. The report I received stated that
5292 data on the command line could provoke the exploit when Exim was running as
5293 exim, but did not say which command line option was involved. All I could
5294 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5295 running as the user.
5298 61. There was a buffer overflow vulnerability in the SPA authentication code
5299 (which came originally from the Samba project). I have added a test to the
5300 spa_base64_to_bits() function which I hope fixes it.
5303 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5304 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5306 63. The daemon start-up calls getloadavg() while still root for those OS that
5307 need the first call to be done as root, but it missed one case: when
5308 deliver_queue_load_max is set with deliver_drop_privilege. This is
5309 necessary for the benefit of the queue runner, because there is no re-exec
5310 when deliver_drop_privilege is set.
5312 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5313 This has been fixed.
5315 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5316 was in use, was not putting the data itself into the right store pool;
5317 consequently, it could be overwritten for a subsequent message in the same
5318 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5321 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5323 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5324 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5325 the function were treating the return as a boolean value, which happened to
5326 work because 0=false and not-0=true, but is not correct code.
5328 68. The host_aton() function was not handling scoped IPv6 addresses (those
5329 with, for example, "%eth0" on the end) correctly.
5331 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5332 negated items (that is, ~something) in unsigned ints. Some compilers
5333 apparently mutter when there is no cast.
5335 70. If an address verification called from an ACL failed, and did not produce a
5336 user-specific message (i.e. there was only a "system" message), nothing was
5337 put in $acl_verify_message. In this situation, it now puts the system
5340 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5341 session; change 4.31/43 added the unwanted input to the log line - except
5342 that it did not do this in the start of session case. It now does.
5344 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5345 This is wrong and can cause the other end to generate a synchronization
5346 error if it is another Exim or anything else that does the synchronization
5347 check. A QUIT command is no longer sent after a timeout.
5349 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5350 during host lookups.
5352 74. The defer_ok option of callout verification was not working if it was used
5353 when verifying addresses in header lines, that is, for this case:
5355 verify = header_sender/callout=defer_ok
5357 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5358 those file descriptors could be used for SMTP connections. If anything
5359 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5360 could be sent to the SMTP client, causing chaos. The daemon now opens
5361 stdin, stdout, and stderr to /dev/null when it puts itself into the
5364 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5365 log by default. The user can override this with suitable Perl magic.
5367 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5368 the log message when discard triggers, was not working for the DATA ACL or
5369 for the non-SMTP ACL.
5371 78. Error message wording change in sieve.c.
5373 79. If smtp_accept_max_per_host was set, the number of connections could be
5374 restricted to fewer than expected, because the daemon was trying to set up
5375 a new connection before checking whether the processes handling previous
5376 connections had finished. The check for completed processes is now done
5377 earlier. On busy systems, this bug wouldn't be noticed because something
5378 else would have woken the daemon, and it would have reaped the completed
5381 80. If a message was submitted locally by a user whose login name contained one
5382 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5383 It caused a spool format error. I have fixed the spool reading code. A
5384 related problem was that the "from" clause in the Received: line became
5385 illegal because of the space(s). It is now covered by ${quote_local_part.
5387 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5390 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5392 83. Patch from Sieve maintainer to fix unterminated string problem in
5393 "vacation" handling.
5395 84. Some minor changes to the Linux configuration files to help with other
5396 OS variants using glibc.
5398 85. One more patch for Sieve to update vacation handling to latest spec.
5401 ----------------------------------------------------
5402 See the note above about the 4.44 and 4.50 releases.
5403 ----------------------------------------------------
5409 1. Change 4.43/35 introduced a bug that caused file counts to be
5410 incorrectly computed when quota_filecount was set in an appendfile
5413 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5414 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5417 3. Give more explanation in the error message when the command for a transport
5418 filter fails to execute.
5420 4. There are several places where Exim runs a non-Exim command in a
5421 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5422 was being done only for the command run by the queryprogram router. It is
5423 now done for all such subprocesses. The other cases are: ${run, transport
5424 filters, and the commands run by the lmtp and pipe transports.
5426 5. Some older OS have a limit of 256 on the maximum number of file
5427 descriptors. Exim was using setrlimit() to set 1000 as a large value
5428 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5429 systems. I've change it so that if it can't get 1000, it tries for 256.
5431 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5432 was an oversight, and furthermore, ever since the addition of extra
5433 controls (e.g. 4.43/32), the checks on when to allow different forms of
5434 "control" were broken. There should now be diagnostics for all cases when a
5435 control that does not make sense is encountered.
5437 7. $recipients is now available in the predata ACL (oversight).
5439 8. Tidy the search cache before the fork to do a delivery from a message
5440 received from the command line. Otherwise the child will trigger a lookup
5441 failure and thereby defer the delivery if it tries to use (for example) a
5442 cached ldap connection that the parent has called unbind on.
5444 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5445 of $address_data from the recipient verification was clobbered by the
5446 sender verification.
5448 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5449 was its contents. (It was OK if the option was not defined at all.)
5451 11. A "Completed" log line is now written for messages that are removed from
5452 the spool by the -Mrm option.
5454 12. $host_address is now set to the target address during the checking of
5455 ignore_target_hosts.
5457 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5458 being passed; this would have caused $sender_host_name to have been used if
5459 matching the list had actually called for a host name (not very likely,
5460 since this list is usually IP addresses). A host name is now passed as
5463 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5464 code that specifies a non-restarting handler (typically sigaction() in
5465 modern systems) in an attempt to fix a rare and obscure crash bug.
5467 15. Narrowed the window for a race in the daemon that could cause it to ignore
5468 SIGCHLD signals. This is not a major problem, because they are used only to
5469 wake it up if nothing else does.
5471 16. A malformed maildirsize file could cause Exim to calculate negative values
5472 for the mailbox size or file count. Odd effects could occur as a result.
5473 The maildirsize information is now recalculated if the size or filecount
5476 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5477 support for a long time. Removed HAVE_SYS_VFS_H.
5479 18. Updated exipick to current release from John Jetmore.
5481 19. Allow an empty sender to be matched against a lookup in an address list.
5482 Previously the only cases considered were a regular expression, or an
5485 20. Exim went into a mad DNS lookup loop when doing a callout where the
5486 host was specified on the transport, if the DNS lookup yielded more than
5489 21. The RFC2047 encoding function was originally intended for short strings
5490 such as real names; it was not keeping to the 75-character limit for
5491 encoded words that the RFC imposes. It now respects the limit, and
5492 generates multiple encoded words if necessary. To be on the safe side, I
5493 have increased the buffer size for the ${rfc2047: expansion operator from
5496 22. Failure to deliver a bounce message always caused it to be frozen, even if
5497 there was an errors_to setting on the router. The errors_to setting is now
5500 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5501 canonical form (fully expanded) before being placed in
5502 $sender_host_address.
5504 24. Updated eximstats to version 1.33
5506 25. Include certificate and key file names in error message when GnuTLS fails
5507 to set them up, because the GnuTLS error message doesn't include the name
5508 of the failing file when there is a problem reading it.
5510 26. Expand error message when OpenSSL has problems setting up cert/key files.
5513 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5514 (this can affect the format of dates).
5516 28. exim_tidydb, when checking for the continued existence of a message for
5517 which it has found a message-specific retry record, was not finding
5518 messages that were in split spool directories. Consequently, it was
5519 deleting retry records that should have stayed in existence.
5521 29. eximstats updated to version 1.35
5522 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5523 1.35 - bugfix such that pie charts by volume are generated correctly
5525 30. The SPA authentication driver was not abandoning authentication and moving
5526 on to the next authenticator when an expansion was forced to fail,
5527 contradicting the general specification for all authenticators. Instead it
5528 was generating a temporary error. It now behaves as specified.
5530 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5531 (the order specifies the preference for clients). The order is now AES256,
5532 AES128, 3DES, ARCFOUR128.
5534 31. Small patch to Sieve code - explicitly set From: when generating an
5537 32. Exim crashed if a remote delivery caused a very long error message to be
5538 recorded - for instance if somebody sent an entire SpamAssassin report back
5539 as a large number of 550 error lines. This bug was coincidentally fixed by
5540 increasing the size of one of Exim's internal buffers (big_buffer) that
5541 happened as part of the Exiscan merge. However, to be on the safe side, I
5542 have made the code more robust (and fixed the comments that describe what
5545 33. Some experimental protocols are using DNS PTR records for new purposes. The
5546 keys for these records are domain names, not reversed IP addresses. The
5547 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5548 leaves it alone. Component reversal etc. now happens only for IP addresses.
5551 34. The host_aton() function is supposed to be passed a string that is known
5552 to be a valid IP address. However, in the case of IPv6 addresses, it was
5553 not checking this. This is a hostage to fortune. Exim now panics and dies
5554 if the condition is not met. A case was found where this could be provoked
5555 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5556 components; fortuitously, this particular loophole had already been fixed
5557 by change 4.50/55 or 4.44/33 above.
5559 If there are any other similar loopholes, the new check in host_aton()
5560 itself should stop them being exploited. The report I received stated that
5561 data on the command line could provoke the exploit when Exim was running as
5562 exim, but did not say which command line option was involved. All I could
5563 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5564 running as the user.
5567 35. There was a buffer overflow vulnerability in the SPA authentication code
5568 (which came originally from the Samba project). I have added a test to the
5569 spa_base64_to_bits() function which I hope fixes it.
5572 36. The daemon start-up calls getloadavg() while still root for those OS that
5573 need the first call to be done as root, but it missed one case: when
5574 deliver_queue_load_max is set with deliver_drop_privilege. This is
5575 necessary for the benefit of the queue runner, because there is no re-exec
5576 when deliver_drop_privilege is set.
5578 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5579 was in use, was not putting the data itself into the right store pool;
5580 consequently, it could be overwritten for a subsequent message in the same
5581 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5584 38. Sometimes the final signoff response after QUIT could fail to get
5585 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5586 < 0 before doing a fflush(). This bug looks as though it goes back to the
5587 introduction of TLS in release 3.20, but "sometimes" must have been rare
5588 because the tests only now provoked it.
5594 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5595 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5596 Now it returns an integer. A number of calls were still expecting a T/F
5597 return. Fortuitously, in all cases, the tests worked in OK situations,
5598 which is the norm. However, things would have gone wrong on any write
5599 failures on the smtp file descriptor. This function is used when sending
5600 messages over SMTP and also when doing verify callouts.
5602 2. When Exim is called to do synchronous delivery of a locally submitted
5603 message (the -odf or -odi options), it no longer closes stderr before doing
5606 3. Implemented the mua_wrapper option.
5608 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5610 5. Implemented the functions header_remove(), header_testname(),
5611 header_add_at_position(), and receive_remove_recipient(), and exported them
5614 6. If an ACL "warn" statement specified the addition of headers, Exim already
5615 inserted X-ACL-Warn: at the start if there was no header name. However, it
5616 was not making this test for the second and subsequent header lines if
5617 there were newlines in the string. This meant that an invalid header could
5618 be inserted if Exim was badly configured.
5620 7. Allow an ACL "warn" statement to add header lines at the start or after all
5621 the Received: headers, as well as at the end.
5623 8. Added the rcpt_4xx retry error code.
5625 9. Added postmaster_mailfrom=xxx to callout verification option.
5627 10. Added mailfrom=xxxx to the callout verification option, for verify=
5630 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5631 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5633 12. Inserted some casts to stop certain compilers warning when using pointer
5634 differences as field lengths or precisions in printf-type calls (mostly
5635 affecting debugging statements).
5637 13. Added optional readline() support for -be (dynamically loaded).
5639 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5640 same clock tick as a message's arrival, so that its received time was the
5641 same as the "first fail" time on the retry record, and that message
5642 remained on the queue past the ultimate address timeout, every queue runner
5643 would try a delivery (because it was past the ultimate address timeout) but
5644 after another failure, the ultimate address timeout, which should have then
5645 bounced the address, did not kick in. This was a "< instead of <=" error;
5646 in most cases the first failure would have been in the next clock tick
5647 after the received time, and all would be well.
5649 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5650 being recognized when the domain list was tested by the match_domain
5651 condition in an expansion string.
5653 16. Added the ${str2b64: operator.
5655 17. Exim was always calling setrlimit() to set a large limit for the number of
5656 processes, without checking whether the existing limit was already
5657 adequate. (It did check for the limit on file descriptors.) Furthermore,
5658 errors from getrlimit() and setrlimit() were being ignored. Now they are
5659 logged to the main and panic logs, but Exim does carry on, to try to do its
5660 job under whatever limits there are.
5662 18. Imported PCRE 5.0.
5664 19. Trivial typo in log message " temporarily refused connection" (the leading
5667 20. If the log selector return_path_on_delivery was set and an address was
5668 redirected to /dev/null, the delivery process crashed because it assumed
5669 that a return path would always be set for a "successful" delivery. In this
5670 case, the whole delivery is bypassed as an optimization, and therefore no
5673 21. Internal re-arrangement: the function for sending a challenge and reading
5674 a response while authentication was assuming a zero-terminated challenge
5675 string. It's now changed to take a pointer and a length, to allow for
5676 binary data in such strings.
5678 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5680 23. Exim was not respecting finduser_retries when seeking the login of the
5681 uid under which it was called; it was always trying 10 times. (The default
5682 setting of finduser_retries is zero.) Also, it was sleeping after the final
5683 failure, which is pointless.
5685 24. Implemented tls_on_connect_ports.
5687 25. Implemented acl_smtp_predata.
5689 26. If the domain in control=submission is set empty, Exim assumes that the
5690 authenticated id is a complete email address when it generates From: or
5691 Sender: header lines.
5693 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5694 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5695 chown and chgrp in /bin and hostname in /usr/bin.
5697 28. Exim was keeping the "process log" file open after each use, just as it
5698 does for the main log. This opens the possibility of it remaining open for
5699 long periods when the USR1 signal hits a daemon. Occasional processlog
5700 errors were reported, that could have been caused by this. Anyway, it seems
5701 much more sensible not to leave this file open at all, so that is what now
5704 29. The long-running daemon process does not normally write to the log once it
5705 has entered its main loop, and it closes the log before doing so. This is
5706 so that log files can straightforwardly be renamed and moved. However,
5707 there are a couple of unusual error situations where the daemon does write
5708 log entries, and I had neglected to close the log afterwards.
5710 30. The text of an SMTP error response that was received during a remote
5711 delivery was being truncated at 512 bytes. This is too short for some of
5712 the long messages that one sometimes sees. I've increased the limit to
5715 31. It is now possible to make retry rules that apply only when a message has a
5716 specific sender, in particular, an empty sender.
5718 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5719 it possible to be selective about when SMTP synchronization is enforced.
5721 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5723 32. Implemented hosts_connection_nolog.
5725 33. Added an ACL for QUIT.
5727 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5730 35. Added mailbox_size and mailbox_filecount to appendfile.
5732 36. Added control = no_multiline_responses to ACLs.
5734 37. There was a bug in the logic of the code that waits for the clock to tick
5735 in the case where the clock went backwards by a substantial amount such
5736 that the microsecond fraction of "now" was more than the microsecond
5737 fraction of "then" (but the whole seconds number was less).
5739 38. Added support for the libradius Radius client library this is found on
5740 FreeBSD (previously only the radiusclient library was supported).
5746 1. When certain lookups returned multiple values in the form name=value, the
5747 quoting of the values was not always being done properly. Specifically:
5748 (a) If the value started with a double quote, but contained no whitespace,
5750 (b) If the value contained whitespace other than a space character (i.e.
5751 tabs or newlines or carriage returns) it was not quoted.
5752 This fix has been applied to the mysql and pgsql lookups by writing a
5753 separate quoting function and calling it from the lookup code. The fix
5754 should probably also be applied to nisplus, ibase and oracle lookups, but
5755 since I cannot test any of those, I have not disturbed their existing code.
5757 2. A hit in the callout cache for a specific address caused a log line with no
5758 reason for rejecting RCPT. Now it says "Previous (cached) callout
5759 verification failure".
5761 3. There was an off-by-one bug in the queryprogram router. An over-long
5762 return line was truncated at 256 instead of 255 characters, thereby
5763 overflowing its buffer with the terminating zero. As well as fixing this, I
5764 have increased the buffer size to 1024 (and made a note to document this).
5766 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5767 when Exim is waiting for an SMTP response from a remote server, Exim
5768 restarts its select() call on the socket, thereby resetting its timeout.
5769 This is not a problem when such interrupts are rare. Somebody set up a cron
5770 job to run exiwhat every 2 minutes, which is less than the normal select()
5771 timeout (5 or 10 minutes). This meant that the select() timeout never
5772 kicked in because it was always reset. I have fixed this by comparing the
5773 time when an interrupt arrives with the time at the start of the first call
5774 to select(). If more time than the timeout has elapsed, the interrupt is
5775 treated as a timeout.
5777 5. Some internal re-factoring in preparation for the addition of Sieve
5778 extensions (by MH). In particular, the "personal" test is moved to a
5779 separate function, and given an option for scanning Cc: and Bcc: (which is
5780 not set for Exim filters).
5782 6. When Exim created an email address using the login of the caller as the
5783 local part (e.g. when creating a From: or Sender: header line), it was not
5784 quoting the local part when it contained special characters such as @.
5786 7. Installed new OpenBSD configuration files.
5788 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5789 try to make them clearer.
5791 9. Callout options, other than the timeout value, were being ignored when
5792 verifying sender addresses in header lines. For example, when using
5794 verify = header_sender/callout=no_cache
5796 the cache was (incorrectly) being used.
5798 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5799 only the Cygwin environment.
5801 11. When return_path_on_delivery was set as a log selector, if different remote
5802 addresses in the same message used different return paths and parallel
5803 remote delivery occurred, the wrong values would sometimes be logged.
5804 (Whenever a remote delivery process finished, the return path value from
5805 the most recently started remote delivery process was logged.)
5807 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5808 lines when AUTH and/or TLS are in use. This is the "received protocol"
5809 field. Exim used to use "asmtp" for authenticated SMTP, without any
5810 indication (in the protocol name) for TLS use. Now it follows the RFC and
5811 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5812 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5813 names appear in log lines as well as in Received: header lines.
5815 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5816 extensions, and comparison tests, and to fix some bugs.
5818 14. Changes to the "personal" filter test:
5820 (1) The test was buggy in that it was just doing the equivalent of
5821 "contains" tests on header lines. For example, if a user's address was
5822 anne@some.where, the "personal" test would incorrectly be true for
5824 To: susanne@some.where
5826 This test is now done by extracting each address from the header in turn,
5827 and checking the entire address. Other tests that are part of "personal"
5828 are now done using regular expressions (for example, to check local parts
5829 of addresses in From: header lines).
5831 (2) The list of non-personal local parts in From: addresses has been
5832 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5833 taken from the Sieve specification recommendations.
5835 (3) If the message contains any header line starting with "List-" it is
5836 treated as non-personal.
5838 (4) The test for "circular" in the Subject: header line has been removed
5839 because it now seems ill-conceived.
5841 15. Minor typos in src/EDITME comments corrected.
5843 16. Installed latest exipick from John Jetmore.
5845 17. If headers_add on a router specified a text string that was too long for
5846 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5847 of string_sprintf() is now avoided.
5849 18. $message_body_size was not set (it was always zero) when running the DATA
5850 ACL and the local_scan() function.
5852 19. For the "mail" command in an Exim filter, no default was being set for
5853 the once_repeat time, causing a random time value to be used if "once" was
5854 specified. (If the value happened to be <= 0, no repeat happened.) The
5855 default is now 0s, meaning "never repeat". The "vacation" command was OK
5856 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5857 (I found it when inspecting the code).
5859 20. There is now an overall timeout for performing a callout verification. It
5860 defaults to 4 times the callout timeout, which applies to individual SMTP
5861 commands during the callout. The overall timeout applies when there is more
5862 than one host that can be tried. The timeout is checked before trying the
5863 next host. This prevents very long delays if there are a large number of
5864 hosts and all are timing out (e.g. when the network connections are timing
5865 out). The value of the overall timeout can be changed by specifying an
5866 additional sub-option for "callout", called "maxwait". For example:
5868 verify = sender/callout=5s,maxwait=20s
5870 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5871 to the end before writing, but this should make it even safer).
5873 22. Exim was forgetting that it had advertised PIPELINING for the second and
5874 subsequent messages on an SMTP connection. It was also not resetting its
5875 memory on STARTTLS and an internal HELO.
5877 23. When Exim logs an SMTP synchronization error within a session, it now
5878 records whether PIPELINING has been advertised or not.
5880 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5881 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5882 rather than long int.
5884 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5886 26. Added the never_mail option to autoreply.
5892 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5893 crash if the getsockname() call failed; this can happen if a connection is
5894 closed very soon after it is established. The problem was simply in the
5895 order in which certain operations were done, causing Exim to try to write
5896 to the SMTP stream before it had set up the file descriptor. The bug has
5897 been fixed by making things happen in the correct order.
5903 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5904 before the connection was closed, thus losing the rejection response.
5906 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5907 some early Solaris releases, but causes trouble in current releases where
5908 socklen_t is defined.
5910 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5913 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5916 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5917 that releasing the top of it at the end releases what was used for sub-
5918 expansions (unless the block got too big). However, discard this block if
5919 the first thing is a variable or header, so that we can use its block when
5920 it is dynamic (useful for very large $message_headers, for example).
5922 6. Lookups now cache *every* query, not just the most recent. A new, separate
5923 store pool is used for this. It can be recovered when all lookup caches are
5924 flushed. Lookups now release memory at the end of their result strings.
5925 This has involved some general refactoring of the lookup sources.
5927 7. Some code has been added to the store_xxx() functions to reduce the amount
5928 of flapping under certain conditions.
5930 8. log_incoming_interface used to affect only the <= reception log lines. Now
5931 it causes the local interface and port to be added to several more SMTP log
5932 lines, for example "SMTP connection from", and rejection lines.
5934 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5936 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5938 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5939 could be overwritten at the end of the current message (or the start of a
5940 new message if it was set in a HELO ACL). The value is now preserved for
5941 the duration of the SMTP connection.
5943 12. If a transport had a headers_rewrite setting, and a matching header line
5944 contained an unqualified address, that address was qualified, even if it
5945 did not match any rewriting rules. The underlying bug was that the values
5946 of the flags that permit the existence of unqualified sender and recipient
5947 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5948 non-local messages, and by -bnq for local messages) were not being
5949 preserved with the message after it was received.
5951 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5952 "next input=" as part of the text comprising the host identity instead of
5953 the correct text. The code was using the same buffer for two different
5954 strings. However, depending on which order the printing function evaluated
5955 its arguments, the bug did not always show up. Under Linux, for example, my
5956 test suite worked just fine.
5958 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5959 doesn't work with some older versions of Perl. It has been changed to "my",
5960 which in any case is probably the better facility to use.
5962 15. A really picky compiler found some instances of statements for creating
5963 error messages that either had too many or two few arguments for the format
5966 16. The size of the buffer for calls to the DNS resolver has been increased
5967 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5968 for addresses that have a lot of PTR records. This alleviates a problem; it
5969 does not fully solve it.
5971 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5972 buffer now truncates the list and logs the incident, which is the same
5973 action as happens when Exim is looking up a host name and its aliases.
5974 Previously in this situation something unpredictable would happen;
5975 sometimes it was "internal error: store_reset failed".
5977 18. If a server dropped the connection unexpectedly when an Exim client was
5978 using GnuTLS and trying to read a response, the client delivery process
5979 crashed while trying to generate an error log message.
5981 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5982 string, for example:
5984 warn message = H1: something\nH2: something
5986 the text was added as a single header line from Exim's point of view
5987 though it ended up OK in the delivered message. However, searching for the
5988 second and subsequent header lines using $h_h2: did not work. This has been
5989 fixed. Similarly, if a system filter added multiple headers in this way,
5990 the routers could not see them.
5992 20. Expanded the error message when iplsearch is called with an invalid key to
5993 suggest using net-iplsearch in a host list.
5995 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5996 ACLs are no longer actually imposed (and a message to that effect is
5999 22. If a "gecos" field in a passwd entry contained escaped characters, in
6000 particular, if it contained a \" sequence, Exim got it wrong when building
6001 a From: or a Sender: header from that name. A second bug also caused
6002 incorrect handling when an unquoted " was present following a character
6003 that needed quoting.
6005 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6006 was not being matched caselessly.
6008 24. Arranged for all hyphens in the exim.8 source to be escaped with
6011 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6012 properly. Recipient callout cache records were still being keyed to include
6013 the sender, even when use_sender was set false. This led to far more
6014 callouts that were necessary. The sender is no longer included in the key
6015 when use_sender is false.
6017 26. Added "control = submission" modifier to ACLs.
6019 27. Added the ${base62d: operator to decode base 62 numbers.
6021 28. dnsdb lookups can now access SRV records.
6023 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6024 the configuration file.
6026 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6027 (-v) mode. This makes the output for a verbose queue run more intelligible.
6029 31. Added a use_postmaster feature to recipient callouts.
6031 32. Added the $body_zerocount variable, containing the number of binary zero
6032 bytes in the message body.
6034 33. The time of last modification of the "new" subdirectory is now used as the
6035 "mailbox time last read" when there is a quota error for a maildir
6038 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6040 35. Added +ignore_unknown as a special item in host lists.
6042 36. Code for decoding IPv6 addresses in host lists is now included, even if
6043 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6044 address was recognized as an IP address, but was then not correctly decoded
6045 into binary, causing unexpected and incorrect effects when compared with
6052 1. Very minor rewording of debugging text in manualroute to say "list of
6053 hosts" instead of "hostlist".
6055 2. If verify=header_syntax was set, and a header line with an unqualified
6056 address (no domain) and a large number of spaces between the end of the
6057 name and the colon was received, the reception process suffered a buffer
6058 overflow, and (when I tested it) crashed. This was caused by some obsolete
6059 code that should have been removed. The fix is to remove it!
6061 3. When running in the test harness, delay a bit after writing a bounce
6062 message to get a bit more predictability in the log output.
6064 4. Added a call to search_tidyup() just before forking a reception process. In
6065 theory, someone could use a lookup in the expansion of smtp_accept_max_
6066 per_host which, without the tidyup, could leave open a database connection.
6068 5. Added the variables $recipient_data and $sender_data which get set from a
6069 lookup success in an ACL "recipients" or "senders" condition, or a router
6070 "senders" option, similar to $domain_data and $local_part_data.
6072 6. Moved the writing of debug_print from before to after the "senders" test
6075 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6076 problems for message scanning, either using a data ACL, or using
6077 local_scan() because the Received: header was not generated till after they
6078 were called (in order to set the time as the time of reception completion).
6079 I have revised the way this works. The header is now generated after the
6080 body is received, but before the ACL or local_scan() are called. After they
6081 are run, the timestamp in the header is updated.
6087 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6088 before starting a queue runner without re-exec. This happened only when
6089 deliver_drop_privilege was set or when the Exim user was set to root. The
6090 effect of the bug was that timeouts during subsequent deliveries caused
6091 crashes instead of being properly handled. The handler is now left at its
6092 default (and expected) setting.
6094 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6095 message, again when deliver_drop_privilege is set or Exim is run as root.
6096 The bug described in (1) was not present in this case, but the tidying up
6097 of the other signals was missing. I have made the two cases consistent.
6099 3. The ignore_target_hosts setting on a manualroute router was being ignored
6100 for hosts that were looked up using the /MX notation.
6102 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6105 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6106 operated on the sender address. After changing the $sender_address to <>
6107 for the sender address verify, Exim was re-instated it as the original
6108 (before rewriting) address, but remembering that it had rewritten it, so it
6109 wasn't rewriting it again. This bug also had the effect of breaking the
6110 sender address verification caching when the sender address was rewritten.
6112 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6113 This has been changed so that if the ip literal address matches
6114 ignore_target_hosts, the router declines.
6116 7. Added expansion conditions match_domain, match_address, and match_local_
6117 part (NOT match_host).
6119 8. The placeholder for the Received: header didn't have a length field set.
6121 9. Added code to Exim itself and to exim_lock to test for a specific race
6122 condition that could lead to file corruption when using MBX delivery. The
6123 issue is with the lockfile that is created in /tmp. If this file is removed
6124 after a process has opened it but before that process has acquired a lock,
6125 there is the potential for a second process to recreate the file and also
6126 acquire a lock. This could lead to two Exim processes writing to the file
6127 at the same time. The added code performs the same test as UW imapd; it
6128 checks after acquiring the lock that its file descriptor still refers to
6129 the same named file.
6131 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6132 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6135 11. Added the smtp_active_hostname option. If used, this will typically be made
6136 to depend on the incoming interface address. Because $interface_address is
6137 not set up until the daemon has forked a reception process, error responses
6138 that can happen earlier (such as "too many connections") no longer contain
6141 12. If an expansion in a condition on a "warn" statement fails because a lookup
6142 defers, the "warn" statement is abandoned, and the next ACL statement is
6143 processed. Previously this caused the whole ACL to be aborted.
6145 13. Added the iplsearch lookup type.
6147 14. Added ident_timeout as a log selector.
6149 15. Added tls_certificate_verified as a log selector.
6151 16. Added a global option tls_require_ciphers (compare the smtp transport
6152 option of the same name). This controls incoming TLS connections.
6154 17. I finally figured out how to make tls_require_ciphers do a similar thing
6155 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6156 before starting the TLS session.
6158 18. Tabs are now shown as \t in -bP output.
6160 19. If the log selector return_path_on_delivery was set, Exim crashed when
6161 bouncing a message because it had too many Received: header lines.
6163 20. If two routers both had headers_remove settings, and the first one included
6164 a superfluous trailing colon, the final name in the first list and the
6165 first name in the second list were incorrectly joined into one item (with a
6166 colon in the middle).
6172 1. Added -C and -D options to the exinext utility, mainly to make it easier
6173 to include in the automated testing, but these could be helpful when
6174 multiple configurations are in use.
6176 2. The exinext utility was not formatting the output nicely when there was
6177 an alternate port involved in the retry record key, nor when there was a
6178 message id as well (for retries that were specific to a specific message
6179 and a specific host). It was also confused by IPv6 addresses, because of
6180 the additional colons they contain. I have fixed the IPv4 problem, and
6181 patched it up to do a reasonable job for IPv6.
6183 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6184 delivery, the log line now contains "pipelined" if PIPELINING was used.
6186 4. An SMTP transport process used to panic and die if the bind() call to set
6187 an explicit outgoing interface failed. This has been changed; it is now
6188 treated in the same way as a connect() failure.
6190 5. A reference to $sender_host_name in the part of a conditional expansion
6191 that was being skipped was still causing a DNS lookup. This no longer
6194 6. The def: expansion condition was not recognizing references to header lines
6195 that used bh_ and bheader_.
6197 7. Added the _cache feature to named lists.
6199 8. The code for checking quota_filecount in the appendfile transport was
6200 allowing one more file than it should have been.
6202 9. For compatibility with Sendmail, the command line option
6210 and sets the incoming protocol and host name (for trusted callers). The
6211 host name and its colon can be omitted when only the protocol is to be set.
6212 Note the Exim already has two private options, -pd and -ps, that refer to
6213 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6214 "s", but I don't think that's a major issue.
6216 10. A number of refactoring changes to the code, none of which should affect
6219 (a) The number of logging options was getting close to filling up the
6220 32-bit word that was used as a bit map. I have split them into two classes:
6221 those that are passed in the argument to log_write(), and those that are
6222 only ever tested independently outside of that function. These are now in
6223 separate 32-bit words, so there is plenty of room for expansion again.
6224 There is no change in the user interface or the logging behaviour.
6226 (b) When building, for example, log lines, the code previously used a
6227 macro that called string_cat() twice, in order to add two strings. This is
6228 not really sufficiently general. Furthermore, there was one instance where
6229 it was actually wrong because one of the argument was used twice, and in
6230 one call a function was used. (As it happened, calling the function twice
6231 did not affect the overall behaviour.) The macro has been replaced by a
6232 function that can join an arbitrary number of extra strings onto a growing
6235 (c) The code for expansion conditions now uses a table and a binary chop
6236 instead of a serial search (which was left over from when there were very
6237 few conditions). Also, it now recognizes conditions like "pam" even when
6238 the relevant support is not compiled in: a suitably worded error message is
6239 given if an attempt is made to use such a condition.
6241 11. Added ${time_interval:xxxxx}.
6243 12. A bug was causing one of the ddress fields not to be passed back correctly
6244 from remote delivery subprocesses. The field in question was not being
6245 subsequently used, so this caused to problems in practice.
6247 13. Added new log selectors queue_time and deliver_time.
6249 14. Might have fixed a bug in maildirsizefile handling that threw up
6250 "unexpected character" debug warnings, and recalculated the data
6251 unnecessarily. In any case, I expanded the warning message to give more
6254 15. Added the message "Restricted characters in address" to the statements in
6255 the default ACL that block characters like @ and % in local parts.
6257 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6258 Three changes have been made:
6260 (a) There was a serious bug; a negative response to MAIL caused the whole
6261 recipient domain to be cached as invalid, thereby blocking all messages
6262 to all local parts at the same domain, from all senders. This bug has
6263 been fixed. The domain is no longer cached after a negative response to
6264 MAIL if the sender used is not empty.
6266 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6269 (c) A new callout option, "use_sender" has been added for people who want
6270 the modified behaviour.
6276 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6279 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6280 indeed breaks things for older releases.
6282 3. Added additional logging to the case where there is a problem reading data
6283 from a filter that is running in a subprocess using a pipe, in order to
6284 try to track down a specific problem.
6286 4. Testing facility fudge: when running in the test harness and attempting
6287 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6288 getting "No route to host". Convert this to a timeout.
6290 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6293 6. Some OS don't have socklen_t but use size_t instead. This affects the
6294 fifth argument of getsockopt() amongst other things. This is now
6295 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6296 can be set for individual OS. I have set it for SunOS5, OSF1, and
6297 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6298 some earlier ones do not.
6300 7. Change 4.30/15 was not doing the test caselessly.
6302 8. The standard form for an IPv6 address literal was being rejected by address
6303 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6304 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6305 this, as well as the form without the "IPv6" on the front (but only when
6306 address literals are enabled, of course).
6308 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6310 10. Exim crashed if a message with an empty sender address specified by -f
6311 encountered a router with an errors_to setting. This could be provoked only
6312 by a command such as
6316 where an empty string was supplied; "<>" did not hit this bug.
6318 11. Installed PCRE release 4.5.
6320 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6321 remained set. It is now erased.
6323 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6324 times from message ids (which are base 36 rather than the normal 62).
6326 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6327 were being counted as actual protocol errors, and logged if the log
6328 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6329 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6330 and DATA following a set of rejected RCPTs do not count as protocol errors.
6331 In other words, Exim assumes they were pipelined, though this may not
6332 actually be the case. Of course, in all cases the client gets an
6333 appropriate error code.
6335 15. If a lookup fails in an ACL condition, a message about the failure may
6336 be available; it is used if testing the ACL cannot continue, because most
6337 such messages specify what the cause of the deferral is. However, some
6338 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6339 that caused an old message to be retained and used if a later statement
6340 caused a defer, replacing the real cause of the deferral.
6342 16. If an IP address had so many PTR records that the DNS lookup buffer
6343 was not large enough to hold them, Exim could crash while trying to process
6344 the truncated data. It now detects and logs this case.
6346 17. Further to 4.21/58, another change has been made: if (and only if) the
6347 first line of a message (the first header line) ends with CRLF, a bare LF
6348 in a subsequent header line has a space inserted after it, so as not to
6349 terminate the header.
6351 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6352 unnecessarily, used atoi() instead of strtol(), and didn't check the
6353 termination when getting file sizes from file names by regex.
6355 19. Completely re-implemented the support for maildirsize files, in the light
6356 of a number of problems with the previous contributed implementation
6357 (4.30/29). In particular:
6359 . If the quota is zero, the maildirsize file is maintained, but no quota is
6362 . If the maildir directory does not exist, it is created before any attempt
6363 to write a maildirsize file.
6365 . The quota value in the file is just a cache; if the quota is changed in
6366 the transport, the new value overrides.
6368 . A regular expression is available for excluding directories from the
6371 20. The autoreply transport checks the characters in options that define the
6372 message's headers; it allows continued headers, but it was checking with
6373 isspace() after an embedded newline instead of explicitly looking for a
6376 21. If all the "regular" hosts to which an address was routed had passed their
6377 expiry times, and had not reached their retry times, the address was
6378 bounced, even if fallback hosts were defined. Now Exim should go on to try
6381 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6382 equivalent code in the SMTP transport. Some hosts send humungous responses
6383 to HELO/EHLO, more than 1024 it seems.
6385 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6386 gives compiler warnings in some environments. I've now done it "properly",
6389 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6390 (because of problems with the built-in one) was declared to return uschar *
6391 instead of char *, causing compiler failure.
6393 25. Fixed a file descriptor leak when processing alias/forward files.
6395 26. Fixed a minor format string issue in dbfn.c.
6397 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6399 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6400 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6403 29. When a sender address is verified, it is cached, to save repeating the test
6404 when there is more than one recipient in a message. However, when the
6405 verification involves a callout, it is possible for different callout
6406 options to be set for different recipients. It is too complicated to keep
6407 track of this in the cache, so now Exim always runs a verification when a
6408 callout is required, relying on the callout cache for the optimization.
6409 The overhead is duplication of the address routing, but this should not be
6412 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6413 to be verified with callout=postmaster, and the main callout worked but the
6414 postmaster check failed, the verification correctly failed. However, if a
6415 subsequent RCPT command asked for sender verification *without* the
6416 postmaster check, incorrect caching caused this verification also to fail,
6419 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6420 it was not caching the DNS options (qualify_single, search_parents) that
6421 were used when the lookup failed. A subsequent lookup with different
6422 options therefore always gave the same answer, though there were cases
6423 where it should not have. (Example: a "domains = !$mx_any" option on a
6424 dnslookup router: the "domains" option is always processed without any
6425 widening, but the router might have qualify_single set.) Now Exim uses the
6426 cached value only when the same options are set.
6428 32. Added John Jetmore's "exipick" utility to the distribution.
6430 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6431 than a timeout (e.g. a certificate is required, and is not provided), an
6432 Exim server now closes the connection immediately. Previously it waited for
6433 the client to close - but if the client is SSL, it seems that they each
6434 wait for each other, leading to a delay before one of them times out.
6436 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6437 maintained 0.8.x compatibility because I don't think many are using it, and
6438 it is clearly obsolete.
6440 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6443 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6444 client certificate was expired. A simple patch fixes this, though I don't
6445 understand the full logic of why the verify callback is called multiple
6448 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6449 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6450 which causes problems with some clients (such as the Certicom SSL Plus
6451 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6452 disables the coutermeasure allowing Eudora to connect."
6454 38. Exim was not checking that a write() to a log file succeeded. This could
6455 lead to Bad Things if a log got too big, in particular if it hit a file
6456 size limit. Exim now panics and dies if it cannot write to a log file, just
6457 as it does if it cannot open a log file.
6459 39. Modified OS/Makefile-Linux so that it now contains
6461 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6463 The two -D definitions ensure that Exim is compiled with large file
6464 support, which makes it possible to handle log files that are bigger than
6467 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6468 instance) a domain was checked against a named list that involved a lookup,
6469 causing $domain_data to be set, then another domain was checked against the
6470 same list, then the first domain was re-checked, the value of $domain_data
6471 after the final check could be wrong. In particular, if the second check
6472 failed, it could be set empty. This bug probably also applied to
6475 41. The strip_trailing_dot option was not being applied to the address given
6476 with the -f command-line option.
6478 42. The code for reading a message's header from the spool was incrementing
6479 $received_count, but never initializing it. This meant that the value was
6480 incorrect (doubled) while delivering a message in the same process in which
6481 it was received. In the most common configuration of Exim, this never
6482 happens - a fresh exec is done - but it can happen when
6483 deliver_drop_privilege is set.
6485 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6486 it now includes up to 150 characters of the unexpected data in the log
6489 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6490 and building data strings. The size of both of these buffers was 10 000
6491 bytes - far larger than anybody would *ever* want, thought I. Needless to
6492 say, somebody hit the limit. I have increased the maximum line length to
6493 20 000 and the maximum data length of concatenated lines to 100 000. I have
6494 also fixed two bugs, because there was no checking on these buffers. Tsk,
6495 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6496 buffer is too small.
6498 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6499 lsearch lookups. Now it does.
6501 46. When parsing a route_list item in a manualroute router, a fixed-length
6502 buffer was used for the list of hosts. I made this 1024 bytes long,
6503 thinking that nobody would ever have a list of hosts that long. Wrong.
6504 Somebody had a whole pile of complicated expansion conditions, and the
6505 string was silently truncated, leading to an expansion error. It turns out
6506 that it is easier to change to an unlimited length (owing to other changes
6507 that have happened since this code was originally written) than to build
6508 structure for giving a limitation error. The length of the item that
6509 expands into the list of hosts is now unlimited.
6511 47. The lsearch lookup could not handle data where the length of text line was
6512 more than 4095 characters. Such lines were truncated, leading to shortened
6513 data being returned. It should now handle lines of any length.
6515 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6516 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6519 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6520 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6521 no longer happens in comment lines. A list of replacements is now placed
6522 at the head of all of the source files, except those whose only change is
6523 to replace PERL_COMMAND in the very first #! line.
6525 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6526 messages on the queue, with a bottom-up merge sort, using code contributed
6527 by Michael Haardt. This should make operations like -bp somewhat faster on
6528 large queues. It won't affect queue runners, except when queue_run_in_order
6531 51. Installed eximstats 1.31 in the distribution.
6533 52. Added support for SRV lookups to the dnslookup router.
6535 53. If an ACL referred to $message_body or $message_body_end, the value was not
6536 reset for any messages that followed in the same SMTP session.
6538 54. The store-handling optimization for building very long strings was not
6539 differentiating between the different store pools. I don't think this
6540 actually made any difference in practice, but I've tidied it.
6542 55. While running the routers to verify a sender address, $sender_address
6543 was still set to the sender address. This is wrong, because when routing to
6544 send a bounce to the sender, it would be empty. Therefore, I have changed
6545 it so that, while verifying a sender address, $sender_address is set to <>.
6546 (There is no change to what happens when verifying a recipient address.)
6548 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6549 target A or AAAA records (if not already returned) without resetting the
6550 qualify_single or search_parents options of the DNS resolver. These are
6551 inappropriate in this case because the targets of MX and SRV records must
6552 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6553 target that, when qualified, matched something in the local domain. These
6554 two options are now turned off when doing these lookups.
6556 57. It seems that at least some releases of Reiserfs (which does not have the
6557 concept of a fixed number of inodes) returns zero and not -1 for the
6558 number of available inodes. This interacted badly with check_spool_inodes,
6559 which assumed that -1 was the "no such thing" setting. What I have done is
6560 to check that the total number of inodes is greater than zero before doing
6561 the test of how many are available.
6563 58. When a "warn" ACL statement has a log_message modifier, the message is
6564 remembered, and not repeated. This is to avoid a lot of repetition when a
6565 message has many recipients that cause the same warning to be written.
6566 However, Exim was preserving the list of already written lines for an
6567 entire SMTP session, which doesn't seem right. The memory is now reset if a
6568 new message is started.
6570 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6571 debugging output unless log_rewrite was also set.
6573 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6574 of (int)(handle) when we know that handle contains (void *)(-1).
6576 61. The Exim daemon panic-logs an error return when it closes the incoming
6577 connection. However "connection reset by peer" seems to be common, and
6578 isn't really an error worthy of noting specially, so that particular error
6581 62. When Exim is trying to find all the local interfaces, it used to panic and
6582 die if the ioctl to get the interface flags failed. However, it seems that
6583 on at least one OS (Solaris 9) it is possible to have an interface that is
6584 included in the list of interfaces, but for which you get a failure error
6585 for this call. This happens when the interface is not "plumbed" into a
6586 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6587 failure of the "get flags" call assumes that the interface is down.
6589 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6590 makes life easier for people who are doing arithmetic on fields extracted
6591 from dates, where you often get leading zeros that should not be
6592 interpreted as octal.
6594 64. Added qualify_domain to the redirect router, to override the global
6597 65. If a pathologically long header line contained very many addresses (the
6598 report of this problem mentioned 10 000) and each of them was rewritten,
6599 Exim could use up a very large amount of memory. (It kept on making new
6600 copies of the header line as it rewrote, and never released the old ones.)
6601 At the expense of a bit more processing, the header rewriting function has
6602 been changed so that it no longer eats memory in this way.
6604 66. The generation of the Received: header has been moved from the time that a
6605 message starts to be received, to the time that it finishes. The timestamp
6606 in the Received: header should now be very close to that of the <= log
6607 line. There are two side-effects of this change:
6609 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6610 logged header lines no longer include the local Received: line, because
6611 it has not yet been created. The same applies to a copy of the message
6612 that is returned to a non-SMTP sender when a message is rejected.
6614 (b) When a filter file is tested using -bf, no additional Received: header
6615 is added to the test message. After some thought, I decided that this
6618 This change does not affect the value of $received_for. It is still set
6619 after address rewriting, but before local_scan() is called.
6621 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6623 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6624 gave an unhelpful panic error message, and a defer error. I have managed to
6625 change this behaviour so that it now rejects any supplied certificate,
6626 which seems right, as the list of acceptable certificates is empty.
6628 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6629 gave an unhelpful defer error. I have not managed to make this reject any
6630 supplied certificates, but the error message it gives is "no certificate
6631 supplied", which is not helpful.
6633 70. exigrep's output now also includes lines that are not associated with any
6634 message, but which match the given pattern. Implemented by a patch from
6635 Martin Sluka, which also tidied up the Perl a bit.
6637 71. Recipient callout verification, like sender verification, was using <> in
6638 the MAIL FROM command. This isn't really the right thing, since the actual
6639 sender may affect whether the remote host accepts the recipient or not. I
6640 have changed it to use the actual sender in the callout; this means that
6641 the cache record is now keyed on a recipient/sender pair, not just the
6642 recipient address. There doesn't seem to be a real danger of callout loops,
6643 since a callout by the remote host to check the sender would use <>.
6644 [SEE ABOVE: changed after hitting problems.]
6646 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6647 temporary errors. However, in the case of such a code being given after
6648 the end of a data transmission (i.e. after ".") Exim was failing to write
6649 a retry record for the message. (Yes, there was some broken host that was
6650 actually sending 8xx at this point.)
6652 73. An unknown lookup type in a host list could cause Exim to panic-die when
6653 the list was checked. (An example that provoked this was putting <; in the
6654 middle of a list instead of at the start.) If this happened during a DATA
6655 ACL check, a -D file could be left lying around. This kind of configuration
6656 error no longer causes Exim to die; instead it causes a defer error. The
6657 incident is still logged to the main and panic logs.
6659 74. Buglet left over from Exim 3 conversion. The message "too many messages
6660 in one connection" was written to the rejectlog but not the mainlog, except
6661 when address rewriting (yes!) was being logged.
6663 75. Added write_rejectlog option.
6665 76. When a system filter was run not as root (that is, when system_filter_user
6666 was set), the values of the $n variables were not being returned to the
6667 main process; thus, they were not subsequently available in the $sn
6670 77. Added +return_path_on_delivery log selector.
6672 78. A connection timeout was being treated differently from recipients deferred
6673 when testing hosts_max_try with a message that was older than the host's
6674 retry timeout. (The host should not be counted, thus allowing all hosts to
6675 be tried at least once before bouncing.) This may have been the cause of an
6676 occasionally reported bug whereby a message would remain on the queue
6677 longer than the retry timeout, but would be bounced if a delivery was
6678 forced. I say "may" because I never totally pinned down the problem;
6679 setting up timeout/retry tests is difficult. See also the next item.
6681 79. The ultimate address timeout was not being applied to errors that involved
6682 a combination of host plus message (for example, a timeout on a MAIL
6683 command). When an address resolved to a number of possible hosts, and they
6684 were not all tried for each delivery (e.g. because of hosts_max_try), a
6685 message could remain on the queue longer than the retry timeout.
6687 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6690 81. Fixed an obscure SMTP outgoing bug which required at least the following
6691 conditions: (a) there was another message waiting for the same server;
6692 (b) the server returned 5xx to all RCPT commands in the first message so
6693 that the message was not completed; (c) the server dropped the connection
6694 or gave a negative response to the RSET that Exim sends to abort the
6695 transaction. The observed case was a dropped connection after DATA that had
6696 been sent in pipelining mode. That is, the server had advertised PIPELINING
6697 but was not implementing it correctly. The effect of the bug was incorrect
6698 behaviour, such as trying another host, and this could lead to a crash.
6704 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6705 and daemon.c were passed as pointers to ints; they should have been
6706 pointers to socklen_t variables (which are typically unsigned ints).
6708 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6711 3. Fixed a really odd bug that affected only the testing scheme; patching a
6712 certain fixed string in the binary changed the value of another string that
6713 happened to be identical to the end of the original first string.
6715 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6716 name", it returns that address as the IP address. On some operating
6717 systems (e.g. Solaris), it also passes back the IP address string as the
6718 "host name". However, on others (e.g. Linux), it passes back an empty
6719 string. Exim wasn't checking for this, and was changing the host name to an
6720 empty string, assuming it had been canonicalized.
6722 5. Although rare, it is permitted to have more than one PTR record for a given
6723 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6724 all the names associated with an address, because they do in Solaris.
6725 However, it seems that they do not in Linux for data that comes from the
6726 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6727 I found this out when I moved to a new Linux workstation and tried to run
6728 the Exim test suite.
6730 To get round this problem I have changed the code so that it now does its
6731 own call to the DNS to look up PTR records when searching for a host name.
6732 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6733 addresses that are only in /etc/hosts are still found.
6735 This behaviour is, however, controlled by an option called host_lookup_
6736 order, which defaults to "bydns:byaddr". If people want to use the other
6737 order, or indeed, just use one or the other means of lookup, they can
6738 specify it in this variable.
6740 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6741 some operating systems, this comes back from gethostbyaddr() as an empty
6742 string, and this is what Exim used to test for. However, it seems that in
6743 other systems, "." is yielded. Exim now tests for this case too.
6745 7. The values of check_spool_space and check_log_space are now held internally
6746 as a number of kilobytes instead of an absolute number of bytes. If a
6747 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6748 kilobyte. This means that much larger values can be stored.
6750 8. Exim monitor: an attempt to get the action menu when not actually pointing
6751 at a message produces an empty menu entitled "No message selected". This
6752 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6753 no entries in it ("Shell widget menu has zero width and/or height"). So I
6754 have added a single, blank menu entry in this case.
6756 9. Added ${quote_local_part.
6758 10. MIME decoding is now applied to the contents of Subject: header lines when
6761 11. Now that a reference to $sender_host_address automatically causes a reverse
6762 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6763 host lookup before query-style lookups in lists that might use this
6764 variable. This has therefore been abolished, and the "net-" prefix is no
6765 longer necessary for query-style lookups.
6767 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6768 have been a typo for LFLAGS, so it has been changed.
6770 13. The install script calls Exim with "-C /dev/null" in order to find the
6771 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6772 to be output. However, since Exim outputs its version number before the
6773 error, it didn't break the script. It just looked ugly. I fixed this by
6774 always allowing "-C /dev/null" if the caller is root.
6776 14. Ignore overlarge ACL variable number when reading spool file - insurance
6777 against a later release with more variables having written the file.
6779 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6780 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6781 this, as well as the form without the "IPv6" on the front.
6783 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6784 OS/Makefile-Darwin file.
6786 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6787 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6789 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6790 variable. However, some OS use ints for inodes. I've added cast to long int
6791 to get rid of the compiler warning.
6793 19. I had forgotten to lock out "/../" in configuration file names when
6794 ALT_CONFIG_PREFIX was set.
6796 20. Routers used for verification do not need to specify transports. However,
6797 if such a router generated a host list, and callout was configured, Exim
6798 crashed, because it could not find a port number from the (non-existent)
6799 transport. It now assumes port 25 in this circumstance.
6801 21. Added the -t option to exigrep.
6803 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6804 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6805 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6806 them is set, LOOKUP_LSEARCH is forced.
6808 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6810 24. Added sender and host information to the "rejected by local_scan()" log
6811 line; previously there was no indication of these.
6813 25. Added .include_if_exists.
6815 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6816 Linux. It turns out that not all file systems support this. Apparently some
6817 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6818 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6819 not supported on the file descriptor, is now ignored when Exim is trying to
6820 sync a directory. This applies only to Linux.
6822 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6824 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6825 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6828 29. Added support for maildirsize files from supplied patch (modified a bit).
6830 30. The use of :fail: followed by an empty string could lead Exim to respond to
6831 sender verification failures with (e.g.):
6833 550 Verification failed for <xxx>
6834 550 Sender verify failed
6836 where the first response line was missing the '-' that indicates it is not
6837 the final line of the response.
6839 31. The loop for finding the name of the user that called Exim had a hardwired
6840 limit of 10; it now uses the value of finduser_retries, which is used for
6841 all other user lookups.
6843 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6846 33. Exim was neglecting to zero errno before one call of strtol() when
6847 expanding a string and expecting an integer value. On some systems this
6848 resulted in spurious "integer overflow" errors. Also, it was casting the
6849 result into an int without checking.
6851 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6852 did not work. The code looks as if it has *never* worked, though it appears
6853 to have been documented since at least release 1.62. I have made it work.
6855 35. The "timeout_DNS" error in retry rules, also documented since at least
6856 1.62, also never worked. As it isn't clear exactly what this means, and
6857 clearly it isn't a major issue, I have abolished the feature by treating it
6858 as "timeout", and writing a warning to the main and panic logs.
6860 36. The display of retry rules for -brt wasn't always showing the error code
6863 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6864 timeout_connect_A, timeout_connect_MX.
6866 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6867 to the empty sender.
6869 39. The daemon was not analysing the content of -oX till after it had closed
6870 stderr and disconnected from the controlling terminal. This meant that any
6871 syntax errors were only noted on the panic log, and the return code from
6872 the command was 0. By re-arranging the code a little, I've made the
6873 decoding happen first, so such errors now appear on stderr, and the return
6874 code is 1. However, the actual setting up of the sockets still happens in
6875 the disconnected process, so errors there are still only recorded on the
6878 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6879 connections (as happens on some IP stacks) was logged at start up time as
6880 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6881 it from "IPv6 and IPv4", which means that two separate sockets are being
6884 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6885 says whether AF_INET or AF_INET6 was passed as an argument.
6887 42. Exiwhat output was messed up when time zones were included in log
6890 43. Exiwhat now gives more information about the daemon's listening ports,
6891 and whether -tls-on-connect was used.
6893 44. The "port" option of the smtp transport is now expanded.
6895 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6896 silently ignored. Now an error message is written to the main and panic
6899 46. There's a new ACL modifier called "logwrite" which writes to a log file
6900 as soon as it is encountered.
6902 47. Added $local_user_uid and $local_user_gid at routing time.
6904 48. Exim crashed when trying to verify a sender address that was being
6907 49. Exim was recognizing only a space character after ".include". It now also
6908 recognizes a tab character.
6910 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6911 extracting the relevant information from the specification. The man page no
6912 longer contains scrambled data for the -d option, and I've added a section
6913 at the front about calling Exim under different names.
6915 51. Added "extra_headers" argument to the "mail" command in filter files.
6917 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6920 53. Installed eximstats 1.29.
6922 54. Added transport_filter_timeout as a generic transport option.
6924 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6925 Cc: header lines. This was required by RFC 822, but it not required by RFC
6928 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6929 incoming messages that did not have them. Now it does so only if the
6930 message originates locally, that is, if there is no associated remote host
6931 address. When Resent- header lines are present, this applies to the Resent-
6932 lines rather than the non-Resent- lines.
6934 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6935 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6937 58. Messages for configuration errors now include the name of the main
6938 configuration file - useful now that there may be more than one file in a
6939 list (.included file names were always shown).
6941 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6942 for those rare installations that do not start the daemon as root or run it
6943 setuid root. I've cut out the call to initgroups() if the daemon is not
6946 60. The Exim user and group can now be bound into the binary as text strings
6947 that are looked up at the start of Exim's processing.
6949 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6951 62. Added $mailstore_basename variable.
6953 63. Installed patch to sieve.c from Michael Haardt.
6955 64. When Exim failed to open the panic log after failing to open the main log,
6956 the original message it was trying to log was written to stderr and debug
6957 output, but if they were not available (the usual case in production), it
6958 was lost. Now it is written to syslog before the two lines that record the
6959 failures to open the logs.
6961 65. Users' Exim filters run in subprocesses under the user's uid. It is
6962 possible for a "deliver" command or an alias in a "personal" command to
6963 provoke an address rewrite. If logging of address rewriting is configured,
6964 this fails because the process is not running as root or exim. There may be
6965 a better way of dealing with this, but for the moment (because 4.30 needs
6966 to be released), I have disabled address rewrite logging when running a
6967 filter in a non-root, non-exim process.
6973 1. The buildconfig auxiliary program wasn't quoting the value set for
6974 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6975 not defined. This bug was masked in 4.22 by the effect that was fixed in
6978 2. Some messages that were rejected after a message id was allocated were
6979 shown as "incomplete" by exigrep. It no longer does this for messages that
6980 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6982 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6983 have allow_domain_literals set, the ID did not get logged in the <= line.
6984 Domain literals are now always recognized in Message-ID: header lines.
6986 4. The first argument for a ${extract expansion item is the key name or field
6987 number. Leading and trailing spaces in this item were not being ignored,
6988 causing some misleading effects.
6990 5. When deliver_drop_privilege was set, single queue runner processes started
6991 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6992 same command in the process it spins off) were not dropping privilege.
6994 6. When the daemon running as "exim" started a queue runner, it always
6995 re-executed Exim in the spun-off process. This is a waste of effort when
6996 deliver_drop_privilege is set. The new process now just calls the
6997 queue-runner function directly.
7003 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7006 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7007 ignored. Though the use of -oP was forcing the writing of a pid file, it
7008 was always written to the default place.
7010 3. If the message "no IP address found for host xxxx" is generated during
7011 incoming verification, it is now followed by identification of the incoming
7012 connection (so you can more easily find what provoked it).
7014 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7016 5. Added some features to "harden" Exim a bit more against certain attacks:
7018 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7019 be put in Local/Makefile. This is like the never_users runtime option,
7020 but it cannot be overridden. The default setting is "root".
7022 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7023 prefix string with which any file named in a -C command line option
7026 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7027 is retained for -C and -D only if the caller of Exim is root. Without
7028 it, the exim user may also use -C and -D and retain privilege.
7030 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7031 command line option is disabled.
7033 6. Macro names set by the -D option must start with an upper case letter, just
7034 like macro names defined in the configuration file.
7036 7. Added "dereference=" facility to LDAP.
7038 8. Two instances of the typo "uknown" in the source files are fixed.
7040 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7041 the Configure-Makefile script screwed up while processing it.
7043 10. Incorporated PCRE 4.4.
7045 11. The SMTP synchronization check was not operating right at the start of an
7046 SMTP session. For example, it could not catch a HELO sent before the client
7047 waited for the greeting. There is now a check for outstanding input at the
7048 point when the greeting is written. Because of the duplex, asynchronous
7049 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7050 way, but not yet received, when the check is performed.
7052 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7053 on TCP/IP sockets, because this apparently causes some broken clients to
7056 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7057 unchanged) from the Cygwin maintainer.
7059 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7060 was supported instead of testing for mbx. Effectively a typo.
7062 15. The spa authenticator server code was not checking that the input it
7063 received was valid base64.
7065 16. The debug output line for the "set" modifier in ACLs was not showing the
7066 name of the variable that was being set.
7068 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7070 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7071 lookup on its own. Something else was needed to trigger the lookup. For
7072 example, a match in host_lookup or the need for a host name in a host list.
7073 Now, if $sender_host_name is referenced and the host name has not yet been
7074 looked up, a lookup is performed. If the lookup fails, the variable remains
7075 empty, and $host_lookup_failed is set to "1".
7077 19. Added "eqi" as a case-independent comparison operator.
7079 20. The saslauthd authentication condition could segfault if neither service
7080 nor realm was specified.
7082 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7083 error message that was logged was misleading, and incoming SMTP
7084 connections were dropped. The message is now more accurate, and temporary
7085 errors are given to SMTP connections.
7087 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7088 (except RSET) with a 421 error, until QUIT is received. However, it was
7089 failing to send a response to QUIT.
7091 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7092 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7093 ACL is now run after the helo_try_verify_hosts code.
7095 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7096 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7097 case-independent, so other case variants are also recognized). Apparently
7098 some systems use these upper case variants.
7100 25. If more than two messages were waiting for the same host, and a transport
7101 filter was specified for the transport, Exim sent two messages over the
7102 same TCP/IP connection, and then failed with "socket operation on non-
7103 socket" when it tried to send the third.
7105 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7107 27. The extern definition of crypt16() in expand.c was not being excluded when
7108 the OS had its own crypt16() function.
7110 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7111 as a preferred synonym for return_size_limit, both as an option and as an
7114 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7116 30. Changed the default configuration ACL to relax the local part checking rule
7117 for addresses that are not in any local domains. For these addresses,
7118 slashes and pipe symbols are allowed within local parts, but the sequence
7119 /../ is explicitly forbidden.
7121 31. SPA server authentication was not clearing the challenge buffer before
7124 32. log_message in a "warn" ACL statement was writing to the reject log as
7125 well as to the main log, which contradicts the documentation and doesn't
7126 seem right (because no rejection is happening). So I have stopped it.
7128 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7129 However, I am unable to do any testing of this.
7131 34. Fixed an infelicity in the appendfile transport. When checking directories
7132 for a mailbox, to see if any needed to be created, it was accidentally
7133 using path names with one or more superfluous leading slashes; tracing
7134 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7136 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7137 contents are added to the log line that is written for every discarded
7138 recipient. (Previously a log_message setting was ignored.)
7140 36. The ${quote: operator now quotes the string if it is empty.
7142 37. The install script runs exim in order to find its version number. If for
7143 some reason other than non-existence or emptiness, which it checks, it
7144 could not run './exim', it was installing it with an empty version number,
7145 i.e. as "exim-". This error state is now caught, and the installation is
7148 38. An argument was missing from the function that creates an error message
7149 when Exim fails to connect to the socket for saslauthd authentication.
7150 This could cause Exim to crash, or give a corrupted message.
7152 39. Added isip, isip4, and isip6 to ${if conditions.
7154 40. The ACL variables $acl_xx are now saved with the message, and can be
7155 accessed later in routers, transports, and filters.
7157 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7158 strings in the file are not string-expanded.
7160 42. If a MAIL command specified a SIZE value that was too large to fit into an
7161 int variable, the check against message_size_limit failed. Such values are
7162 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7163 day this will have to be increased, but I don't think I want to be around
7164 when emails are that large.
7171 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7172 iconv() is not standard in FreeBSD.
7174 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7175 IPv6 enabled. The observed symptom was a segmentation fault on return from
7176 the function os_common_find_running_interfaces() in src/os.c.
7178 3. In the check_special_case() function in daemon.c I had used "errno" as an
7179 argument name, which causes warnings on some systems. This was basically a
7180 typo, since it was named "eno" in the comments!
7182 4. The code that waits for the clock to tick (at a resolution of some fraction
7183 of a second) so as to ensure message-id uniqueness was always waiting for
7184 at least one whole tick, when it could have waited for less. [This is
7185 almost certainly not relevant at current processor speeds, where it is
7186 unlikely to ever wait at all. But we try to future-proof.]
7188 5. The function that sleeps for a time interval that includes fractions of a
7189 second contained a race. It did not block SIGALRM between setting the
7190 timer, and suspending (a couple of lines later). If the interval was short
7191 and the sigsuspend() was delayed until after it had expired, the suspension
7192 never ended. On busy systems this could lead to processes getting stuck for
7195 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7196 process, before it forks any delivery processes. The open lookup caching
7197 mechanism meant that the open file or database connection was passed into
7198 the delivery process. The problem was that delivery processes always tidy
7199 up cached lookup data. This could cause a problem for the next delivery
7200 process started by the queue runner, because the external queue runner
7201 process does not know about the closure. So the next delivery process
7202 still has data in the lookup cache. In the case of a file lookup, there was
7203 no problem because closing a file descriptor in a subprocess doesn't affect
7204 the parent. However, if the lookup was caching a connection to a database,
7205 the connection was closed, and the second delivery process was likely to
7206 see errors such as "PGSQL: query failed: server closed the connection
7207 unexpectedly". The problem has been fixed by closing all cached lookups
7208 in a queue runner before running a delivery process.
7210 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7211 seem to have the "const" qualifier which it has on other OS. I've
7214 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7215 *of the same type* (client or server) with the same public name that an
7216 error should be diagnosed.
7218 9. When Exim looked up a host name for an IP address, but failed to find the
7219 original IP address when looking up the host name (a safety check), it
7220 output the message "<ip address> does not match any IP for NULL", which was
7221 confusing, to say the least. The bug was that the host name should have
7222 appeared instead of "NULL".
7224 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7225 user that is built into the binary, and the -C or -D options is used, root
7226 privilege is dropped before the configuration file is read. In addition,
7227 logging is switched to stderr instead of the normal log files. If the
7228 configuration then re-defines the Exim user, the unprivileged environment
7229 is probably not what is expected, so Exim logs a panic warning message (but
7232 However, if deliver_drop_privilege is set, the unprivileged state may well
7233 be exactly what is intended, so the warning has been cut out in that case,
7234 and Exim is allowed to try to write to its normal log files.
7240 1. smtp_return_error_details was not giving details for temporary sender
7241 or receiver verification errors.
7243 2. Diagnose a configuration error if two authenticators have the same public
7246 3. Exim used not to create the message log file for a message until the first
7247 delivery attempt. This could be confusing when incoming messages were held
7248 for policy or load reasons. The message log file is now created at the time
7249 the message is received, and an initial "Received" line is written to it.
7251 4. The automatically generated man page for command line options had a minor
7252 bug that caused no ill effects; however, a more serious problem was that
7253 the procedure for building the man page automatically didn't always
7254 operate. Consequently, release 4.20 contains an out-of-date version. This
7255 shouldn't happen again.
7257 5. When building Exim with embedded Perl support, the script that builds the
7258 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7259 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7261 6. The freeze_tell option was not being used for messages that were frozen on
7262 arrival, either by an ACL or by local_scan().
7264 7. Added the smtp_incomplete_transaction log selector.
7266 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7267 was accepting AUTH without a new EHLO.
7269 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7270 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7273 10. Logging of TCP/IP connections (when configured) now happens in the main
7274 daemon process instead of the child process, so that the TCP/IP connection
7275 count is more accurate (but it can never be perfect).
7277 11. The use of "drop" in a nested ACL was not being handled correctly in the
7278 outer ACL. Now, if condition failure induced by the nested "drop" causes
7279 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7280 or "require"), the connection is dropped.
7282 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7283 that yield "discard" can now be used with an "accept" or a "discard" verb,
7284 but an error is generated for any others (because I can't see a useful way
7285 to define what should happen).
7287 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7288 are now processed in the same way as lines in the Exim configuration file.
7289 In particular, continuation lines are supported.
7291 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7293 15. Added -ti meaning -t -i.
7295 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7296 domains, and warn by logging if others are found.
7298 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7299 returned from GIFCONF when Exim is trying to find the list of interfaces on
7300 a host. The code in os.c has been modified to copy each ifreq to an aligned
7301 structure in all cases.
7303 Also, in some cases, the returned ifreq's were being copied to a 'struct
7304 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7305 means the last couple of bytes of an IPv6 address could be chopped if the
7306 ifreq contained only a normal sockaddr (14 bytes storage).
7308 18. Named domain lists were not supported in the hosts_treat_as_local option.
7309 An entry such as +xxxx was not recognized, and was treated as a literal
7312 19. Ensure that header lines added by a DATA ACL are included in the reject log
7313 if the ACL subsequently rejects the message.
7315 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7316 MD5 (which is deprecated).
7318 21. When testing a filter file using -bf, Exim was writing a message when it
7319 took the sender from a "From " line in the message, but it was not doing so
7320 when it took $return_path from a Return-Path: header line. It now does.
7322 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7323 with a valid header line field name (a series of printing characters
7324 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7326 23. Changed "disc" in the source to "disk" to conform to the documentation and
7327 the book and for uniformity.
7329 24. Ignore Sendmail's -Ooption=value command line item.
7331 25. When execve() failed while trying to run a command in a pipe transport,
7332 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7333 could be confused with a return value of 69 from the command itself. This
7334 has been changed to 127, the value the shell returns if it is asked to run
7335 a non-existent command. The wording for the related log line suggests a
7336 non-existent command as the problem.
7338 26. If received_header_text expands to an empty string, do not add a Received:
7339 header line to the message. (Well, it adds a token one on the spool, but
7340 marks it "old" so that it doesn't get used or transmitted.)
7342 27. Installed eximstats 1.28 (addition of -nt option).
7344 28. There was no check for failure on the call to getsockname() in the daemon
7345 code. This can fail if there is a shortage of resources on the system, with
7346 ENOMEM, for example. A temporary error is now given on failure.
7348 29. Contrary to the C standard, it seems that in some environments, the
7349 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7350 program. Exim now does this explicitly; it affects the formatting of
7351 timestamps using strftime().
7353 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7354 complaints. I've now initialized all the variables, to avoid this.
7356 32. Header lines added by a system filter were not being "seen" during
7357 transport-time rewrites.
7359 33. The info_callback() function passed to OpenSSL is set up with type void
7360 (*)(SSL *, int, int), as described somewhere. However, when calling the
7361 function (actually a macro) that sets it up, the type void(*)() is
7362 expected. I've put in a cast to prevent warnings from picky compilers.
7364 34. If a DNS black list lookup found a CNAME record, but there were no A
7365 records associated with the domain it pointed at, Exim crashed.
7367 35. If a DNS black list lookup returned more than one A record, Exim ignored
7368 all but the first. It now scans all returned addresses if a particular IP
7369 value is being sought. In this situation, the contents of the
7370 $dnslist_value variable are a list of all the addresses, separated by a
7373 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7374 to accept a host name and all its aliases if the forward lookup for any of
7375 them yielded the IP address of the incoming connection. Now it accepts only
7376 those names whose forward lookup yields the correct IP address. Any other
7377 names are discarded. This closes a loophole whereby a rogue DNS
7378 administrator could create reverse DNS records to break through a
7379 wildcarded host restriction in an ACL.
7381 37. If a user filter or a system filter that ran in a subprocess used any of
7382 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7383 the wrong values were passed to the pipe command ($thisaddress had the
7384 value of $0, $0 had the value of $1, etc). This bug was introduced by
7385 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7387 38. Improved the line breaking for long SMTP error messages from ACLs.
7388 Previously, if there was no break point between 40 and 75 characters, Exim
7389 left the rest of the message alone. Two changes have been made: (a) I've
7390 reduced the minimum length to 35 characters; (b) if it can't find a break
7391 point between 35 and 75 characters, it looks ahead and uses the first one
7392 that it finds. This may give the occasional overlong line, but at least the
7393 remaining text gets split now.
7395 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7396 file descriptors might be low, and that setting 1000 would always raise it.
7397 It turns out that in some environments, the limit is already over 1000 and
7398 that lowering it causes trouble. So now Exim takes care not to decrease it.
7400 40. When delivering a message, the value of $return_path is set to $sender_
7401 address at the start of routing (routers may change the value). By an
7402 oversight, this default was not being set up when an address was tested by
7403 -bt or -bv, which affected the outcome if any router or filter referred to
7406 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7407 log only when "message" or "log_message" are set. However, if one of the
7408 conditions was an address verification, or a call to a nested ACL, the
7409 messages generated by the underlying test were being passed through. This
7410 no longer happens. The underlying message is available in $acl_verify_
7411 message for both "message" and "log_message" expansions, so it can be
7412 passed through if needed.
7414 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7415 new expansion $bh_ to give the encoded byte string without charset
7416 translation. Translation happens only if iconv() is available; HAVE_ICONV
7417 indicates this at build time. HEADERS_CHARSET gives the charset to
7418 translate to; headers_charset can change it in the configuration, and
7419 "headers charset" can change it in an individual filter file.
7421 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7422 that creates RFC 2047 encoded "words" labels them as that charset instead
7423 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7424 expansion operator; (ii) when Exim creates a From: line for a local
7425 message; (iii) when a header line is rewritten to include a "phrase" part.
7427 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7428 buggy, causing it to skip the first lines of messages whose message ID
7429 ended in 'D'. This would not have bitten before Exim release 4.14, because
7430 message IDs were unlikely to end in 'D' before then. The effect was to have
7431 incorrect size information for certain domains.
7433 45. #include "config.h" was missing at the start of the crypt16.c module. This
7434 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7437 46. If there was a timeout during a "random" callout check, Exim treated it as
7438 a failure of the random address, and carried on sending RSET and the real
7439 address. If the delay was just some slowness somewhere, the response to the
7440 original RCPT would be taken as a response to RSET and so on, causing
7441 mayhem of various kinds.
7443 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7444 when I implemented it. It didn't allow for the fact that some option values
7445 may legitimately be negative (e.g. size_addition), and it didn't even do
7446 the right test for positive values.
7448 48. Domain names in DNS records are case-independent. Exim always looks them up
7449 in lower case. Some resolvers return domain names in exactly the case they
7450 appear in the zone file, that is, they may contain uppercase letters. Not
7451 all resolvers do this - some return always lower case. Exim was treating a
7452 change of case by a resolver as a change of domain, similar to a widening
7453 of a domain abbreviation. This triggered its re-routing code and so it was
7454 trying to route what was effectively the same domain again. This normally
7455 caused routing to fail (because the router wouldn't handle the domain
7456 twice). Now Exim checks for this case specially, and just changes the
7457 casing of the domain that it ultimately uses when it transmits the message
7460 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7463 50. If a filter generated a file delivery with a non-absolute name (possible if
7464 no home directory exists for the router), the forbid_file option was not
7467 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7468 the existing equality matching.
7470 52. Exim was using ints instead of ino_t variables in some places where it was
7471 dealing with inode numbers.
7473 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7474 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7475 called TMPDIR, and if it finds it is different, it changes its value.
7477 54. The smtp_printf() function is now made available to local_scan() so
7478 additional output lines can be written before returning. There is also an
7479 smtp_fflush() function to enable the detection of a dropped connection.
7480 The variables smtp_input and smtp_batched_input are exported to
7483 55. Changed the default runtime configuration: the message "Unknown user"
7484 has been removed from the ACL, and instead placed on the localuser router,
7485 using the cannot_route_message feature. This means that any verification
7486 failures that generate their own messages won't get overridden. Similarly,
7487 the "Unrouteable address" message that was in the ACL for unverifiable
7488 relay addresses has also been removed.
7490 56. Added hosts_avoid_esmtp to the smtp transport.
7492 57. The exicyclog script was not checking for the esoteric option
7493 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7494 will work only if exicyclog is run under the appropriate euid.
7496 58. Following a discussion on the list, the rules by which Exim recognises line
7497 endings on incoming messages have been changed. The -dropcr and drop_cr
7498 options are now no-ops, retained only for backwards compatibility. The
7499 following line terminators are recognized: LF CRLF CR. However, special
7500 processing applies to CR:
7502 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7503 nor a local message in the state where . is a terminator.
7505 (ii) If a bare CR is encountered in a header line, an extra space is added
7506 after the line terminator so as not to end the header. The reasoning
7507 behind this is that bare CRs in header lines are most likely either
7508 to be mistakes, or people trying to play silly games.
7510 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7511 was being incorrectly given as 18 bytes larger than it should have been.
7512 This is a VOB (very old bug).
7514 60. This may never have affected anything current, but just in case it has:
7515 When the local host is found other than at the start of a list of hosts,
7516 the local host, those with the same MX, and any that follow, are discarded.
7517 When the list in question was part of a longer list of hosts, the following
7518 hosts (not currently being processed) were also being discarded. This no
7519 longer happens. I'm not sure if this situation could ever has previously
7522 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7525 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7526 header. This is something that is recommended in a new Internet Draft, and
7527 is something that is documented as being done by Sendmail. There are two
7528 possible values. For messages generated by the autoreply transport, Exim
7531 Auto-Submitted: auto-replied
7533 whereas for all other generated messages (e.g. bounces) it adds
7535 Auto-Submitted: auto-generated
7537 63. The "personal" condition in filters now includes a test for the
7538 Auto-Submitted: header. If it contains the string "auto-" the message it
7539 not considered personal.
7541 64. Added rcpt_include_affixes as a generic transport option.
7543 65. Added queue_only_override (default true).
7545 66. Added the syslog_duplication option.
7547 67. If what should have been the first header line of a message consisted of
7548 a space followed by a colon, Exim was mis-interpreting it as a header line.
7549 It isn't of course - it is syntactically invalid and should therefore be
7550 treated as the start of the message body. The misbehaviour could have
7551 caused a number of strange effects, including loss of data in subsequent
7552 header lines, and spool format errors.
7554 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7555 client host had authenticated. This control can now be exercised by an ACL
7556 for more flexibility.
7558 69. By default, callouts do not happen when testing with -bh. There is now a
7559 variant, -bhc, which does actually run the callout code, including
7560 consulting and updating the callout cache.
7562 70. Added support for saslauthd authentication, courtesy of Alexander
7565 71. If statvfs() failed on the spool or log directories while checking their
7566 size for availability, Exim confusingly gave the error "space shortage".
7567 Furthermore, in debugging mode it crashed with a floating point exception.
7568 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7569 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7570 serious problem, Exim now writes to the main and panic logs when this
7571 happens, with details of the failure. It then refuses to accept the
7572 incoming message, giving the message "spool directory problem" or "log
7573 directory problem" with a 421 code for SMTP messages.
7575 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7576 0, 1, and 2 exist, because some OS complain for execs without them (see
7577 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7578 descriptors. However, the code omitted to check that the open succeeded,
7579 causing mysterious errors if for some reason the permissions on /dev/null
7580 got screwed. Now Exim writes a message to the main and panic logs, and
7581 bombs out if it can't open /dev/null.
7583 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7584 interact so that it is all more flexible. It is supposed to remain
7585 backwards compatible. Also added extra_local_interfaces.
7587 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7588 to bomb out with an assertion failure - to the client this appears as a
7589 connection drop. This problem occurs in the part of the code that was taken
7590 from the Samba project. Fortunately, the assertion is in a very simple
7591 function, so I have fixed this by reproducing the function inline in the
7592 one place where it is called, and arranging for authentication to fail
7593 instead of killing the process with assert().
7595 75. The SPA client code was not working when the server requested OEM rather
7596 than Unicode encoding.
7598 76. Added code to make require_files with a specific uid setting more usable in
7599 the case where statting the file as root fails - usually a non-root-mounted
7600 NFS file system. When this happens and the failure is EACCES, Exim now
7601 forks a subprocess and does the per-uid checking as the relevant uid.
7603 77. Added process_log_path.
7605 78. If log_file_path was not explicitly set, a setting of check_log_space or
7606 check_log_inodes was ignored.
7608 79. If a space check for the spool or log partitions fails, the incident is now
7609 logged. Of course, in the latter case the data may get lost...
7611 80. Added the %p formatting code to string_format() so that it can be used to
7612 print addresses in debug_print(). Adjusted all the address printing in the
7613 debugging in store.c to use %p rather than %d.
7615 81. There was a concern that a line of code in smtp_in.c could overflow a
7616 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7617 initially expressed, the concern was not well-founded, because trailing
7618 spaces are removed early. However, if the trailing spaces were followed by
7619 a NULL, they did not get removed, so the overflow was possible. Two fixes
7622 (a) I re-wrote the offending code in a cleaner fashion.
7623 (b) If an incoming SMTP command contains a NULL character, it is rejected
7626 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7627 runs initgroups(), so that if the Exim user is in any additional groups,
7628 they will be used during message reception.
7634 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob