1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
6 Exim version 4.93+fixes
7 -----------------------
8 This is not an official release. It is just a branch, collecting
9 proposed bugfixes. Depending on your environment the fixes may be
10 necessary to build and/or run Exim successfully.
12 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
13 PAM library frees one of the arguments given to it, despite the
14 documentation. Therefore a plain malloc must be used.
20 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
21 side, to decode a TLS 1.3 packet capture.
23 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
24 Previously the default library behaviour applied, sending two, each in
27 JH/03 Debug output for ACL now gives the config file name and line number for
30 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
32 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
34 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
35 buffer overrun for (non-chunking) other transports.
37 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
38 TLS1.3, means that a server rejecting a client certificate is not visible
39 to the client until the first read of encrypted data (typically the
40 response to EHLO). Add detection for that case and treat it as a failed
41 TLS connection attempt, so that the normal retry-in-clear can work (if
44 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
45 and/or domain. Found and fixed by Jason Betts.
47 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
48 configuration). If a CNAME target was not a wellformed name pattern, a
51 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
52 the OS reports them interleaved with other addresses.
54 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
55 used both for input and for a verify callout, both encrypted, SMTP
56 responses being sent by the server could be lost. This resulted in
57 dropped connections and sometimes bounces generated by a peer sending
60 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
61 string. Previously it was possible to cause undefined behaviour in a
62 library routine (usually a crash). Found by "zerons".
64 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
67 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
68 API was removed, so update to use the newer ones.
70 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
71 any timeout set, is taking a long time. Previously we would hang on to a
72 rotated logfile "forever" if the input was arriving with long gaps
73 (a previous attempt to fix addressed lack, for a long time, of initial
76 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
77 shared (NFS) environment. The length of the tempfile name is now
78 4 + 16 ("hdr.$message_exim_id") which might break on file
79 systems which restrict the file name length to lower values.
82 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
83 shared (NFS) environment.
85 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
86 did for all versions <4.90). Notably -M, -m, --invert, -I may be
89 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
90 on some platforms for bit 31.
92 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
93 to changes apparently associated with TLS1.3 handling some of the APIs
94 previously used were either nonfunctional or inappropriate. Strings
95 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
96 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
97 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
98 This affects log line X= elements, the $tls_{in,out}_cipher variables,
99 and the use of specific cipher names in the encrypted= ACL condition.
101 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
103 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
104 verification result was not updated unless hosts_require_ocsp applied.
106 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
107 queue_list_requires_admin set to false, non-admin users were denied the
110 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
111 directory-of-certs mode. Previously they were advertised despite the
114 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
115 A single TCP connection by a client will now hold a TLS connection open
116 for multiple message deliveries, by default. Previoud the default was to
119 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
120 default. If built with the facility, DANE will be used. The facility
121 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
123 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
124 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
125 must be defined and you must still, unless you define DISABLE_TLS, manage
126 the the include-dir and library-file requirements that go with that
127 choice. Non-TLS builds are still supported.
129 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
132 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
133 default. If the platform supports and has the facility enabled, it will
134 be requested on all coneections.
136 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
137 controlled by the build-time option SUPPORT_PIPE_CONNECT.
139 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
141 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
142 success-DSN messages. Previously the From: header was always the default
143 one for these; the option was ignored.
145 JH/28 Fix the timeout on smtp response to apply to the whole response.
146 Previously it was reset for every read, so a teergrubing peer sending
147 single bytes within the time limit could extend the connection for a
148 long time. Credit to Qualsys Security Advisory Team for the discovery.
150 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
151 delivery address, which leaked information of the results of local
152 forwarding. Change to the original envelope recipient address, per
155 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
156 requested. Previously not bounce was generated and a log entry of
157 error ignored was made.
159 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
161 JH/32 Introduce a general tainting mechanism for values read from the input
162 channel, and values derived from them. Refuse to expand any tainted
163 values, to catch one form of exploit.
165 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
166 was unused and the unexpanded text used for the test. Found and
167 fixed by Ruben Jenster.
169 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
170 an attempt to use a TLS library read routine dereffed a nul pointer,
173 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
174 them in our smtp response.
176 JH/36 Have the general router option retry_use_local_part default to true when
177 any of the restrictive preconditions are set (to anything). Previously it
178 was only for check_local user. The change removes one item of manual
179 configuration which is required for proper retries when a remote router
180 handles a subset of addresses for a domain.
182 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
183 link count into consideration.
185 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
186 caused the extension of big_buffer, the following lines were ignored.
188 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
189 accordance with RFC 2308. Previously there was no expiry, so a longlived
190 receive process (eg. due to ACL delays) versus a short SOA value could
193 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
195 JH/39 Promote DMARC support to mainline.
197 JH/40 Bug 2452: Add a References: header to DSNs.
199 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
200 parameters. The relevant library call is documented as "Deprecated: This
201 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
202 3.6.0, DH parameters are negotiated following RFC7919."
204 HS/06 Change the default of dnssec_request_domains to "*"
206 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
207 carried on and emitted a BDAT command, even when PIPELINING was not
210 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
211 buffer was used for the filename, resulting in a trap when tainted
212 arguments (eg. $domain) were used.
214 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
215 recommended to avoid a possible server-load attack. The feature can be
216 re-enabled via the openssl_options main cofiguration option.
218 JH/45 local_scan API: documented the current smtp_printf() call. This changed
219 for version 4.90 - adding a "more data" boolean to the arguments.
220 Bumped the ABI version number also, this having been missed previously;
221 release versions 4.90 to 4.92.3 inclusive were effectively broken in
222 respect of usage of smtp_printf() by either local_scan code or libraries
223 accessed via the ${dlfunc } expansion item. Both will need coding
224 adjustment for any calls to smtp_printf() to match the new function
225 signature; a FALSE value for the new argument is always safe.
227 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
228 the file-offset (which the Linux syscall does, and exim expects); this
229 resulted in an indefinite loop.
231 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
232 to do ARC verification. The Authentication-Results: header line added
233 by the configuration then had no ARC item.
239 JH/01 Remove code calling the customisable local_scan function, unless a new
240 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
242 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
243 non-signal-safe functions being used.
245 JH/03 Bug 2269: When presented with a received message having a stupidly large
246 number of DKIM-Signature headers, disable DKIM verification to avoid
247 a resource-consumption attack. The limit is set at twenty.
249 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
250 report of oldest_pass in ${authres } in consequence, and separate out
251 some descriptions of reasons for verification fail.
253 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
254 files in the spool were present and unlocked. A queue-runner could spot
255 them, resulting in a duplicate delivery. Fix that by doing the unlock
256 after the unlink. Investigation by Tim Stewart. Take the opportunity to
257 add more error-checking on spoolfile handling while that code is being
260 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
261 No known attacks, no CVE, this is defensive hardening.
263 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
264 a queue-runner could start a delivery while other operations were ongoing.
265 Cutthrough delivery was a common victim, resulting in duplicate delivery.
266 Found and investigated by Tim Stewart. Fix by using the open message data
267 file handle rather than opening another, and not locally closing it (which
268 releases a lock) for that case, while creating the temporary .eml format
269 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
271 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
272 $sender_verify_failure/$recipient_verify_failure to "random".
274 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
277 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
278 Previously this would segfault.
280 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
283 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
284 like zero, since the resolver should be doing this for us, But we need one
285 as a CNAME but no MX presence gets the CNAME returned; we need to check
286 that doesn't point to an MX to declare it "no MX returned" rather than
287 "error, loop". A new main option is added so the older capability of
288 following some limited number of chain links is maintained.
290 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
292 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
293 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
294 now-deprecated earlier definitions used only specified the range up to TLS
295 1.2 (in the older-version library docs).
297 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
299 JH/15 Rework TLS client-side context management. Stop using a global, and
300 explicitly pass a context around. This enables future use of TLS for
301 connections to service-daemons (eg. malware scanning) while a client smtp
302 connection is using TLS; with cutthrough connections this is quite likely.
304 JH/16 Fix ARC verification to do AS checks in reverse order.
306 JH/17 Support a "tls" option on the ${readsocket } expansion item.
308 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
309 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
310 Previously the "utf8" would be re-prepended for every additional message.
312 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
313 Previously thery were accepted, resulting in issues when attempting to
314 forward messages to a non-supporting MTA.
316 PP/02 Let -n work with printing macros too, not just options.
318 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
319 one parent address was copied, and bogus data was used at delivery-logging
320 time. Either a crash (after delivery) or bogus log data could result.
321 Discovery and analysis by Tim Stewart.
323 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
324 Previously if the string ended mid-character, we did not insert the
325 promised '?' replacement.
327 PP/04 Documentation: current string operators work on bytes, not codepoints.
329 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
330 should pack well giving a smaller memory footprint so better caching and
331 therefore performance. Group the declarations where this can't be done so
332 that the byte-sized flag variables are not interspersed among pointer
333 variables, giving a better chance of good packing by the compiler.
335 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
336 non-null, to avoid issues with sites running BATV. Previously reports were
337 sent with an empty envelope sender so looked like bounces.
339 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
340 The ignore_error flag wasn't being returned from the filter subprocess so
341 was not set for later routers. Investigation and fix by Matthias Kurz.
343 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
344 and a msg:complete for the whole, when a message is manually removed using
345 -Mrm. Developement by Matthias Kurz, hacked on by JH.
347 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
348 a "Gnu special" function, asprintf() in the DB utility binary builds; I
349 hope that is portable enough.
351 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
352 requiring a known-CA anchor certificate; make it now rely entirely on the
353 TLSA as an anchor. Checking the name on the leaf cert against the name
354 on the A-record for the host is still done for TA (but not for EE mode).
356 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
357 included in delivery lines for non-proxied connections, when compiled with
358 SUPPORT_SOCKS and running with proxy logging enabled.
360 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
361 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
362 move the existing event to fire before the normal logging of message
363 failure so that custom logging is bracketed by normal logging.
365 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
366 msg:fail:internal event. Developement by Matthias Kurz.
368 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
369 far too small for todays use of crypto signatures stored there. Go all
370 the way to the max DNS message size of 64kB, even though this might be
371 overmuch for IOT constrained device use.
373 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
374 copy a string over itself. The library routine is documented as not
375 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
377 JH/32 For main options check_spool_space and check_inode_space, where the
378 platform supports 64b integers, support more than the previous 2^31 kB
379 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
380 the previous G, M, k.
382 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
383 $authenticated_fail_id variable on authentication failure. Previously
386 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
387 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
388 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
389 GNUTLS_SEC_PARAM_MEDIUM.
391 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
392 side. Previously we would continue as if no SNI had been received.
394 JH/36 Harden the handling of string-lists. When a list consisted of a sole
395 "<" character, which should be a list-separator specification, we walked
396 off past the nul-terimation.
398 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
399 causes) even when the retry time is not yet met. Previously they were
400 not, meaning that when (say) an account was over-quota and temp-rejecting,
401 and multiple senders' messages were queued, only one sender would get
402 notified on each configured delay_warning cycle.
404 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
406 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
407 cipher-suites, an error can be left on the stack even for a succeeding
408 accept; this results in impossible error messages when a later operation
411 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
412 return error codes indicating retry. Under TLS1.3 this becomes required.
414 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
415 it only wrote the new authenticators, resulting in a lack of tracking of
416 peer changes of ESMTP extensions until the next cache flush.
418 JH/41 Fix the loop reading a message header line to check for integer overflow,
419 and more-often against header_maxsize. Previously a crafted message could
420 induce a crash of the recive process; now the message is cleanly rejected.
422 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
423 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
429 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
430 When redis_servers is set to a list of > 1 element, and the Redis servers
431 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
432 case of MOVED into a DEFER case instead, thus moving the query onto the
433 next server in the list. For a cluster of N elements, all N servers must
434 be defined in redis_servers.
436 GF/02 Catch and remove uninitialized value warning in exiqsumm
437 Check for existence of @ARGV before looking at $ARGV[0]
439 JH/01 Replace the store_release() internal interface with store_newblock(),
440 which internalises the check required to safely use the old one, plus
441 the allocate and data copy operations duplicated in both (!) of the
442 extant use locations.
444 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
445 modifier. This matches the restriction on the commandline.
447 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
448 Previously only the last row was returned.
450 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
451 we assumed that tags in the header were well-formed, and parsed the
452 element content after inspecting only the first char of the tag.
453 Assumptions at that stage could crash the receive process on malformed
456 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
457 While running the DKIM ACL we operate on the Permanent memory pool so that
458 variables created with "set" persist to the DATA ACL. Also (at any time)
459 DNS lookups that fail create cache records using the Permanent pool. But
460 expansions release any allocations made on the current pool - so a dnsdb
461 lookup expansion done in the DKIM ACL releases the memory used for the
462 DNS negative-cache, and bad things result. Solution is to switch to the
463 Main pool for expansions.
464 While we're in that code, add checks on the DNS cache during store_reset,
465 active in the testsuite.
466 Problem spotted, and debugging aided, by Wolfgang Breyha.
468 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
469 When none of the hosts presented to a transport match an already-open
470 connection, close it and proceed with the list. Previously we would
471 queue the message. Spotted by Lena with Yahoo, probably involving
474 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
475 Previously a spurious "250 OK id=" response was appended to the proper
478 JH/08 The "support for" informational output now, which built with Content
479 Scanning support, has a line for the malware scanner interfaces compiled
480 in. Interface can be individually included or not at build time.
482 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
483 by the template makefile "src/EDITME". The "STREAM" support for an older
484 ClamAV interface method is removed.
486 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
487 rows affected is given instead).
489 JH/11 The runtime Berkeley DB library version is now additionally output by
490 "exim -d -bV". Previously only the compile-time version was shown.
492 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
493 SMTP connection. Previously, when one had more receipients than the
494 first, an abortive onward connection was made. Move to full support for
495 multiple onward connections in sequence, handling cutthrough connection
496 for all multi-message initiating connections.
498 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
499 routers. Previously, a multi-recipient message would fail to match the
500 onward-connection opened for the first recipient, and cause its closure.
502 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
503 a timeout on read on a GnuTLS initiating connection, resulting in the
504 initiating connection being dropped. This mattered most when the callout
505 was marked defer_ok. Fix to keep the two timeout-detection methods
508 JH/15 Relax results from ACL control request to enable cutthrough, in
509 unsupported situations, from error to silently (except under debug)
510 ignoring. This covers use with PRDR, frozen messages, queue-only and
513 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
515 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
516 metadata, resulting in a crash in free().
518 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
519 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
520 Broken also in d185889f4, with init system revamp.
522 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
523 Previously we abruptly closed the connection after reading a malware-
524 found indication; now we go on to read the "scan ok" response line,
527 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
528 ACL. Previously, a crash would result.
530 JH/19 Speed up macro lookups during configuration file read, by skipping non-
531 macro text after a replacement (previously it was only once per line) and
532 by skipping builtin macros when searching for an uppercase lead character.
534 JH/20 DANE support moved from Experimental to mainline. The Makefile control
535 for the build is renamed.
537 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
538 was allocated for every new TLS startup, meaning one per message. Fix
539 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
541 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
542 reported the original. Fix to report (as far as possible) the ACL
543 result replacing the original.
545 JH/23 Fix memory leak during multi-message connections using STARTTLS under
546 OpenSSL. Certificate information is loaded for every new TLS startup,
547 and the resources needed to be freed.
549 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
551 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
554 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
555 DATA response info to the (existing) per-recipient response info for
556 the "C=" log element. It can have useful tracking info from the
557 destination system. Patch from Simon Arlott.
559 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
560 length value. Previously this would segfault.
562 HS/02 Support Avast multiline protoocol, this allows passing flags to
563 newer versions of the scanner.
565 JH/28 Ensure that variables possibly set during message acceptance are marked
566 dead before release of memory in the daemon loop. This stops complaints
567 about them when the debug_store option is enabled. Discovered specifically
568 for sender_rate_period, but applies to a whole set of variables.
569 Do the same for the queue-runner and queue-list loops, for variables set
570 from spool message files. Do the same for the SMTP per-message loop, for
571 certain variables indirectly set in ACL operations.
573 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
574 as a multi-recipient message from a mailinglist manager). The coding had
575 an arbitrary cutoff number of characters while checking for more input;
576 enforced by writing a NUL into the buffer. This corrupted long / fast
577 input. The problem was exposed more widely when more pipelineing of SMTP
578 responses was introduced, and one Exim system was feeding another.
579 The symptom is log complaints of SMTP syntax error (NUL chars) on the
580 receiving system, and refused recipients seen by the sending system
581 (propating to people being dropped from mailing lists).
582 Discovered and pinpointed by David Carter.
584 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
585 replaced by the ${authresults } expansion.
587 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
589 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
590 allows proper process termination in container environments.
592 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
593 Previously the "final dot" had a newline after it; ensure it is CR,LF.
595 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
596 and "err_perm", deprecated since 4.83 when the RFC-defined words
597 "temperror" and "permerror" were introduced.
599 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
600 transport-filters or DKIM-signing. The restriction was lost in the
601 consolidation of verify-callout and delivery SMTP handling.
602 Extend the restriction to also cover ARC-signing.
604 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
605 in defer=pass mode supply a 450 to the initiator. Previously the message
608 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
609 tls_require_ciphers is used as before.
611 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
612 "pass_unscanned". Only tmpfails from the scanner are written to
613 the paniclog, as they may require admin intervention (permission
614 denied, license issues). Other scanner errors (like decompression
615 bombs) do not cause a paniclog entry.
617 JH/36 Fix reinitialisation of DKIM logging variable between messages.
618 Previously it was possible to log spurious information in receive log
621 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
622 triggered odd behaviour from Outlook Express clients.
624 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
627 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
628 since the IETF WG has not yet settled on that versus the original
629 "bare" representation.
631 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
632 Previously the millisecond value corrupted the output.
633 Fix also for syslog_pid=no and log_selector +pid, for which the pid
634 corrupted the output.
640 JH/01 Rework error string handling in TLS interface so that the caller in
641 more cases is responsible for logging. This permits library-sourced
642 string to be attached to addresses during delivery, and collapses
643 pairs of long lines into single ones.
645 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
646 during configuration. Wildcards are allowed and expanded.
648 JH/02 Rework error string handling in DKIM to pass more info back to callers.
649 This permits better logging.
651 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
652 do not close it down and have the child transport start it up again on
653 the passed-on TCP connection. Instead, proxy the child (and any
654 subsequent ones) for TLS via a unix-domain socket channel. Logging is
655 affected: the continued delivery log lines do not have any DNSSEC, TLS
656 Certificate or OCSP information. TLS cipher information is still logged.
658 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
659 identical IP addresses on different listening ports. Will also affect
662 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
663 add noisy ifdef guards to special-case this sillyness.
664 Patch from Bernd Kuhls.
666 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
667 than 255 are no longer allowed.
669 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
670 Disable the session-cache too, which might reduce our load. Since we
671 currrectly use a new context for every connection, both as server and
672 client, there is no benefit for these.
673 GnuTLS appears to not support tickets server-side by default (we don't
674 call gnutls_session_ticket_enable_server()) but client side is enabled
675 by default on recent versions (3.1.3 +) unless the PFS priority string
678 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
679 <https://reproducible-builds.org/specs/source-date-epoch/>.
681 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
682 the check for any unsuccessful recipients did not notice the limit, and
683 erroneously found still-pending ones.
685 JH/08 Pipeline CHUNKING command and data together, on kernels that support
686 MSG_MORE. Only in-clear (not on TLS connections).
688 JH/09 Avoid using a temporary file during transport using dkim. Unless a
689 transport-filter is involved we can buffer the headers in memory for
690 creating the signature, and read the spool data file once for the
691 signature and again for transmission.
693 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
694 4.77 as the kernel support then wasn't solid, having issues in 64bit
695 mode. Now, it's been long enough. Add support for FreeBSD also.
697 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
698 case where the routing stage had gathered several addresses to send to
699 a host before calling the transport for the first, we previously failed
700 to close down TLS in the old transport process before passing the TCP
701 connection to the new process. The new one sent a STARTTLS command
702 which naturally failed, giving a failed delivery and bloating the retry
703 database. Investigation and fix prototype from Wolfgang Breyha.
705 JH/12 Fix check on SMTP command input synchronisation. Previously there were
706 false-negatives in the check that the sender had not preempted a response
707 or prompt from Exim (running as a server), due to that code's lack of
708 awareness of the SMTP input buffering.
710 PP/04 Add commandline_checks_require_admin option.
711 Exim drops privileges sanely, various checks such as -be aren't a
712 security problem, as long as you trust local users with access to their
713 own account. When invoked by services which pass untrusted data to
714 Exim, this might be an issue. Set this option in main configuration
715 AND make fixes to the calling application, such as using `--` to stop
718 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
719 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
720 responses to those, into a single TLS record each way (this usually means
721 a single packet). As a side issue, smtp_enforce_sync now works on TLS
724 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
725 affects you only if you're dancing at the edge of the param size limits.
726 If you are, and this message makes sense to you, then: raise the
727 configured limit or use OpenSSL 1.1. Nothing we can do for older
730 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
731 cmdline element to get the documented default one. Previously it was
734 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
735 are unsuable for non-message-specific reasons - eg. connection timeout,
736 banner-time rejection.
738 JH/16 Fix logging of delivery remote port, when specified by router, under
741 PP/06 Repair manualroute's ability to take options in any order, even if one
742 is the name of a transport.
745 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
747 JH/17 Change the list-building routines interface to use the expanding-string
748 triplet model, for better allocation and copying behaviour.
750 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
751 Previously it was constructed the first time a possibly-matching string
752 was met in the configuration file input during startup; now it is done
755 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
756 compatible one, to avoid the (poorly documented) possibility of a config
757 file in the working directory redirecting the DB files, possibly correpting
758 some existing file. CVE-2017-10140 assigned for BDB.
760 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
761 cache-hot. Previously, although the result was properly cached, the
762 initial verify call returned a defer.
764 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
765 the main verify for receipient in uncached-mode.
767 JH/22 Retire historical build files to an "unsupported" subdir. These are
768 defined as "ones for which we have no current evidence of testing".
770 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
771 if present. Previously it was ignored.
773 JH/24 Start using specified-initialisers in C structure init coding. This is
774 a C99 feature (it's 2017, so now considered safe).
776 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
777 if was a fixed-sized field and bitmask ops via macros; it is now more
780 PP/07 GitHub PR 56: Apply MariaDB build fix.
781 Patch provided by Jaroslav Škarvada.
783 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
784 during Coverity cleanups [4.87 JH/47]
785 Diagnosis and fix provided by Michael Fischer v. Mollard.
787 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
788 the right size to place the terminating semicolon on its own folded
789 line, the header hash was calculated to an incorrect value thanks to
790 the (relaxed) space the fold became.
792 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
793 and confused the parent.
795 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
796 which could crash as a result. This could lead to undeliverable messages.
798 JH/28 Logging: "next input sent too soon" now shows where input was truncated
801 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
802 matters on fast-turnover and PID-randomising systems, which were getting
803 out-of-order delivery.
805 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
806 a possibly-overlapping copy. The symptom was that "Remote host closed
807 connection in response to HELO" was logged instead of the actual 4xx
810 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
811 Previously only that bufferd was discarded, resulting in SYMTP command
814 JH/32 DKIM: when a message has multiple signatures matching an identity given
815 in dkim_verify_signers, run the dkim acl once for each. Previously only
816 one run was done. Bug 2189.
818 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
819 "panic the current process" to "deliberately defer". The panic log is
820 still written with the problem list name; the mail and reject logs now
821 get a temp-reject line for the message that was being handled, saying
822 something like "domains check lookup or other defer". The SMTP 451
823 message is still "Temporary local problem".
825 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
826 A crafted sequence of BDAT commands could result in in-use memory beeing
827 freed. CVE-2017-16943.
829 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
830 from SMTP input. Previously it was always done; now only done for DATA
831 and not BDAT commands. CVE-2017-16944.
833 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
834 to the message (such as an overlong header line). Previously this was
835 not done and we did not exit BDAT mode. Followon from the previous item
836 though a different problem.
842 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
843 than -2003 did; needs libidn2 in addition to libidn.
845 JH/02 The path option on a pipe transport is now expanded before use.
847 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
848 Patch provided by "Björn", documentation fix added too.
850 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
851 missing a wire-to-host endian conversion.
853 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
854 close after a BDAT command line could be taken as a following command,
855 giving a synch failure. Fix by only checking for synch immediately
856 before acknowledging the chunk.
858 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
859 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
860 macro. Patches provided by Josh Soref.
862 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
863 Previously we did not; the RFC seems ambiguous and VRFY is not listed
864 by IANA as a service extension. However, John Klensin suggests that we
867 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
868 the dkim code may be unix-mode line endings rather than smtp wire-format
869 CRLF, so prepend a CR to any bare LF.
871 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
872 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
874 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
875 the first were themselves being wrongly included in the feed into dkim
876 processing; with most chunk sizes in use this resulted in an incorrect
877 body hash calculated value.
879 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
880 DKIM signature block, for verification. Although advised against by
881 standards it is specifically not ruled illegal.
883 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
885 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
886 missing a body hash (the bh= tag).
888 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
889 It seems that HAProxy sends the Proxy Protocol information in clear and
890 only then does a TLS startup, so do the same.
892 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
893 TCP connections (such as for Spamd) unless the daemon successfully set
894 Fast Open mode on its listening sockets. This fixes breakage seen on
895 too-old kernels or those not configured for Fast Open, at the cost of
896 requiring both directions being enabled for TFO, and TFO never being used
897 by non-daemon-related Exim processes.
899 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
900 endings, at least on the first header line. Try to canonify any that get
901 past that check, despite the cost.
903 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
904 now limited to an arbitrary five deep, while parsing addresses with the
905 strip_excess_angle_brackets option enabled.
907 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
908 instead leave the unprompted TLS handshake in socket buffer for the
909 TLS library to consume.
911 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
913 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
915 JH/16 Drop variables when they go out of scope. Memory management drops a whole
916 region in one operation, for speed, and this leaves assigned pointers
917 dangling. Add checks run only under the testsuite which checks all
918 variables at a store-reset and panics on a dangling pointer; add code
919 explicitly nulling out all the variables discovered. Fixes one known
920 bug: a transport crash, where a dangling pointer for $sending_ip_address
921 originally assigned in a verify callout, is re-used.
923 PP/06 Drop '.' from @INC in various Perl scripts.
925 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
927 PP/08 Reduce a number of compilation warnings under clang; building with
928 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
929 should be warning-free.
931 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
933 HS/01 Fix portability problems introduced by PP/08 for platforms where
934 realloc(NULL) is not equivalent to malloc() [SunOS et al].
936 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
937 chunk. This allows us to accept broken chunked messages. We need a more
938 general solution here.
940 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
941 already-broken messages in the queue.
943 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
945 JH/19 Fix reference counting bug in routing-generated-address tracking.
951 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
952 supports it and a size is available (ie. the sending peer gave us one).
954 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
955 years of being deprecated). The replacements are the ACLs
956 acl_smtp_mime and acl_not_smtp_mime.
958 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
959 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
960 or even an in-clear connection were permitted. Now, if the host lookup
961 was dnssec and dane was requested then the host is only used if the
962 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
963 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
964 if one fails this test.
965 This means that a poorly-configured remote DNS will make it incommunicado;
966 but it protects against a DNS-interception attack on it.
968 JH/04 Bug 1810: make continued-use of an open smtp transport connection
969 non-noisy when a race steals the message being considered.
971 JH/05 If main configuration option tls_certificate is unset, generate a
972 self-signed certificate for inbound TLS connections.
974 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
975 in rewrites and routers.
977 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
978 and logged a warning sing 4.83; now they are a configuration file error.
980 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
981 (lacking @domain). Apply the same qualification processing as RCPT.
983 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
985 JH/10 Support ${sha256:} applied to a string (as well as the previous
988 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
989 a cutthrough deliver is pending, as we always want to make a connection.
990 This also avoids re-routing the message when later placing the cutthrough
991 connection after a verify cache hit.
992 Do not update it with the verify result either.
994 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
995 when routing results in more than one destination address.
997 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
998 signing (which inhibits the cutthrough capability). Previously only
999 the presence of an option was tested; now an expansion evaluating as
1000 empty is permissible (obviously it should depend only on data available
1001 when the cutthrough connection is made).
1003 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1004 the relevant preceding SMTP command did not note the pipelining mode.
1006 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1007 Previously they were not counted.
1009 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1010 as one having no matching records. Previously we deferred the message
1011 that needed the lookup.
1013 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1014 distinguished as "(=".
1016 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1017 for missing MX records. Previously it only worked for missing A records.
1019 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1021 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1022 after the data-go-ahead and data-ack. Patch from Jason Betts.
1024 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1025 even for a "none" policy. Patch from Tony Meyer.
1027 JH/22 Fix continued use of a connection for further deliveries. If a port was
1028 specified by a router, it must also match for the delivery to be
1031 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1032 When one of the recipients of a message was unsuitable for the connection
1033 (has no matching addresses), we lost track of needing to mark it
1034 deferred. As a result mail would be lost.
1036 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1038 JH/25 Decoding ACL controls is now done using a binary search; the source code
1039 takes up less space and should be simpler to maintain. Merge the ACL
1040 condition decode tables also, with similar effect.
1042 JH/26 Fix problem with one_time used on a redirect router which returned the
1043 parent address unchanged. A retry would see the parent address marked as
1044 delivered, so not attempt the (identical) child. As a result mail would
1047 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1048 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1049 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1052 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1053 Main config options check_{log,spool}_{inodes,space} are now
1054 100 inodes, 10MB unless set otherwise in the configuration.
1056 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1057 Previously it only applied to the main-section connection policy
1060 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1062 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1063 by me. Added RFC7919 DH primes as an alternative.
1065 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1066 are not in the system include path.
1068 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1069 GnuTLS, when a session startup failed (eg because the client disconnected)
1070 Exim did stdio operations after fclose. This was exposed by a recent
1071 change which nulled out the file handle after the fclose.
1073 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1074 signed directly by the cert-signing cert, rather than an intermediate
1075 OCSP-signing cert. This is the model used by LetsEncrypt.
1077 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1079 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1080 an incoming connection.
1082 HS/02 Bug 1802: Do not half-close the connection after sending a request
1085 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1086 fallback to "prime256v1".
1088 JH/34 SECURITY: Use proper copy of DATA command in error message.
1089 Could leak key material. Remotely exploitable. CVE-2016-9963.
1095 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1096 and 3.4.4 - once the server is enabled to respond to an OCSP request
1097 it does even when not requested, resulting in a stapling non-aware
1098 client dropping the TLS connection.
1100 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1101 support variable-length bit vectors. No functional change.
1103 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1104 The I= interface field on outgoing lines is now after the H= remote
1105 host field, same as incoming lines. There is a separate
1106 outgoing_interface log selector which allows you to disable the
1109 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1110 If not running log_selector +smtp_connection the mainlog would be held
1111 open indefinitely after a "too many connections" event, including to a
1112 deleted file after a log rotate. Leave the per net connection logging
1113 leaving it open for efficiency as that will be quickly detected by the
1114 check on the next write.
1116 HS/01 Bug 1671: Fix post transport crash.
1117 Processing the wait-<transport> messages could crash the delivery
1118 process if the message IDs didn't exist for some reason. When
1119 using 'split_spool_directory=yes' the construction of the spool
1120 file name failed already, exposing the same netto behaviour.
1122 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1123 mime_regex ACL conditions.
1125 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1126 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1127 to HELO, local diagnostic string.
1129 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1130 log line to debug. Even when configured with a tls authenticator many
1131 client connections are expected to not authenticate in this way, so
1132 an authenticate fail is not an error.
1134 HS/02 Add the Exim version string to the process info. This way exiwhat
1135 gives some more detail about the running daemon.
1137 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1138 matter for fast-change records such as DNSBLs.
1140 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1141 retry record, even if constant. There may be multiple transports with
1142 different interface settings and the retry behaviour needs to be kept
1145 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1147 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1149 JH/10 Bug 840: fix log_defer_output option of pipe transport
1151 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1152 hosts_randomize. This matters a lot when combined with mua_wrapper.
1154 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1155 ${quote_pgsql:<string>} operator.
1157 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1159 JH/14 Change header file name for freeradius-client. Relevant if compiling
1160 with Radius support; from the Gentoo tree and checked under Fedora.
1162 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1164 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1165 option transport_filter, meaning no filtering.
1167 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1169 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1170 defaults to "*" (all hosts). The variable is now available when not built
1171 with TLS, default unset, mainly to enable keeping the testsuite sane.
1172 If a server certificate is not supplied (via tls_certificate) an error is
1173 logged, and clients will find TLS connections fail on startup. Presumably
1174 they will retry in-clear.
1175 Packagers of Exim are strongly encouraged to create a server certificate
1176 at installation time.
1178 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1179 with the $config_file variable.
1181 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1182 in transport context, after the attempt, and per-recipient. The latter type
1183 is per host attempted. The event data is the error message, and the errno
1184 information encodes the lookup type (A vs. MX) used for the (first) host,
1185 and the trailing two digits of the smtp 4xx response.
1187 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1188 to write to mainlog (or rejectlog, paniclog) in the window between file
1189 creation and permissions/ownership being changed. Particularly affects
1190 installations where exicyclog is run as root, rather than exim user;
1191 result is that the running daemon panics and dies.
1193 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1195 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1196 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1197 "pri" and "weight". Note that the previous implicit priority given by the
1198 list order is no longer honoured.
1200 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1201 for DKIM processing.
1203 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1204 by defining SUPPORT_SOCKS.
1206 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1207 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1208 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1209 variables are renamed to proxy_{local,external}_{address,port}.
1211 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1212 for a build by defining SUPPORT_I18N
1214 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1215 of the query string, and make ${quote_redis:} do that quoting.
1217 JH/29 Move Events support from Experimental to mainline, enabled by default
1218 and removable for a build by defining DISABLE_EVENT.
1220 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1222 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1223 cached by the daemon.
1225 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1226 by defining LOOKUP_REDIS. The libhiredis library is required.
1228 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1229 keys are given for lookup.
1231 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1232 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1233 only supported when built with TLS support. The PolarSSL SHA routines
1234 are still used when the TLS library is too old for convenient support.
1236 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1237 openssl_options), for security. OpenSSL forces this from version 1.1.0
1238 server-side so match that on older versions.
1240 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1241 allocation for $value could be released as the expansion processing
1242 concluded, but leaving the global pointer active for it.
1244 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1245 and to use the domains and local_parts ACL conditions.
1247 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1248 incorrectly not doubled on cutthrough transmission, hence seen as a
1249 body-termination at the receiving system - resulting in truncated mails.
1250 Commonly the sender saw a TCP-level error, and retransmitted the message
1251 via the normal store-and-forward channel. This could result in duplicates
1252 received - but deduplicating mailstores were liable to retain only the
1253 initial truncated version.
1255 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1257 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1259 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1260 we're in there, support oversigning also; bug 1309.
1262 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1264 HS/04 Add support for keep_environment and add_environment options.
1266 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1267 either intentional arithmetic overflow during PRNG, or testing config-
1270 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1271 delivery resulted in actual delivery. Cancel cutthrough before DATA
1274 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1275 rejects a recipient: pass the reject to the originator.
1277 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1278 Many were false-positives and ignorable, but it's worth fixing the
1281 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1282 for the new environment-manipulation done at startup. Move the routines
1283 from being local to tls.c to being global via the os.c file.
1285 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1286 an extract embedded as result-arg for a map, the first arg for extract
1287 is unavailable so we cannot tell if this is a numbered or keyed
1288 extraction. Accept either.
1294 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1297 JH/02 The smtp transport option "multi_domain" is now expanded.
1299 JH/03 The smtp transport now requests PRDR by default, if the server offers
1302 JH/04 Certificate name checking on server certificates, when exim is a client,
1303 is now done by default. The transport option tls_verify_cert_hostnames
1304 can be used to disable this per-host. The build option
1305 EXPERIMENTAL_CERTNAMES is withdrawn.
1307 JH/05 The value of the tls_verify_certificates smtp transport and main options
1308 default to the word "system" to access the system default CA bundle.
1309 For GnuTLS, only version 3.0.20 or later.
1311 JH/06 Verification of the server certificate for a TLS connection is now tried
1312 (but not required) by default. The verification status is now logged by
1313 default, for both outbound TLS and client-certificate supplying inbound
1316 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1319 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1320 Status Notification (bounce) messages are now MIME format per RFC 3464.
1321 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1322 under the control of the dsn_advertise_hosts option, and routers may
1323 have a dsn_lasthop option.
1325 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1326 default, modifiable by a malware= option. The list separator for
1327 the options can now be changed in the usual way. Bug 68.
1329 JH/10 The smtp_receive_timeout main option is now expanded before use.
1331 JH/11 The incoming_interface log option now also enables logging of the
1332 local interface on delivery outgoing connections.
1334 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1335 if the interface and destination host and port all match.
1337 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1340 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1341 Patch from Andrew Lewis.
1343 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1344 now supports optional time-restrictions, weighting, and priority
1345 modifiers per server. Patch originally by <rommer@active.by>.
1347 JH/16 The spamd_address main option now supports a mixed list of local
1348 and remote servers. Remote servers can be IPv6 addresses, and
1349 specify a port-range.
1351 JH/17 Bug 68: The spamd_address main option now supports an optional
1352 timeout value per server.
1354 JH/18 Bug 1581: Router and transport options headers_add/remove can
1355 now have the list separator specified.
1357 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1360 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1363 JH/21 Support for the A6 type of dns record is withdrawn.
1365 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1366 rather than the verbs used.
1368 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1369 from 255 to 1024 chars.
1371 JH/24 Verification callouts now attempt to use TLS by default.
1373 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1374 are generic router options now. The defaults didn't change.
1376 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1377 Original patch from Alexander Shikoff, worked over by JH.
1379 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1380 files if ZCAT_COMMAND is not executable.
1382 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1384 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1386 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1387 Normally benign, it bites when the pair was led to by a CNAME;
1388 modern usage is to not canonicalize the domain to a CNAME target
1389 (and we were inconsistent anyway for A-only vs AAAA+A).
1391 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1393 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1394 when evaluating $sender_host_dnssec.
1396 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1397 $sender_helo_dnssec variable.
1399 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1401 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1403 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1405 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1406 documented as working, but never had. Support all but $spam_report.
1408 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1409 added for tls authenticator.
1411 HS/03 Add perl_taintmode main config option
1417 TL/01 When running the test suite, the README says that variables such as
1418 no_msglog_check are global and can be placed anywhere in a specific
1419 test's script, however it was observed that placement needed to be near
1420 the beginning for it to behave that way. Changed the runtest perl
1421 script to read through the entire script once to detect and set these
1422 variables, reset to the beginning of the script, and then run through
1423 the script parsing/test process like normal.
1425 TL/02 The BSD's have an arc4random API. One of the functions to induce
1426 adding randomness was arc4random_stir(), but it has been removed in
1427 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1428 function when detected.
1430 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1431 cause callback expansion.
1433 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1434 syntax errors in an expansion can be treated as a string instead of
1435 logging or causing an error, due to the internal use of bool_lax
1436 instead of bool when processing it.
1438 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1439 server certificates when making smtp deliveries.
1441 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1443 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1445 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1447 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1448 Merged patch from Sebastian Wiedenroth.
1450 JH/05 Fix results-pipe from transport process. Several recipients, combined
1451 with certificate use, exposed issues where response data items split
1452 over buffer boundaries were not parsed properly. This eventually
1453 resulted in duplicates being sent. This issue only became common enough
1454 to notice due to the introduction of connection certificate information,
1455 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1457 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1458 size buffer was used, resulting in syntax errors when an expansion
1461 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1462 version 3.3.6 or later.
1464 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1465 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1466 both become "event_action", the variables become $event_name, $event_data
1467 and $event_defer_errno. There is a new variable $verify_mode, usable in
1468 routers, transports and related events. The tls:cert event is now also
1469 raised for inbound connections, if the main configuration event_action
1472 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1473 early OCSP support, but no stapling (appears to be less than 1.0.0).
1475 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1476 server certificate names available under the smtp transport option
1477 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1480 JH/10 Time-related extraction expansions from certificates now use the main
1481 option "timezone" setting for output formatting, and are consistent
1482 between OpenSSL and GnuTLS compilations. Bug 1541.
1484 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1485 encoded parameter in the incoming message. Bug 1558.
1487 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1488 include certificate info, eximon was claiming there were spoolfile
1491 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1493 JH/14 Log delivery-related information more consistently, using the sequence
1494 "H=<name> [<ip>]" wherever possible.
1496 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1497 are problematic for Debian distribution, omit them from the release
1500 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1502 JH/16 Fix string representation of time values on 64bit time_t architectures.
1505 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1506 output list separator was used.
1511 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1512 checkers that were complaining about end of non-void function with no
1515 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1516 This was a regression introduced in 4.83 by another bugfix.
1518 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1520 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1521 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1527 TF/01 Correctly close the server side of TLS when forking for delivery.
1529 When a message was received over SMTP with TLS, Exim failed to clear up
1530 the incoming connection properly after forking off the child process to
1531 deliver the message. In some situations the subsequent outgoing
1532 delivery connection happened to have the same fd number as the incoming
1533 connection previously had. Exim would try to use TLS and fail, logging
1534 a "Bad file descriptor" error.
1536 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1537 utilities have not been installed.
1539 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1540 temporary space as the ACL may create new global variables.
1542 TL/01 LDAP support uses per connection or global context settings, depending
1543 upon the detected version of the libraries at build time.
1545 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1546 to extract and use the src ip:port in logging and expansions as if it
1547 were a direct connection from the outside internet. PPv2 support was
1548 updated based on HAProxy spec change in May 2014.
1550 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1552 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1553 Properly escape header and check for NULL return.
1555 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1558 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1560 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1561 characters in header names, implemented as a verify condition.
1562 Contributed by Michael Fischer v. Mollard.
1564 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1565 results permerror and temperror. Previous values are deprecated but
1566 still accepted. In a future release, err_perm and err_temp will be
1567 completely removed, which will be a backward incompatibility if the
1568 ACL tests for either of these two old results. Patch contributed by
1569 user bes-internal on the mailing list.
1571 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1573 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1574 selectors, in both main and reject logs.
1576 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1579 JH/07 Add malware type "sock" for talking to simple daemon.
1581 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1583 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1584 routers/transports under cutthrough routing.
1586 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1587 numbers. Touch up "bool" conditional to keep the same definition.
1589 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1591 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1593 JH/12 Expand items in router/transport headers_add or headers_remove lists
1594 individually rather than the list as a whole. Bug 1452.
1596 Required for reasonable handling of multiple headers_ options when
1597 they may be empty; requires that headers_remove items with embedded
1598 colons must have them doubled (or the list-separator changed).
1600 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1601 view the policy declared in the DMARC record. Currently, $dmarc_status
1602 is a combined value of both the record presence and the result of the
1605 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1607 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1608 dnslookup router and the smtp transport (applying to the forward
1611 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1612 of ldap servers used for a specific lookup. Patch provided by Heiko
1615 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1616 New variable $lookup_dnssec_authenticated for observability.
1618 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1619 Patch submitted by Lars Timman.
1621 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1623 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1624 Requires trusted mode and valid format message id, aborts otherwise.
1625 Patch contributed by Heiko Schlichting.
1627 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1628 certextract with support for various fields. Bug 1358.
1630 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1631 is requested by default, modifiable by smtp transport option
1634 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1635 operate on certificate variables to give certificate fingerprints
1636 Also new ${sha256:cert_variable}.
1638 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1640 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1641 Christian Aistleitner.
1643 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1645 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1646 file. Patch from Wolfgang Breyha.
1648 JH/25 Expand the coverage of the delivery $host and $host_address to
1649 client authenticators run in verify callout. Bug 1476.
1651 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1652 align with daemon_smtp_ports. Bug 72.
1654 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1655 support and error reporting did not work properly.
1657 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1658 and is readable. Patch from Andrew Colin Kissa.
1660 TL/14 Enhance documentation of ${run expansion and how it parses the
1661 commandline after expansion, particularly in the case when an
1662 unquoted variable expansion results in an empty value.
1664 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1666 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1667 by renaming away from it. Addresses GH issue 15, reported by
1670 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1672 TL/15 SECURITY: prevent double expansion in math comparison functions
1673 (can expand unsanitized data). Not remotely exploitable.
1680 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1682 PP/02 Make -n do something, by making it not do something.
1683 When combined with -bP, the name of an option is not output.
1685 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1688 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1689 $sender_host_name and config options to manage this, and basic check
1692 PP/05 DSCP support for outbound connections and control modifier for inbound.
1694 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1695 (Only plugin which currently uses this is kerberos4, which nobody should
1696 be using, but we should make it available and other future plugins might
1697 conceivably use it, even though it would break NAT; stuff *should* be
1698 using channel bindings instead).
1700 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1701 name; added for Sendmail compatibility; requires admin caller.
1702 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1703 just ignore it); requires trusted caller.
1704 Also parse but ignore: -Ac -Am -X<logfile>
1707 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1709 TL/02 Add +smtp_confirmation as a default logging option.
1711 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1712 Patch by Magnus Holmgren from 2007-02-20.
1714 TL/04 Bugzilla 1281 - Spec typo.
1715 Bugzilla 1283 - Spec typo.
1716 Bugzilla 1290 - Spec grammar fixes.
1718 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1720 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1722 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1723 mailing list by by Dmitry Isaikin.
1725 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1727 JH/02 Support "G" suffix to numbers in ${if comparisons.
1729 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1731 NM/01 Bugzilla 1197 - Spec typo
1732 Bugzilla 1196 - Spec examples corrections
1734 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1736 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1737 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1740 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1741 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1743 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1744 "acl {{name}{arg}...}", and optional args on acl condition
1747 JH/05 Permit multiple router/transport headers_add/remove lines.
1749 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1751 JH/07 Avoid using a waiting database for a single-message-only transport.
1752 Performance patch from Paul Fisher. Bugzilla 1262.
1754 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1757 JH/09 Add $headers_added variable, with content from use of ACL modifier
1758 add_header (but not yet added to the message). Bugzilla 199.
1760 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1761 Pulled from Bugzilla 817 by Wolfgang Breyha.
1763 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1765 (nb: this is the same fix as in Exim 4.80.1)
1767 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1770 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1773 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1775 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1776 advertise SMTP AUTH mechanism to us, instead of a generic
1777 protocol violation error. Also, make Exim more robust to bad
1778 data from the Dovecot auth socket.
1780 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1782 When a queue runner is handling a message, Exim first routes the
1783 recipient addresses, during which it prunes them based on the retry
1784 hints database. After that it attempts to deliver the message to
1785 any remaining recipients. It then updates the hints database using
1788 So if a recipient address works intermittently, it can get repeatedly
1789 deferred at routing time. The retry hints record remains fresh so the
1790 address never reaches the final cutoff time.
1792 This is a fairly common occurrence when a user is bumping up against
1793 their storage quota. Exim had some logic in its local delivery code
1794 to deal with this. However it did not apply to per-recipient defers
1795 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1797 This change adds a proper retry rule check during routing so that the
1798 final cutoff time is checked against the message's age. We only do
1799 this check if there is an address retry record and there is not a
1800 domain retry record; this implies that previous attempts to handle
1801 the address had the retry_use_local_parts option turned on. We use
1802 this as an approximation for the destination being like a local
1803 delivery, as in LMTP.
1805 I suspect this new check makes the old local delivery cutoff check
1806 redundant, but I have not verified this so I left the code in place.
1808 TF/02 Correct gecos expansion when From: is a prefix of the username.
1810 Test 0254 submits a message to Exim with the header
1814 When I ran the test suite under the user fanf2, Exim expanded
1815 the header to contain my full name, whereas it should have added
1816 a Resent-Sender: header. It erroneously treats any prefix of the
1817 username as equal to the username.
1819 This change corrects that bug.
1821 GF/01 DCC debug and logging tidyup
1822 Error conditions log to paniclog rather than rejectlog.
1823 Debug lines prefixed by "DCC: " to remove any ambiguity.
1825 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1827 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1828 Bug spotted by Jeremy Harris; was flawed since initial commit.
1829 Would have resulted in OCSP responses post-SNI triggering an Exim
1830 NULL dereference and crash.
1832 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1834 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1835 Bug detection, analysis and fix by Samuel Thibault.
1836 Bugzilla 1331, Debian bug #698092.
1838 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1840 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1841 Server implementation by Todd Lyons, client by JH.
1842 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1843 config variable "prdr_enable" controls whether the server
1844 advertises the facility. If the client requests PRDR a new
1845 acl_data_smtp_prdr ACL is called once for each recipient, after
1846 the body content is received and before the acl_smtp_data ACL.
1847 The client is controlled by both of: a hosts_try_prdr option
1848 on the smtp transport, and the server advertisement.
1849 Default client logging of deliveries and rejections involving
1850 PRDR are flagged with the string "PRDR".
1852 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1853 fclose(). Diagnosis by Todd Lyons.
1855 PP/17 Update configure.default to handle IPv6 localhost better.
1856 Patch by Alain Williams (plus minor tweaks).
1859 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1860 This is now consistent with GnuTLS, and is now documented: the
1861 previous undocumented portable approach to treating the option as
1862 unset was to force an expansion failure. That still works, and
1863 an empty string is now equivalent.
1865 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1866 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1867 not performing validation itself.
1869 PP/20 Added force_command boolean option to pipe transport.
1870 Patch from Nick Koston, of cPanel Inc.
1872 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1875 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1877 PP/21 Fix eximon continuous updating with timestamped log-files.
1878 Broken in a format-string cleanup in 4.80, missed when I repaired the
1879 other false fix of the same issue.
1880 Report and fix from Heiko Schlichting.
1883 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1884 Report from Prashanth Katuri.
1886 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1887 It's SecureTransport, so affects any MacOS clients which use the
1888 system-integrated TLS libraries, including email clients.
1890 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1891 using a MIME ACL for non-SMTP local injection.
1892 Report and assistance in diagnosis by Warren Baker.
1894 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1896 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1898 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1899 last id that failed so it may be referenced in subsequent ACL's.
1901 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1904 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1905 ldap library initialization, allowing self-signed CA's to be
1906 used. Also properly sets require_cert option later in code by
1907 using NULL (global ldap config) instead of ldap handle (per
1908 session). Bug diagnosis and testing by alxgomz.
1910 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1911 the src/util/ subdirectory.
1913 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1914 renamed to Transport Post Delivery Action by Jeremy Harris, as
1917 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1918 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1919 redis_servers = needs to be configured which will be used by the redis
1920 lookup. Patch from Warren Baker, of The Packet Hub.
1922 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1924 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1925 hostname or reverse DNS when processing a host list. Used suggestions
1926 from multiple comments on this bug.
1928 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1930 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1931 Missed a few lines, added it to make the runtest require no keyboard
1934 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1935 contains upper case chars. Make router use caseful_local_part.
1937 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1938 support when GnuTLS has been built with p11-kit.
1944 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1946 This, or similar/improved, will also be change PP/11 of 4.82.
1952 PP/01 Handle short writes when writing local log-files.
1953 In practice, only affects FreeBSD (8 onwards).
1954 Bugzilla 1053, with thanks to Dmitry Isaikin.
1956 NM/01 Bugzilla 949 - Documentation tweak
1958 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1961 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1963 PP/02 Implemented gsasl authenticator.
1965 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1967 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1968 `pkg-config foo` for cflags/libs.
1970 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1971 with rest of GSASL and with heimdal_gssapi.
1973 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1974 `pkg-config foo` for cflags/libs for the TLS implementation.
1976 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1977 properties get this fed in as external SSF. A number of robustness
1978 and debugging improvements to the cyrus_sasl authenticator.
1980 PP/08 cyrus_sasl server now expands the server_realm option.
1982 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1983 Patch by Jeremy Harris.
1985 PP/10 Added dbmjz lookup type.
1987 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1989 PP/12 MAIL args handles TAB as well as SP, for better interop with
1990 non-compliant senders.
1991 Analysis and variant patch by Todd Lyons.
1993 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1994 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1995 Patch from Dirk Mueller <dmueller@suse.com>
1997 PP/13 tls_peerdn now print-escaped for spool files.
1998 Observed some $tls_peerdn in wild which contained \n, which resulted
1999 in spool file corruption.
2001 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2002 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2003 or write after TLS renegotiation, which otherwise led to messages
2006 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2007 as a tracking header (ie: a signed header comes before the signature).
2008 Patch from Wolfgang Breyha.
2010 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2011 comma-sep list; embedded commas doubled.
2013 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2015 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2017 Report and patch from Dmitry Banschikov.
2019 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2020 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2021 failures. We appear to now support TLS1.1+ with Exim.
2023 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2024 lets Exim select keys and certificates based upon TLS SNI from client.
2025 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2026 before an outbound SMTP session. New log_selector, +tls_sni.
2028 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2029 NULL dereference. Report and patch from Alun Jones.
2031 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2032 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2033 Not seeing resolver debug output on NetBSD, but suspect this is a
2034 resolver implementation change.
2036 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2037 Left warnings. Added "eximon gdb" invocation mode.
2039 PP/21 Defaulting "accept_8bitmime" to true, not false.
2041 PP/22 Added -bw for inetd wait mode support.
2043 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2044 locate the relevant includes and libraries. Made this the default.
2046 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2047 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2049 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2050 This may cause build issues on older platforms.
2052 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2053 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2054 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2055 Added SNI support via GnuTLS too.
2056 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2058 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2060 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2061 Applied second patch from Janne, implementing suggestion to default
2062 multiple-strings-in-record handling to match SPF spec.
2064 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2066 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2067 read-only, out of scope).
2068 Patch from Wolfgang Breyha, report from Stuart Northfield.
2070 PP/29 Fix three issues highlighted by clang analyser static analysis.
2071 Only crash-plausible issue would require the Cambridge-specific
2072 iplookup router and a misconfiguration.
2073 Report from Marcin Mirosław.
2075 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2077 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2078 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2079 As part of this, removing so much warning spew let me fix some minor
2080 real issues in debug logging.
2082 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2083 assignment on my part. Fixed.
2085 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2086 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2087 Janne Snabb (who went above and beyond: thank you).
2089 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2090 string otherwise requires a connection and a bunch more work and it's
2091 relatively easy to get wrong. Should also expose TLS library linkage
2094 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2095 64-bit ${eval} (JH/03).
2097 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2098 GNU libc to support some of the 64-bit stuff, should not lead to
2099 conflicts. Defined before os.h is pulled in, so if a given platform
2100 needs to override this, it can.
2102 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2103 protection layer was required, which is not implemented.
2104 Bugzilla 1254, patch from Wolfgang Breyha.
2106 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2107 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2108 tls_dhparam take prime identifiers. Also unbreak combination of
2109 OpenSSL+DH_params+TLSSNI.
2111 PP/39 Disable SSLv2 by default in OpenSSL support.
2117 PP/01 Solaris build fix for Oracle's LDAP libraries.
2118 Bugzilla 1109, patch from Stephen Usher.
2120 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2122 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2125 TF/02 Fix a couple more cases where we did not log the error message
2126 when unlink() failed. See also change 4.74-TF/03.
2128 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2129 lock up or crash if it happened to be inside a call to libc when it
2130 got a SIGUSR1 from exiwhat.
2132 The SIGUSR1 handler appends the current process status to the process
2133 log which is later printed by exiwhat. It used to use the general
2134 purpose logging code to do this, but several functions it calls are
2135 not safe for signals.
2137 The new output code in the SIGUSR1 handler is specific to the process
2138 log, and simple enough that it's easy to inspect for signal safety.
2139 Removing some special cases also simplifies the general logging code.
2140 Removing the spurious timestamps from the process log simplifies
2143 TF/04 Improved ratelimit ACL condition.
2145 The /noupdate option has been deprecated in favour of /readonly which
2146 has clearer semantics. The /leaky, /strict, and /readonly update modes
2147 are mutually exclusive. The update mode is no longer included in the
2148 database key; it just determines when the database is updated. (This
2149 means that when you upgrade Exim will forget old rate measurements.)
2151 Exim now checks that the per_* options are used with an update mode that
2152 makes sense for the current ACL. For example, when Exim is processing a
2153 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2154 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2155 must specify per_mail/readonly. If you omit the update mode it defaults to
2156 /leaky where that makes sense (as before) or /readonly where required.
2158 The /noupdate option is now undocumented but still supported for
2159 backwards compatibility. It is equivalent to /readonly except that in
2160 ACLs where /readonly is required you may specify /leaky/noupdate or
2161 /strict/noupdate which are treated the same as /readonly.
2163 A useful new feature is the /count= option. This is a generalization
2164 of the per_byte option, so that you can measure the throughput of other
2165 aggregate values. For example, the per_byte option is now equivalent
2166 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2168 The per_rcpt option has been generalized using the /count= mechanism
2169 (though it's more complicated than the per_byte equivalence). When it is
2170 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2171 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2172 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2173 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2174 Note that using per_rcpt with a non-readonly update mode in more than
2175 one ACL will cause the recipients to be double-counted. (The per_mail
2176 and per_byte options don't have this problem.)
2178 The handling of very low rates has changed slightly. If the computed rate
2179 is less than the event's count (usually one) then this event is the first
2180 after a long gap. In this case the rate is set to the same as this event's
2181 count, so that the first message of a spam run is counted properly.
2183 The major new feature is a mechanism for counting the rate of unique
2184 events. The new per_addr option counts the number of different
2185 recipients that someone has sent messages to in the last time period. It
2186 behaves like per_rcpt if all the recipient addresses are different, but
2187 duplicate recipient addresses do not increase the measured rate. Like
2188 the /count= option this is a general mechanism, so the per_addr option
2189 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2190 example, measure the rate that a client uses different sender addresses
2191 with the options per_mail/unique=$sender_address. There are further
2192 details in the main documentation.
2194 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2196 TF/06 Removed a few PCRE remnants.
2198 TF/07 Automatically extract Exim's version number from tags in the git
2199 repository when doing development or release builds.
2201 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2202 Bugzilla 879. Patch from Paul Fisher.
2204 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2205 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2208 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2210 PP/05 Variable $av_failed, true if the AV scanner deferred.
2211 Bugzilla 1078. Patch from John Horne.
2213 PP/06 Stop make process more reliably on build failure.
2214 Bugzilla 1087. Patch from Heiko Schlittermann.
2216 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2217 Bugzilla 1089. Patch from Heiko Schlittermann.
2219 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2220 Bugzilla 1131. Patch from Holger Weiß.
2222 PP/09 Handle IPv6 addresses with SPF.
2223 Bugzilla 860. Patch from Wolfgang Breyha.
2225 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2227 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2230 PP/11 match_* no longer expand right-hand-side by default.
2231 New compile-time build option, EXPAND_LISTMATCH_RHS.
2232 New expansion conditions, "inlist", "inlisti".
2234 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2236 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2238 PP/14 fix log_write() format string regression from TF/03.
2239 Bugzilla 1152. Patch from Dmitry Isaikin.
2245 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2247 PP/02 Harmonised TLS library version reporting; only show if debugging.
2248 Layout now matches that introduced for other libraries in 4.74 PP/03.
2250 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2252 PP/04 New "dns_use_edns0" global option.
2254 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2257 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2258 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2260 TK/01 Updated PolarSSL code to 0.14.2.
2261 Bugzilla 1097. Patch from Andreas Metzler.
2263 PP/07 Catch divide-by-zero in ${eval:...}.
2264 Fixes bugzilla 1102.
2266 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2269 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2270 format-string attack -- SECURITY: remote arbitrary code execution.
2272 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2273 time unintentionally subject to list matching rules, letting the header
2274 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2275 arbitrary string expansion). This allowed for information disclosure.
2277 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2278 INT_MIN/-1 -- value coerced to INT_MAX.
2284 NM/01 Workaround for PCRE version dependency in version reporting
2287 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2288 This fixes portability to compilers other than gcc, notably
2289 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2291 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2292 makefiles for portability to HP-UX and POSIX correctness.
2294 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2295 Also via indented variable definition in the Makefile.
2296 (Debugging by Oliver Heesakkers).
2298 PP/02 Restore caching of spamd results with expanded spamd_address.
2299 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2301 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2302 Improves build reliability. Fix from: Frank Elsner
2304 NM/02 Fix wide character breakage in the rfc2047 coding
2305 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2307 NM/03 Allow underscore in dnslist lookups
2308 Fixes bug 1026. Patch from Graeme Fowler
2310 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2311 Code patches from Adam Ciarcinski of NetBSD.
2313 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2316 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2317 is logged, to avoid truncation. Patch from John Horne.
2319 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2320 Patch from Jakob Hirsch.
2322 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2323 SQL string expansion failure details.
2324 Patch from Andrey Oktyabrski.
2326 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2327 Patch from Simon Arlott.
2329 PP/09 New lookups functionality failed to compile on old gcc which rejects
2330 extern declarations in function scope.
2331 Patch from Oliver Fleischmann
2333 PP/10 Use sig_atomic_t for flags set from signal handlers.
2334 Check getgroups() return and improve debugging.
2335 Fixed developed for diagnosis in bug 927 (which turned out to be
2338 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2339 Patch from Mark Zealey.
2341 PP/12 Bugzilla 1056: Improved spamd server selection.
2342 Patch from Mark Zealey.
2344 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2345 Based on patch from Heiko Schlittermann.
2347 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2348 Patch from Uwe Doering, sign-off by Michael Haardt.
2350 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2351 variable declaration deep within a block. Bug and patch from
2354 PP/15 lookups-Makefile IRIX compatibility coercion.
2356 PP/16 Make DISABLE_DKIM build knob functional.
2358 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2359 Patch by Simon Arlott
2361 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2362 variable argument macros. Our copy now differs from upstream.
2368 TF/01 Failure to get a lock on a hints database can have serious
2369 consequences so log it to the panic log.
2371 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2372 controlled using the smtp_confirmation log selector.
2374 TF/03 Include the error message when we fail to unlink a spool file.
2376 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2377 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2378 for maintaining out-of-tree patches for some time.
2380 PP/01 Bugzilla 139: Documentation and portability issues.
2381 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2382 Handle per-OS dynamic-module compilation flags.
2384 PP/02 Let /dev/null have normal permissions.
2385 The 4.73 fixes were a little too stringent and complained about the
2386 permissions on /dev/null. Exempt it from some checks.
2387 Reported by Andreas M. Kirchwitz.
2389 PP/03 Report version information for many libraries, including
2390 Exim version information for dynamically loaded libraries. Created
2391 version.h, now support a version extension string for distributors
2392 who patch heavily. Dynamic module ABI change.
2394 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2395 privilege escalation vulnerability whereby the Exim run-time user
2396 can cause root to append content of the attacker's choosing to
2399 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2402 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2403 If dropping privileges for untrusted macros, we disabled normal logging
2404 on the basis that it would fail; for the Exim run-time user, this is not
2405 the case, and it resulted in successful deliveries going unlogged.
2406 Fixed. Reported by Andreas Metzler.
2412 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2413 only prepend for the Resent-* case. Fixes regression introduced in
2414 Exim 4.70 by NM/22 for Bugzilla 607.
2416 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2417 increasing numbers of administrators be bitten by this.
2419 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2421 PP/03 Bugzilla 994: added openssl_options main configuration option.
2423 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2425 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2427 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2429 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2430 without a peer certificate, leading to a segfault because of an
2431 assumption that peers always have certificates. Be a little more
2432 paranoid. Problem reported by Martin Tscholak.
2434 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2435 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2436 NB: ClamAV planning to remove STREAM in "middle of 2010".
2437 CL also introduces -bmalware, various -d+acl logging additions and
2438 more caution in buffer sizes.
2440 PP/09 Implemented reverse_ip expansion operator.
2442 PP/10 Bugzilla 937: provide a "debug" ACL control.
2444 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2446 PP/12 Bugzilla 973: Implement --version.
2448 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2450 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2452 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2454 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2455 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2456 ignore trailing whitespace.
2458 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2460 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2463 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2464 Notification from Dr Andrew Aitchison.
2466 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2467 ExtendedDetectionInfo response format.
2468 Notification from John Horne.
2470 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2473 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2474 XSL and documented dependency on system catalogs, with examples of how
2477 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2480 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2481 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2482 configuration file which is writeable by the Exim user or group.
2484 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2485 of configuration files to cover files specified with the -C option if
2486 they are going to be used with root privileges, not just the default
2489 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2490 option (effectively making it always true).
2492 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2493 files to be used while preserving root privileges.
2495 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2496 that rogue child processes cannot use them.
2498 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2499 run-time user, instead of root.
2501 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2502 Exim run-time user without dropping privileges.
2504 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2505 result string, instead of calling string_vformat() twice with the same
2508 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2509 for other users. Others should always drop root privileges if they use
2510 -C on the command line, even for a whitelisted configure file.
2512 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2514 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2520 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2521 $data_path, and $header_path variables; fixed documentation bugs and
2524 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2525 exipick to access non-standard spools, including the "frozen" queue
2528 NM/01 Bugzilla 965: Support mysql stored procedures.
2529 Patch from Alain Williams
2531 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2533 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2534 Patch from Andreas Metzler
2536 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2537 Patch from Kirill Miazine
2539 NM/05 Bugzilla 671: Added umask to procmail example.
2541 JJ/03 installed exipick 20100323.0, fixing doc bug
2543 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2544 directory. Notification and patch from Dan Rosenberg.
2546 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2548 TK/02 Improve log output when DKIM signing operation fails.
2550 MH/01 Treat the transport option dkim_domain as a colon separated
2551 list, not as a single string, and sign the message with each element,
2552 omitting multiple occurences of the same signer.
2554 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2555 Bugzilla 985, 986. Patch by Simon Arlott
2557 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2558 Patch by Simon Arlott
2560 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2561 MBX locking. Notification from Dan Rosenberg.
2567 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2569 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2571 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2573 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2575 NM/04 Bugzilla 915: Flag broken perl installation during build.
2581 TK/01 Added patch by Johannes Berg that expands the main option
2582 "spamd_address" if it starts with a dollar sign.
2584 TK/02 Write list of recipients to X-Envelope-Sender header when building
2585 the mbox-format spool file for content scanning (suggested by Jakob
2588 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2589 (http://www.dcc-servers.net/) support via dccifd. Activated by
2590 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2592 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2593 by Mark Daniel Reidel <mr@df.eu>.
2595 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2596 When building exim an external PCRE library is now needed -
2597 PCRE is a system library on the majority of modern systems.
2598 See entry on PCRE_LIBS in EDITME file.
2600 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2601 conversation. Added nologin parameter to request.
2602 Patch contributed by Kirill Miazine.
2604 TF/01 Do not log submission mode rewrites if they do not change the address.
2606 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2608 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2609 log files in place. Contributed by Roberto Lima.
2611 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2613 TF/03 Bugzilla 615: When checking the local_parts router precondition
2614 after a local_part_suffix or local_part_prefix option, Exim now
2615 does not use the address's named list lookup cache, since this
2616 contains cached lookups for the whole local part.
2618 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2619 Robert Millan. Documentation is in experimental-spec.txt.
2621 TF/04 Bugzilla 668: Fix parallel build (make -j).
2623 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2625 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2626 Patch provided by Jan Srzednicki.
2628 TF/05 Leading white space used to be stripped from $spam_report which
2629 wrecked the formatting. Now it is preserved.
2631 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2632 that they are available at delivery time.
2634 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2636 TF/08 TLS error reporting now respects the incoming_interface and
2637 incoming_port log selectors.
2639 TF/09 Produce a more useful error message if an SMTP transport's hosts
2640 setting expands to an empty string.
2642 NM/06 Bugzilla 744: EXPN did not work under TLS.
2643 Patch provided by Phil Pennock.
2645 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2646 Patch provided by Richard Godbee.
2648 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2649 acl_smtp_notquit, added index entry.
2651 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2652 Patch provided by Eugene Bujak.
2654 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2655 accept(). Patch provided by Maxim Dounin.
2657 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2658 Patch provided by Phil Pennock.
2660 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2662 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2663 Patch provided by Brad "anomie" Jorsch.
2665 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2666 Patch provided by Dean Brooks.
2668 TK/05 Add native DKIM support (does not depend on external libraries).
2670 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2671 Patch provided by Graeme Fowler.
2673 NM/16 Bugzilla 851: Documentation example syntax fix.
2675 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2677 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2680 NM/19 Bugzilla 745: TLS version reporting.
2681 Patch provided by Phil Pennock.
2683 NM/20 Bugzilla 167: bool: condition support.
2684 Patch provided by Phil Pennock.
2686 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2687 clients. Patch provided by Phil Pennock.
2689 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2690 Patch provided by Brad "anomie" Jorsch.
2692 NM/23 Bugzilla 687: Fix misparses in eximstats.
2693 Patch provided by Heiko Schlittermann.
2695 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2696 Patch provided by Heiko Schlittermann.
2698 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2699 plus update to original patch.
2701 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2703 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2704 Patch provided by David Brownlee.
2706 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2708 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2710 NM/30 Bugzilla 888: TLS documentation bugfixes.
2712 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2714 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2715 Unlike the original bugzilla I have changed all shell scripts in src tree.
2717 NM/33 Bugzilla 898: Transport filter timeout fix.
2718 Patch by Todd Rinaldo.
2720 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2721 Patch by Serge Demonchaux.
2723 NM/35 Bugzilla 39: Base64 decode bug fixes.
2724 Patch by Jakob Hirsch.
2726 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2728 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2730 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2732 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2738 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2739 ALT-N's libdkim that I have put here:
2740 http://duncanthrax.net/exim-experimental/
2742 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2743 were called 'true' and it seems that C99 defines that as a reserved
2744 keyword to be used with 'bool' variable types. That means you could
2745 not include C99-style headers which use bools without triggering
2746 build errors in sieve.c.
2748 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2749 as mailq or other aliases. Changed the --help handling significantly
2750 to do whats expected. exim_usage() emits usage/help information.
2752 SC/01 Added the -bylocaldomain option to eximstats.
2754 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2756 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2758 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2764 PH/01 Another patch from the Sieve maintainer.
2766 PH/02 When an IPv6 address is converted to a string for single-key lookup
2767 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2768 dots are used instead of colons so that keys in lsearch files need not
2769 contain colons. This was done some time before quoting was made available
2770 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2771 (notated using the quote facility) so as to distinguish them from IPv4
2772 keys. This meant that lookups for IP addresses in host lists did not work
2773 for iplsearch lookups.
2775 This has been fixed by arranging for IPv6 addresses to be expressed with
2776 colons if the lookup type is iplsearch. This is not incompatible, because
2777 previously such lookups could never work.
2779 The situation is now rather anomalous, since one *can* have colons in
2780 ordinary lsearch keys. However, making the change in all cases is
2781 incompatible and would probably break a number of configurations.
2783 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2786 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2787 conversion specification without a maximum field width, thereby enabling
2788 a rogue spamd server to cause a buffer overflow. While nobody in their
2789 right mind would setup Exim to query an untrusted spamd server, an
2790 attacker that gains access to a server running spamd could potentially
2791 exploit this vulnerability to run arbitrary code as the Exim user.
2793 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2794 $primary_hostname instead of what libspf2 thinks the hosts name is.
2796 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2797 a directory entry by the name of the lookup key. Previously, if a
2798 symlink pointed to a non-existing file or a file in a directory that
2799 Exim lacked permissions to read, a lookup for a key matching that
2800 symlink would fail. Now it is enough that a matching directory entry
2801 exists, symlink or not. (Bugzilla 503.)
2803 PH/03 The body_linecount and body_zerocount variables are now exported in the
2806 PH/04 Added the $dnslist_matched variable.
2808 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2809 This means they are set thereafter only if the connection becomes
2812 PH/06 Added the client_condition to authenticators so that some can be skipped
2813 by clients under certain conditions.
2815 PH/07 The error message for a badly-placed control=no_multiline_responses left
2816 "_responses" off the end of the name.
2818 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2820 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2821 (without spaces) instead of just copying the configuration text.
2823 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2825 PH/11 Added $max_received_linelength.
2827 PH/12 Added +ignore_defer and +include_defer to host lists.
2829 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2830 way in which PCRE > 7.0 is built.
2832 PH/14 Implemented queue_only_load_latch.
2834 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2835 MAIL command. The effect was to mangle the value on 64-bit systems.
2837 PH/16 Another patch from the Sieve maintainer.
2839 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2841 PH/18 If a system quota error occurred while trying to create the file for
2842 a maildir delivery, the message "Mailbox is full" was not appended to the
2843 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2844 only to a quota excession during the actual writing of the file.
2846 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2847 characters?) which causes problems in log lines. The DN values are now
2848 passed through string_printing() before being added to log lines.
2850 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2851 and InterBase are left for another time.)
2853 PH/21 Added message_body_newlines option.
2855 PH/22 Guard against possible overflow in moan_check_errorcopy().
2857 PH/23 POSIX allows open() to be a macro; guard against that.
2859 PH/24 If the recipient of an error message contained an @ in the local part
2860 (suitably quoted, of course), incorrect values were put in $domain and
2861 $local_part during the evaluation of errors_copy.
2867 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2868 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2871 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2872 issue a MAIL command.
2874 PH/02 In an ACL statement such as
2876 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2878 if a client was not listed at all, or was listed with a value other than
2879 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2880 the condition was not true (as it should be), so access was not denied.
2881 The bug was that the ! inversion was incorrectly passed on to the second
2882 item. This has been fixed.
2884 PH/03 Added additional dnslists conditions == and =& which are different from
2885 = and & when the dns lookup returns more than one IP address.
2887 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2888 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2890 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2891 FSYNC, which compiles an option called disable_fsync that allows for
2892 bypassing fsync(). The documentation is heavily laced with warnings.
2894 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2896 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2897 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2898 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2899 including adding "make clean"; (3) Added -fPIC when compiling the test
2900 dynamically loaded module, to get rid of a warning.
2902 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2903 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2904 = 0s. The bug is otherwise harmless.
2906 PH/07 There was a bug in the dovecot authenticator such that the value of
2907 $auth1 could be overwritten, and so not correctly preserved, after a
2908 successful authentication. This usually meant that the value preserved by
2909 the server_setid option was incorrect.
2911 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2913 PH/09 Installed PCRE release 7.0.
2915 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2916 run for batched SMTP input. It is now run at the start of every message
2917 in the batch. While fixing this I discovered that the process information
2918 (output by running exiwhat) was not always getting set for -bs and -bS
2919 input. This is fixed, and it now also says "batched" for BSMTP.
2921 PH/11 Added control=no_pipelining.
2923 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2924 patch, slightly modified), and move the expansion of helo_data till after
2925 the connection is made in the smtp transport (so it can use these
2928 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2930 PH/14 Added log_selector = +pid.
2932 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2934 PH/16 Add ${if forany and ${if forall.
2936 PH/17 Added dsn_from option to vary the From: line in DSNs.
2938 PH/18 Flush SMTP output before performing a callout, unless control =
2939 no_callout_flush is set.
2941 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2942 was true (the default) a successful delivery failed to delete the retry
2943 item, thus causing premature timeout of the address. The bug is now
2946 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2948 PH/21 Long custom messages for fakedefer and fakereject are now split up
2949 into multiline responses in the same way that messages for "deny" and
2950 other ACL rejections are.
2952 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2953 with slight modification.
2955 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2956 draft, changing the syntax and factoring some duplicate code".
2958 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2959 for deliveries of the second and subsequent messages over the same SMTP
2962 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2963 ${reduce, with only minor "tidies".
2965 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2967 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2968 expansion side effects.
2970 PH/27 When a message times out after an over-quota error from an Exim-imposed
2971 quota, the bounce message says "mailbox is full". This message was not
2972 being given when it was a system quota that was exceeded. It now should
2975 MH/03 Made $recipients available in local_scan(). local_scan() already has
2976 better access to the recipient list through recipients_list[], but
2977 $recipients can be useful in postmaster-provided expansion strings.
2979 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2980 in the case of a MAIL command with additional options following the
2981 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2982 were accidentally chopped off.
2984 PH/29 SMTP synchronization checks are implemented when a command is read -
2985 there is a check that no more input is waiting when there shouldn't be
2986 any. However, for some commands, a delay in an ACL can mean that it is
2987 some time before the response is written. In this time, more input might
2988 arrive, invalidly. So now there are extra checks after an ACL has run for
2989 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2990 pipelining has not been advertised.
2992 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2994 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2995 returned in the SMTP response when smtp_return_error_details was false.
2996 This has been fixed.
2998 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2999 instead of the C I/O that was originally supplied, because problems were
3000 reported on Solaris.
3002 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3003 Exim which did not show up earlier: it was assuming that a call to
3004 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3005 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3006 was a macro that became an assignment, so it seemed to work. This has
3007 changed to a proper function call with a void return, hence the compile
3008 error. Exim's code has been fixed.
3010 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3013 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3015 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3017 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3018 because it turns out to be too big or there is a timeout) there is no
3019 "Completed" line in the log. When some messages of this type were
3020 selected by exigrep, they were listed as "not completed". Others were
3021 picked up by some special patterns. I have improved the selection
3022 criteria to be more general.
3024 PH/38 The host_find_failed option in the manualroute router can now be set
3025 to "ignore", to completely ignore a host whose IP address cannot be
3026 found. If all hosts are ignored, the behaviour is controlled by the new
3027 host_all_ignored option.
3029 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3030 homing or because of multiple MX records with /mx) generated more than
3031 one IP address, and the following item turned out to be the local host,
3032 all the secondary addresses of the first item were incorrectly removed
3033 from the list, along with the local host and any following hosts (which
3034 is what is supposed to happen).
3036 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3037 whoever called Exim into the -H file. In the case of the daemon it was
3038 behaving confusingly. When first started, it used values for whoever
3039 started the daemon, but after a SIGHUP it used the Exim user (because it
3040 calls itself on a restart). I have changed the code so that it now always
3043 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3044 message are rejected with the same error (e.g. no authentication or bad
3045 sender address), and a DATA command is nevertheless sent (as can happen
3046 with PIPELINING or a stupid MUA), the error message that was given to the
3047 RCPT commands is included in the rejection of the DATA command. This is
3048 intended to be helpful for MUAs that show only the final error to their
3051 PH/42 Another patch from the Sieve maintainer.
3053 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3054 Eximstats - Fixed some broken HTML links and added missing column headers
3056 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3057 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3059 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3061 PH/43 Yet another patch from the Sieve maintainer.
3063 PH/44 I found a way to check for a TCP/IP connection going away before sending
3064 the response to the final '.' that terminates a message, but only in the
3065 case where the client has not sent further data following the '.'
3066 (unfortunately, this is allowed). However, in many cases there won't be
3067 any further data because there won't be any more messages to send. A call
3068 to select() can be used: if it shows that the input is "ready", there is
3069 either input waiting, or the socket has been closed. An attempt to read
3070 the next input character can distinguish the two cases. Previously, Exim
3071 would have sent an OK response which the client would never have see.
3072 This could lead to message repetition. This fix should cure that, at
3073 least in a lot of common cases.
3075 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3076 advertised in response to EHLO.
3082 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3083 fixed by 4.65/MH/01 (is this a record?) are fixed:
3085 (i) An empty string was always treated as zero by the numeric comparison
3086 operators. This behaviour has been restored.
3088 (ii) It is documented that the numeric comparison operators always treat
3089 their arguments as decimal numbers. This was broken in that numbers
3090 starting with 0 were being interpreted as octal.
3092 While fixing these problems I realized that there was another issue that
3093 hadn't been noticed. Values of message_size_limit (both the global option
3094 and the transport option) were treated as octal if they started with 0.
3095 The documentation was vague. These values are now always treated as
3096 decimal, and I will make that clear in the documentation.
3102 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3103 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3106 MH/01 Don't check that the operands of numeric comparison operators are
3107 integers when their expansion is in "skipping" mode (fixes bug
3108 introduced by 4.64-PH/07).
3110 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3111 child addresses, Exim now panics and dies. Previously, because the count
3112 is held in a short int, deliveries were likely to be lost. As such a
3113 large number of recipients for a single message is ridiculous
3114 (performance will be very, very poor), I have chosen to impose a limit
3115 rather than extend the field.
3121 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3122 leftover -K file (the existence of which was triggered by #402).
3123 While we were at it, introduced process PID as part of the -K
3124 filename. This should rule out race conditions when creating
3127 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3128 processing considerably. Previous code took too long for large mails,
3129 triggering a timeout which in turn triggers #401.
3131 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3132 in the DK code in transports.c. sendfile() is not really portable,
3133 hence the _LINUX specificness.
3135 TF/01 In the add_headers option to the mail command in an Exim filter,
3136 there was a bug that Exim would claim a syntax error in any
3137 header after the first one which had an odd number of characters
3140 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3141 callout verification, Exim cached a "reject" for the entire domain. This
3142 is correct for most verifications, but it is not correct for a recipient
3143 verification with use_sender or use_postmaster set, because in that case
3144 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3145 case of MAIL FROM:<> rejection from other early rejections (e.g.
3146 rejection of HELO). When verifying a recipient using a non-null MAIL
3147 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3148 Whatever the result of the callout, the value of the domain cache is
3149 left unchanged (for any other kind of callout, getting as far as trying
3150 RCPT means that the domain itself is ok).
3152 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3155 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3156 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3157 session, the daemon ignores ECONNECT errors and logs others; it now
3158 ignores EPIPE as well.
3160 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3161 (quoted-printable decoding).
3163 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3164 later the small subsequent patch to fix an introduced bug.
3166 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3168 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3170 PH/08 An error is now given if message_size_limit is specified negative.
3172 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3173 to be given (somewhat) arbitrary names.
3175 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3178 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3179 miscellaneous code fixes
3181 PH/10 Added the log_reject_target ACL modifier to specify where to log
3184 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3185 hostname. This is wrong, because it relates to the incoming message (and
3186 probably the interface on which it is arriving) and not to the outgoing
3187 callout (which could be using a different interface). This has been
3188 changed to use the value of the helo_data option from the smtp transport
3189 instead - this is what is used when a message is actually being sent. If
3190 there is no remote transport (possible with a router that sets up host
3191 addresses), $smtp_active_hostname is used.
3193 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3194 tweaks were necessary in order to get it to work (see also 21 below):
3195 (a) The code assumed that strncpy() returns a negative number on buffer
3196 overflow, which isn't the case. Replaced with Exim's string_format()
3198 (b) There were several signed/unsigned issues. I just did the minimum
3199 hacking in of casts. There is scope for a larger refactoring.
3200 (c) The code used strcasecmp() which is not a standard C function.
3201 Replaced with Exim's strcmpic() function.
3202 (d) The code set only $1; it now sets $auth1 as well.
3203 (e) A simple test gave the error "authentication client didn't specify
3204 service in request". It would seem that Dovecot has changed its
3205 interface. Fortunately there's a specification; I followed it and
3206 changed what the client sends and it appears to be working now.
3208 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3211 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3212 address A is aliased to B and C, where B exists and C does not. Without
3213 -v the output is "A verified" because verification stops after a
3214 successful redirection if more than one address is generated. However,
3215 with -v the child addresses are also verified. Exim was outputting "A
3216 failed to verify" and then showing the successful verification for C,
3217 with its parentage. It now outputs "B failed to verify", showing B's
3218 parentage before showing the successful verification of C.
3220 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3221 look up a TXT record in a specific list after matching in a combined
3224 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3225 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3226 they consult the DNS. I had assumed they would set it the way they
3227 wanted; and indeed my experiments on Linux seem to show that in some
3228 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3229 To be on the safe side, however, I have now made the interface to
3230 host_find_byname() similar to host_find_bydns(), with an argument
3231 containing the DNS resolver options. The host_find_byname() function now
3232 sets these options at its start, just as host_find_bydns() does. The smtp
3233 transport options dns_qualify_single and dns_search_parents are passed to
3234 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3235 of host_find_byname() use the default settings of RES_DEFNAMES
3236 (qualify_single) but not RES_DNSRCH (search_parents).
3238 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3239 spool_read_header() do less string testing, by means of a preliminary
3240 switch on the second character of optional "-foo" lines. (This is
3241 overdue, caused by the large number of possibilities that now exist.
3242 Originally there were few.) While I was there, I also converted the
3243 str(n)cmp tests so they don't re-test the leading "-" and the first
3244 character, in the hope this might squeeze out yet more improvement.
3246 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3247 flag allowing group syntax was set by the header_syntax check but not
3248 turned off, possible causing trouble later; (2) The flag was not being
3249 set at all for the header_verify test, causing "group"-style headers to
3250 be rejected. I have now set it in this case, and also caused header_
3251 verify to ignore an empty address taken from a group. While doing this, I
3252 came across some other cases where the code for allowing group syntax
3253 while scanning a header line wasn't quite right (mostly, not resetting
3254 the flag correctly in the right place). These bugs could have caused
3255 trouble for malformed header lines. I hope it is now all correct.
3257 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3258 with the "reply" argument non-NULL. The code, however (which originally
3259 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3260 but it didn't always do it. This confused somebody who was copying the
3261 code for some other use. I have removed all the tests.
3263 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3264 feature that was used to support insecure browsers during the U.S. crypto
3265 embargo. It requires special client support, and Exim is probably the
3266 only MTA that supported it -- and would never use it because real RSA is
3267 always available. This code has been removed, because it had the bad
3268 effect of slowing Exim down by computing (never used) parameters for the
3269 RSA_EXPORT functionality.
3271 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3272 authenticator to fail if there's a tab character in the incoming data
3273 (there should never be unless someone is messing about, as it's supposed
3274 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3275 if the connection is using TLS or if the remote IP is the same as the
3276 local IP, and the "valid-client-cert option" if a client certificate has
3279 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3280 authenticators. This can be used for authorization after authentication
3281 succeeds. (In the case of plaintext, it servers for both authentication
3284 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3285 if any retry times were supplied.
3287 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3288 connection, where there is no client IP address to check. In this
3289 situation, the verify now always succeeds.
3291 PH/25 Applied John Jetmore's -Mset patch.
3293 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3295 PH/27 In a string expansion for a processed (not raw) header when multiple
3296 headers of the same name were present, leading whitespace was being
3297 removed from all of them, but trailing whitespace was being removed only
3298 from the last one. Now trailing whitespace is removed from each header
3299 before concatenation. Completely empty headers in a concatenation (as
3300 before) are ignored.
3302 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3303 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3305 PH/29 [Removed. This was a change that I later backed out, and forgot to
3306 correct the ChangeLog entry (that I had efficiently created) before
3307 committing the later change.]
3309 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3310 address errors (4xx response to RCPT) over the same connection as other
3311 messages routed to the same hosts. Such deliveries are always "forced",
3312 so retry times are not inspected. This resulted in far too many retries
3313 for the affected addresses. The effect occurred only when there were more
3314 hosts than the hosts_max_try setting in the smtp transport when it had
3315 the 4xx errors. Those hosts that it had tried were not added to the list
3316 of hosts for which the message was waiting, so if all were tried, there
3317 was no problem. Two fixes have been applied:
3319 (i) If there are any address or message errors in an SMTP delivery, none
3320 of the hosts (tried or untried) are now added to the list of hosts
3321 for which the message is waiting, so the message should not be a
3322 candidate for sending over the same connection that was used for a
3323 successful delivery of some other message. This seems entirely
3324 reasonable: after all the message is NOT "waiting for some host".
3325 This is so "obvious" that I'm not sure why it wasn't done
3326 previously. Hope I haven't missed anything, but it can't do any
3327 harm, as the worst effect is to miss an optimization.
3329 (ii) If, despite (i), such a delivery is accidentally attempted, the
3330 routing retry time is respected, so at least it doesn't keep
3331 hammering the server.
3333 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3334 in ${readsocket because some servers need this prod.
3336 PH/32 Added some extra debug output when updating a wait-xxx database.
3338 PH/33 The hint "could be header name not terminated by colon", which has been
3339 given for certain expansion errors for a long time, was not being given
3340 for the ${if def:h_colon_omitted{... case.
3342 PH/34 The spec says: "With one important exception, whenever a domain list is
3343 being scanned, $domain contains the subject domain." There was at least
3344 one case where this was not true.
3346 PH/35 The error "getsockname() failed: connection reset by peer" was being
3347 written to the panic log as well as the main log, but it isn't really
3348 panic-worthy as it just means the connection died rather early on. I have
3349 removed the panic log writing for the ECONNRESET error when getsockname()
3352 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3353 runs only) independently of the message's sender address. This meant
3354 that, if the 4xx error was in fact related to the sender, a different
3355 message to the same recipient with a different sender could confuse
3356 things. In particular, this can happen when sending to a greylisting
3357 server, but other circumstances could also provoke similar problems.
3358 I have changed the default so that the retry time for these errors is now
3359 based a combination of the sender and recipient addresses. This change
3360 can be overridden by setting address_retry_include_sender=false in the
3363 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3364 remote server are returned as part of bounce messages. This was not
3365 happening for LMTP over a pipe (the lmtp transport), but now it is the
3366 same for both kinds of LMTP.
3368 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3369 in header lines that were in fact CNAMEs. This is no longer the case.
3371 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3372 and queue runs started by the daemon processed all messages. This has
3373 been fixed so that -R and -S can now usefully be given with -q<time>.
3375 PH/40 Import PCRE release 6.7 (fixes some bugs).
3377 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3379 PH/42 Give an error if -q is specified more than once.
3381 PH/43 Renamed the variables $interface_address and $interface_port as
3382 $received_ip_address and $received_port, to make it clear that these
3383 values apply to message reception, and not to the outgoing interface when
3384 a message is delivered. (The old names remain recognized, of course.)
3386 PH/44 There was no timeout on the connect() call when using a Unix domain
3387 socket in the ${readsocket expansion. There now is.
3389 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3390 be meaningful with "accept".
3392 SC/01 Eximstats V1.43
3393 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3395 SC/02 Eximstats V1.44
3396 Use a glob alias rather than an array ref in the generated
3397 parser. This improves both readability and performance.
3399 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3400 Collect SpamAssassin and rejection statistics.
3401 Don't display local sender or destination tables unless
3402 there is data to show.
3403 Added average volumes into the top table text output.
3405 SC/04 Eximstats V1.46
3406 Collect data on the number of addresses (recipients)
3407 as well as the number of messages.
3409 SC/05 Eximstats V1.47
3410 Added 'Message too big' to the list of mail rejection
3411 reasons (thanks to Marco Gaiarin).
3413 SC/06 Eximstats V1.48
3414 Mainlog lines which have GMT offsets and are too short to
3415 have a flag are now skipped.
3417 SC/07 Eximstats V1.49 (Alain Williams)
3418 Added the -emptyok flag.
3420 SC/08 Eximstats V1.50
3421 Fixes for obtaining the IP address from reject messages.
3423 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3424 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3425 whitespace changes from 4.64-PH/27
3427 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3430 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3431 are found, allow negative numbers in numeric criteria)
3433 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3435 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3436 to match changes made in 4.64-PH/43
3438 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3440 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3441 "domain" is actually an IP address, and force a failure. This locks out
3442 those revolvers/nameservers that support "A-for-A" lookups, in
3443 contravention of the specifications.
3445 PH/48 When a host name was looked up from an IP address, and the subsequent
3446 forward lookup of the name timed out, the host name was left in
3447 $sender_host_name, contrary to the specification.
3449 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3450 restricted to single-key lookups, Exim was not diagnosing an error if
3451 * or *@ was used with a query-style lookup.
3453 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3455 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3456 long ago, but noone interested enough thought of it. Let's just say that
3457 the "1.1" means that there are some new functions that weren't there at
3458 some point in the past.
3460 PH/51 Error processing for expansion failure of helo_data from an smtp
3461 transport during callout processing was broken.
3463 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3464 tested/used via the -bh/-bhc/-bs options.
3466 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3467 bug, fixed in subsequent PCRE releases).
3469 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3470 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3472 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3478 SC/01 Use a glob alias rather than an array ref in eximstats generated
3479 parser. This improves both readability and performance.
3481 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3482 Don't display local sender or destination tables in eximstats unless
3483 there is data to show.
3484 Added average volumes into the eximstats top table text output.
3486 SC/03 Collect data on the number of addresses (recipients) as well
3487 as the number of messages in eximstats.
3489 TF/01 Correct an error in the documentation for the redirect router. Exim
3490 does (usually) call initgroups() when daemonizing.
3492 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3493 with consistent privilege compared to when running as a daemon.
3495 TF/03 Note in the spec that $authenticated_id is not set for local
3496 submissions from trusted users.
3498 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3499 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3501 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3502 by adding some example configuration directives to the default
3503 configuration file. A little bit of work is required to uncomment the
3504 directives and define how usernames and passwords are checked, but
3505 there is now a framework to start from.
3507 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3508 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3509 without this. I don't know how relevant this is to other LDAP libraries.
3511 PH/02 Add the verb name to the "unknown ACL verb" error.
3513 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3515 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3517 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3518 directory not expanded when it should be if an expanded home directory
3519 was set for the address (which is overridden by the transport).
3521 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3524 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3525 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3526 because it is too late at that time, and has no effect.
3528 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3529 security issue with \' (bugzilla #107). I could not use the
3530 PQescapeStringConn() function, because it needs a PGconn value as one of
3533 PH/08 When testing addresses using -bt, indicate those final addresses that
3534 are duplicates that would not cause an additional delivery. At least one
3535 person was confused, thinking that -bt output corresponded to deliveries.
3536 (Suppressing duplicates isn't a good idea as you lose the information
3537 about possibly different redirections that led to the duplicates.)
3539 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3540 systems where poll() doesn't work, in particular OS X.
3542 PH/10 Added more information to debugging output for retry time not reached.
3544 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3545 operations in malware.c.
3547 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3550 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3551 syslog_duplication set false, log lines that would normally be written
3552 both the the main log and to the reject log were not written to syslog at
3555 PH/14 In the default configuration, change the use of "message" in ACL warn
3556 statements to "add_header".
3558 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3559 not followed by a command (e.g. "seen endif").
3561 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3562 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3565 PH/17 Added extra conditions to the default value of delay_warning_condition
3569 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3570 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3571 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3574 The Auto-Submitted: and various List- headers are standardised, whereas I
3575 don't think Precedence: ever was.
3577 PH/18 Refactored debugging code in route_finduser() to show more information,
3578 in particular, the error code if getpwnam() issues one.
3580 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3581 This is apparently needed in addition to the PH/07 change above to avoid
3582 any possible encoding problems.
3584 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3585 but not after initializing Perl.
3587 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3588 output them only if debugging. By default they are written stderr,
3589 apparently, which is not desirable.
3591 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3594 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3597 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3599 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3600 authenticated or an ident call has been made. Suppress the default
3601 values for $authenticated_id and $authenticated_sender (but permit -oMai
3602 and -oMas) when testing with -bh.
3604 PH/24 Re-jigged the order of the tests in the default configuration so that the
3605 tests for valid domains and recipients precede the DNS black list and CSA
3606 tests, on the grounds that those ones are more expensive.
3608 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3609 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3610 This bug exists in every version of Exim that I still have, right back to
3613 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3614 However, an attempt to turn on case-sensitivity in a regex key by
3615 including (?-i) didn't work because the subject string was already
3616 lowercased, and the effects were non-intuitive. It turns out that a
3617 one-line patch can be used to allow (?-i) to work as expected.
3623 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3624 other effects) broke the use of negated acl sub-conditions.
3626 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3629 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3630 "Deny" causes Exim to reject the incoming connection with a 554 error.
3631 Unfortunately, if there is a major crisis, such as a disk failure,
3632 tcp-wrappers gives "deny", whereas what one would like would be some
3633 kind of temporary error. A kludge has been added to help with this.
3634 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3635 554 error is used if errno is still zero or contains ENOENT (which occurs
3636 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3639 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3641 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3642 errors. Otherwise a message that provokes a temporary error (when other
3643 messages do not) can cause a whole host to time out.
3645 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3646 addresses were routed directly to files or pipes from a redirect router.
3647 File deliveries just didn't batch; pipe deliveries might have suffered
3650 PH/06 A failure to get a lock for a hints database would erroneously always say
3651 "Failed to get write lock", even when it was really a read lock.
3653 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3654 of 0600. This has been changed to use the value of the lockfile_mode
3655 option (which defaults to 0600).
3657 PH/08 Applied small patch from the Sieve maintainer.
3659 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3660 folder from quota calculations, a direct delivery into this folder messed
3661 up the contents of the maildirsize file. This was because the regex was
3662 used only to exclude .Trash (or whatever) when the size of the mailbox
3663 was calculated. There was no check that a delivery was happening into an
3664 excluded directory. This bug has been fixed by ignoring all quota
3665 processing for deliveries into excluded directories.
3667 PH/10 Added the maildirfolder_create_regex option to appendfile.
3673 PH/01 The code for finding all the local interface addresses on a FreeBSD
3674 system running IPv6 was broken. This may well have applied to all BSD
3675 systems, as well as to others that have similar system calls. The broken
3676 code found IPv4 interfaces correctly, but gave incorrect values for the
3677 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3678 that it would not match correctly against @[] and not recognize the IPv6
3681 PH/02 The ipliteral router was not recognizing addresses of the form user@
3682 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3684 PH/03 Added disable_ipv6.
3686 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3687 decoded form, because it is most often used to construct To: headers
3688 lines in autoreplies, and the decoded form may well be syntactically
3689 invalid. However, $reply_address has leading white space removed, and all
3690 newlines turned into spaces so that the autoreply transport does not
3693 PH/05 If group was specified without a user on a router, and no group or user
3694 was specified on a transport, the group from the router was ignored.
3696 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3697 for visible compile-time settings that can be used to change these
3698 numbers, for those that want even more. Backwards compatibility with old
3699 spool files has been maintained. However, going back to a previous Exim
3700 release will lost any variables that are in spool files.
3702 PH/07 Two small changes when running in the test harness: increase delay when
3703 passing a TCP/IP connection to a new process, in case the original
3704 process has to generate a bounce, and remove special handling of
3705 127.0.0.2 (sic), which is no longer necessary.
3707 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3708 be the same on different OS.
3710 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3713 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3714 whether --show-vars was specified or not
3716 JJ/02 exipick: Added support for new ACL variable spool format introduced
3719 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3720 syntactically invalid From: or Reply-to: line, and a filter used this to
3721 generate an autoreply, and therefore failed to obtain an address for the
3722 autoreply, Exim could try to deliver to a non-existent relative file
3723 name, causing unrelated and misleading errors. What now happens is that
3724 it logs this as a hard delivery error, but does not attempt to create a
3727 PH/11 The exinext utility has a -C option for testing purposes, but although
3728 the given file was scanned by exinext itself; it wasn't being passed on
3729 when Exim was called.
3731 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3732 an end-of-file indication when reading a command response.
3734 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3735 compiled. In many other places in Exim, IPv6 addresses are always
3736 recognized, so I have changed this. It also means that IPv4 domain
3737 literals of the form [IPV4:n.n.n.n] are now always recognized.
3739 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3740 used if the router is not running as root, for example, when verifying at
3741 ACL time, or when using -bh. The debugging output from this situation was
3742 non-existent - all you got was a failure to exec. I have made two
3745 (a) Failures to set uid/gid, the current directory, or a process leader
3746 in a subprocess such as that created by queryprogram now generate
3747 suitable debugging output when -d is set.
3749 (b) The queryprogram router detects when it is not running as root,
3750 outputs suitable debugging information if -d is set, and then runs
3751 the subprocess without attempting to change uid/gid.
3753 PH/15 Minor change to Makefile for building test_host (undocumented testing
3756 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3757 additional section of a DNS packet that returns MX or SRV records.
3758 Instead, it always explicitly searches for A/AAAA records. This avoids
3759 major problems that occur when a DNS server includes only records of one
3760 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3761 fixed another bug: if SRV records were looked up and the corresponding
3762 address records were *not* found in the additional section, the port
3763 values from the SRV records were lost.
3765 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3766 using the correct key (the original address) when searching the retry
3767 rules in order to find which one to use for generating the retry hint.
3769 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3770 adding the default one. Similarly, if it contains a Reply-To: header, the
3771 errors_reply_to option, if set, is not used.
3773 PH/19 When calculating a retry time, Exim used to measure the "time since
3774 failure" by looking at the "first failed" field in the retry record. Now
3775 it does not use this if it is later than than the arrival time of the
3776 message. Instead it uses the arrival time. This makes for better
3777 behaviour in cases where some deliveries succeed, thus re-setting the
3778 "first failed" field. An example is a quota failure for a huge message
3779 when small messages continue to be delivered. Without this change, the
3780 "time since failure" will always be short, possible causing more frequent
3781 delivery attempts for the huge message than are intended.
3782 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3784 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3785 $1, $2, $3) because the numerical variables can be reset during some
3786 expansion items (e.g. "match"), thereby losing the authentication data.
3788 PH/21 Make -bV show the size of off_t variables so that the test suite can
3789 decide whether to run tests for quotas > 2G.
3791 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3792 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3793 filecount value is greater than 2G or if a quota value is greater than 2G
3794 on a system where the size of off_t is not greater than 4, a panic error
3797 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3798 never match. The debug and -bh output now contains an explicit error
3799 message indicating a malformed IPv4 address or mask.
3801 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3802 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3803 PH/23 above applies.
3805 PH/25 Do not write to syslog when running in the test harness. The only
3806 occasion when this arises is a failure to open the main or panic logs
3807 (for which there is an explicit test).
3809 PH/26 Added the /no_tell option to "control=freeze".
3811 PH/27 If a host name lookup failed very early in a connection, for example, if
3812 the IP address matched host_lookup and the reverse lookup yielded a name
3813 that did not have a forward lookup, an error message of the form "no IP
3814 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3815 could be logged. Now it outputs the IP address instead of "NULL".
3817 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3818 allows the sender and the authenticated sender to be set when
3819 submitting a message from within Exim. Since child_open_exim() is
3820 documented for local_scan(), the new function should be too.
3822 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3823 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3824 results in an empty string is now treated as unset.
3826 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3828 PH/31 Added sender_verify_fail logging option.
3830 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3831 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3832 tidied the source and removed it altogether.
3834 PH/33 When a queue run was abandoned because the load average was too high, a
3835 log line was always written; now it is written only if the queue_run log
3836 selector is set. In addition, the log line for abandonment now contains
3837 information about the queue run such as the pid. This is always present
3838 in "start" and "stop" lines but was omitted from the "abandon" line.
3840 PH/34 Omit spaces between a header name and the colon in the error message that
3841 is given when verify = headers_syntax fails (if there are lots of them,
3842 the message gets confusing).
3844 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3845 names, as there are now some PTR records that contain slashes. This check
3846 is only to protect against broken name servers that fall over on strange
3847 characters, so the fact that it applies to all lookups doesn't matter.
3849 PH/36 Now that the new test suite is complete, we can remove some of the
3850 special code in Exim that was needed for the old test suite. For example,
3851 sorting DNS records because real resolvers return them in an arbitrary
3852 order. The new test suite's fake resolver always returns records in the
3855 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3856 bounces) except when queue_only is set, to avoid logging races between
3857 the different processes.
3859 PH/38 Panic-die if .include specifies a non-absolute path.
3861 PH/39 A tweak to the "H" retry rule from its user.
3863 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3864 a label. They prevented compilation on older perls.
3866 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3867 a warning to be raised on newish perls.
3869 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3870 on queue. Changes to match documented behaviour of showing count of
3871 messages matching specified criteria.
3873 PH/40 Changed the default ident timeout from 30s to 5s.
3875 PH/41 Added support for the use of login_cap features, on those BSD systems
3876 that have them, for controlling the resources used by pipe deliveries.
3878 PH/42 The content-scanning code uses fopen() to create files in which to put
3879 message data. Previously it was not paying any attention to the mode of
3880 the files. Exim runs with umask(0) because the rest of the code creates
3881 files with open(), and sets the required mode explicitly. Thus, these
3882 files were ending up world-writeable. This was not a big issue, because,
3883 being within the spool directory, they were not world-accessible. I have
3884 created a function called modefopen, which takes an additional mode
3885 argument. It sets umask(777), creates the file, chmods it to the required
3886 mode, then resets the umask. All the relevant calls to fopen() in the
3887 content scanning code have been changed to use this function.
3889 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3890 to 24 hours. This avoids potential overflow problems when processing G
3891 and H retry rules. I suspect nobody ever tinkers with this value.
3893 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3895 PH/45 When the plaintext authenticator is running as a client, the server's
3896 challenges are checked to ensure they are valid base64 strings. By
3897 default, the authentication attempt is cancelled if an invalid string is
3898 received. Setting client_ignore_invalid_base64 true ignores these errors.
3899 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3900 they are received. Thus, the responses can be made to depend on the
3901 challenges. If an invalid string is ignored, an empty string is placed in
3904 PH/46 Messages that are created by the autoreply transport now contains a
3905 References: header, in accordance with RFCs 2822 and 3834.
3907 PH/47 Added authenticated_sender_force to the smtp transport.
3909 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3911 PH/49 Installed latest patch from the Sieve maintainer.
3913 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3914 was also set, the appendfile transport was unnecessarily scanning a
3915 directory of message files (e.g. for maildir delivery) to find the count
3916 of files (along with the size), even though it did not need this
3917 information. It now does the scan only if it needs to find either the
3918 size of the count of files.
3920 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3922 PH/52 Two bugs concerned with error handling when the smtp transport is
3925 (i) Exim was not creating retry information for temporary errors given
3926 for individual recipients after the DATA command when the smtp transport
3927 was used in LMTP mode. This meant that they could be retried too
3928 frequently, and not timed out correctly.
3930 (ii) Exim was setting the flag that allows error details to be returned
3931 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3932 recipients that were returned after the DATA command.
3934 PH/53 This is related to PH/52, but is more general: for any failing address,
3935 when detailed error information was permitted to be returned to the
3936 sender, but the error was temporary, then after the final timeout, only
3937 "retry timeout exceeded" was returned. Now it returns the full error as
3938 well as "retry timeout exceeded".
3940 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3941 do this, and (what is worse) MTAs that accept it.
3943 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3944 will now be deprecated.
3946 PH/56 New os.c-cygwin from the Cygwin maintainer.
3948 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3949 formats (previously only available in exim formats via -bpr, -bpru,
3950 and -bpra. Now also available in native and exiqgrep formats)
3952 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3953 with very large, slow to parse queues
3955 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3957 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3959 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3960 responses to authentication challenges, though it was showing the
3961 challenges; (ii) I've removed the CR characters from the debug output for
3964 PH/58 Allow for the insertion of a newline as well as a space when a string
3965 is turned into more than one encoded-word during RFC 2047 encoding. The
3966 Sieve code now uses this.
3968 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3969 data_4xx, lost_connection, tls_required.
3971 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3972 message was being sent as an SMTP response.
3974 PH/61 Add -l and -k options to exicyclog.
3976 PH/62 When verifying, if an address was redirected to one new address, so that
3977 verification continued, and the new address failed or deferred after
3978 having set something in $address_data, the value of $address_data was not
3979 passed back to the ACL. This was different to the case when no
3980 redirection occurred. The value is now passed back in both cases.
3982 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3983 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3984 use login_cap.h, so on its own it isn't the distinguishing feature. The
3985 new name refers directly to the setclassresources() function.
3987 PH/65 Added configuration files for NetBSD3.
3989 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3991 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3992 is preferred over IPv4.
3994 PH/68 The bounce_return_message and bounce_return_body options were not being
3995 honoured for bounces generated during the reception of non-SMTP messages.
3996 In particular, this applied to messages rejected by the ACL. This bug has
3997 been fixed. However, if bounce_return_message is true and bounce_return_
3998 body is false, the headers that are returned for a non-SMTP message
3999 include only those that have been read before the error was detected.
4000 (In the case of an ACL rejection, they have all been read.)
4002 PH/69 The HTML version of the specification is now built in a directory called
4003 spec_html instead of spec.html, because the latter looks like a path with
4004 a MIME-type, and this confuses some software.
4006 PH/70 Catch two compiler warnings in sieve.c.
4008 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4009 function verify_get_ident() calls ip_connect() to connect a socket, but
4010 if the "connect()" function timed out, ip_connect() used to close the
4011 socket. However, verify_get_ident() also closes the socket later, and in
4012 between Exim writes to the log, which may get opened at this point. When
4013 the socket was closed in ip_connect(), the log could get the same file
4014 descriptor number as the socket. This naturally causes chaos. The fix is
4015 not to close the socket in ip_connect(); the socket should be closed by
4016 the function that creates it. There was only one place in the code where
4017 this was missing, in the iplookup router, which I don't think anybody now
4018 uses, but I've fixed it anyway.
4020 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4021 well as to direct DNS lookups. Otherwise the handling of names in host
4022 lists is inconsistent and therefore confusing.
4028 PH/01 Two changes to the default runtime configuration:
4030 (1) Move the checks for relay_from_hosts and authenticated clients from
4031 after to before the (commented out) DNS black list checks.
4033 (2) Add control=submission to the relay_from_hosts and authenticated
4034 clients checks, on the grounds that messages accepted by these
4035 statements are most likely to be submissions.
4037 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4039 (1) Generate an error if the third argument for the ${prvs expansion is
4042 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4045 (3) Reset the variables that are obtained from the first argument of
4046 ${prvscheck and used in the second argument before leaving the code,
4047 because their memory is reclaimed, so using them afterwards may do
4050 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4051 one (it's much easier than Tom thought :-).
4053 (5) Because of (4), we can now allow for the use of $prvscheck_result
4054 inside the third argument.
4056 PH/03 For some reason, the default setting of PATH when running a command from
4057 a pipe transport was just "/usr/bin". I have changed it to
4060 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4061 anything to be listed in the output from -bV.
4063 PH/05 When a filter generated an autoreply, the entire To: header line was
4064 quoted in the delivery log line, like this:
4066 => >A.N.Other <ano@some.domain> <original@ddress> ...
4068 This has been changed so that it extracts the operative address. There
4069 may be more than one such address. If so, they are comma-separated, like
4072 => >ano@some.domain,ona@other.domain <original@ddress> ...
4074 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4075 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4076 not being looked up in the rDNS to get a host name, Exim was showing the
4077 IP address twice in Received: lines, even though the IP addresses were
4078 identical. For example:
4080 Received: from [1.2.3.4] (helo=[1.2.3.4])
4082 However, if the real host name was known, it was omitting the HELO data
4083 if it matched the actual IP address. This has been tidied up so that it
4084 doesn't show the same IP address twice.
4086 PH/07 When both +timestamp and +memory debugging was on, the value given by
4087 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4088 called by the debug printing, thereby overwriting the timestamp buffer.
4089 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4091 PH/08 When the original message was included in an autoreply transport, it
4092 always said "this is a copy of the message, including all the headers",
4093 even if body_only or headers_only was set. It now gives an appropriate
4096 PH/09 Applied a patch from the Sieve maintainer which:
4098 o fixes some comments
4099 o adds the (disabled) notify extension core
4100 o adds some debug output for the result of if/elsif tests
4101 o points to the current vacation draft in the documentation
4102 and documents the missing references header update
4106 o fixes a bug in processing the envelope test (when testing
4107 multiple envelope elements, the last element determined the
4110 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4111 Electronic Mail") by including:
4113 Auto-submitted: auto-generated
4115 in the messages that it generates (bounce messages and others, such as
4116 warnings). In the case of bounce messages for non-SMTP messages, there was
4117 also a typo: it was using "Auto_submitted" (underscore instead of
4118 hyphen). Since every message generated by Exim is necessarily in response
4119 to another message, thes have all been changed to:
4121 Auto-Submitted: auto-replied
4123 in accordance with these statements in the RFC:
4125 The auto-replied keyword:
4127 - SHOULD be used on messages sent in direct response to another
4128 message by an automatic process,
4130 - MUST NOT be used on manually-generated messages,
4132 - MAY be used on Delivery Status Notifications (DSNs) and Message
4133 Disposition Notifications (MDNs),
4135 - MUST NOT be used on messages generated by automatic or periodic
4136 processes, except for messages which are automatic responses to
4139 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4140 to the default Received: header definition.
4142 PH/12 Added log selector acl_warn_skipped (default on).
4144 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4145 variables because (a) they are in the wrong storage pool and (b) even if
4146 they were copied, it wouldn't work properly because of the caching.
4148 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4149 checking when decoding. Apparently there are clients that generate
4150 overlong encoded strings. Why am I not surprised?
4152 PH/15 If the first argument of "${if match_address" was not empty, but did not
4153 contain an "@" character, Exim crashed. Now it writes a panic log message
4154 and treats the condition as false.
4156 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4158 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4159 extension "envelope-auth". The code is finished and in agreement with
4160 other implementations, but there is no documentation so far and in fact,
4161 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4162 not changing the active code.
4164 Print executed "if" and "elsif" statements when debugging is used. This
4165 helps a great deal to understand what a filter does.
4167 Document more things not specified clearly in RFC3028. I had all this
4168 sorted out, when out of a sudden new issues came to my mind. Oops."
4170 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4173 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4174 canonical form (as documented). However, after a host name lookup from
4175 the IP address, check_host() was doing a simple string comparison with
4176 addresses acquired from the DNS when checking that the found name did
4177 have the original IP as one of its addresses. Since any found IPv6
4178 addresses are likely to be in abbreviated form, the comparison could
4179 fail. Luckily, there already exists a function for doing the comparison
4180 by converting both addresses to binary, so now that is used instead of
4181 the text comparison.
4183 PH/20 There was another similar case to PH/19, when a complete host name was
4184 given in a host list; looking up its IP address could give an abbreviated
4185 form, whereas the current host's name might or might not be abbreviated.
4186 The same fix has been applied.
4192 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4193 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4196 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4197 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4199 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4201 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4202 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4203 Also, for "Auto-Submitted", treat anything other than "no" as
4204 non-personal, in accordance with RFC 3834. (Previously it treated
4205 anything starting "auto-" as non-personal.)
4207 TF/01 The control=submission/name=... option had a problem with syntax
4208 errors if the name included a slash character. The /name= option
4209 now slurps the rest of the string, so it can include any characters
4210 but it must come last in the list of options (after /sender_retain
4213 PH/05 Some modifications to the interface to the fake nameserver for the new
4221 TK/01 Added the "success_on_redirect" address verification option. See
4222 NewStuff for rationale and an example.
4224 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4226 PH/02 Patch to exigrep to allow it to work on syslog lines.
4228 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4229 fread() to skip over the body file's header line, because in Cygwin the
4230 header line is locked and is inaccessible.
4232 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4233 co-exist for some time) to make it clear that it is the Exim ID that is
4234 referenced, not the Message-ID: header line.
4236 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4237 string_format() function, because snprintf() does not exist on all
4240 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4241 ${stat: expansion item.
4243 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4244 protocol synchronization error", to keep the pedants happy.
4246 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4247 well as for IRIX systems, when gcc is being used. See the host.c source
4250 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4252 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4255 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4256 transport and to the smtp transport in LMTP mode.
4258 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4260 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4261 run a filter in a subprocess. This could lead to confusion in subsequent
4262 lookups in the parent process. There should also be a search_tidyup() at
4263 the end of the subprocess.
4265 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4266 only if the host matched helo_try_verify_hosts, which caused the
4267 verification to occur when the EHLO/HELO command was issued. The ACL just
4268 tested the remembered result. Now, if a previous verification attempt has
4269 not happened, "verify = helo" does it there and then.
4271 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4273 TK/03 Fix log output including CR from clamd.
4275 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4276 exist provoked a memory error which could cause a segfault.
4278 PH/15 Installed PCRE 6.2
4280 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4282 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4283 of the problem. Specifically, suggested +O2 rather than +O1 for the
4286 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4288 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4289 record, the port was not correctly logged when the outgoing_port log
4290 selector was set (it logged the transort's default port).
4292 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4293 fallback_hosts, and "hosts" in the smtp transport.
4295 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4296 host errors such as "Connection refused".
4298 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4299 authentication with radiusclient 0.4.9:
4301 - Error returned from rc_read_config was caught wrongly
4302 - Username/password not passed on to radius server due to wrong length.
4304 The presumption is that some radiusclient API changes for 4.51/PH/17
4305 were not taken care of correctly. The code is still untested by me (my
4306 Linux distribution still has 0.3.2 of radiusclient), but it was
4307 contributed by a Radius user.
4309 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4310 expanding the "port" option of the smtp transport.
4312 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4313 while reading a MIME header. Thanks to Tom Hughes for a patch.
4315 PH/24 Include config.h inside local_scan.h so that configuration settings are
4318 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4319 that in an ACL for RCPT (for example), you can examine exactly what was
4322 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4323 commands, but it was not correctly comparing the address with the actual
4324 client host address. Thus, it would show the EHLO address in Received:
4325 header lines when this was not necessary.
4327 PH/27 Added the % operator to ${eval:}.
4329 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4330 it should be ignoring failures (because with -C, for example, it has lost
4331 privilege). It wasn't ignoring creation failures other than "already
4334 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4337 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4338 because an input line was too long, either on its own, or by virtue of
4339 too many continuations, the temporary file was not being removed, and the
4340 return code was incorrect.
4342 PH/31 Missing "BOOL" in function definition in filtertest.c.
4344 PH/32 Applied Sieve patches from the maintainer.
4346 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4348 PH/33 Added "verify = not_blind".
4350 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4351 Local/Makefile (with some defaults set). These are used in built scripts
4352 such as exicyclog, but they have never been used in the exim_install
4353 script (though there are many overriding facilities there). I have
4354 arranged that the exim_install script now takes note of these two
4357 PH/35 Installed configuration files for Dragonfly.
4359 PH/36 When a locally submitted message by a trusted user did not contain a
4360 From: header, and the sender address was obtained from -f or from an SMTP
4361 MAIL command, and the trusted user did not use -F to supply a sender
4362 name, $originator_name was incorrectly used when constructing a From:
4363 header. Furthermore, $originator_name was used for submission mode
4364 messages from external hosts without From: headers in a similar way,
4365 which is clearly wrong.
4367 PH/37 Added control=suppress_local_fixups.
4369 PH/38 When log_selector = +received_sender was set, and the addition of the
4370 sender made the log line's construction buffer exactly full, or one byte
4371 less than full, an overflow happened when the terminating "\n" was
4374 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4375 when the result of a list match is failure because a DNS lookup failed.
4377 PH/40 RM_COMMAND is now used in the building process.
4379 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4380 the "build-* directories that it finds.
4382 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4383 address in a domain literal was a prefix of an interface address.
4385 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4386 when verifying a sender address, unless rewrite_headers is false.
4388 PH/44 Wrote a long comment about why errors_to addresses are verified as
4389 recipients, not senders.
4391 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4392 the ratelimit ACL was added.
4394 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4396 PH/46 Added extra information about PostgreSQL errors to the error string.
4398 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4399 avoiding the need to install special zones in a real server. This is
4400 backwards compatible; if it can't find the fake resolver, it drops back.
4401 Thus, both old and new test suites can be run.
4403 TF/02 Added util/ratelimit.pl
4405 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4406 clock is set back in time.
4408 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4409 Candler <B.Candler@pobox.com>.
4411 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4412 OK for addresses that are the result of redirections.
4414 PH/48 A number of further additions for the benefit of the new test suite,
4415 including a fake gethostbyname() that interfaces to the fake DNS resolver
4418 TF/06 The fix for widen_domains has also been applied to qualify_single and
4419 search_parents which are the other dnslookup options that can cause
4422 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4425 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4427 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4428 service" if it encountered an MX record with an empty target hostname.
4429 The message is now "an MX or SRV record indicated no SMTP service".
4431 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4432 if the DNS of the sending site is misconfigured. This is quite a
4433 common situation. This change restores the behaviour of treating a
4434 helo verification defer as a failure.
4436 PH/51 If self=fail was set on a router, the bounce message did not include the
4437 actual error message.
4443 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4445 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4446 command itself ended in error, the underlying message about the transport
4447 filter timeout was being overwritten with the pipe command error. Now the
4448 underlying error message should be appended to the second error message.
4450 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4452 PH/02 Reduce the amount of output that "make" produces by default. Full output
4453 can still be requested.
4455 PH/03 The warning log line about a condition test deferring for a "warn" verb
4456 was being output only once per connection, rather than after each
4457 occurrence (because it was using the same function as for successful
4458 "warn" verbs). This seems wrong, so I have changed it.
4460 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4461 it should not have, which might have caused a crash in the right
4462 circumstances, but probably never did.
4464 PH/04 Installed a modified version of Tony Finch's patch to make submission
4465 mode fix the return path as well as the Sender: header line, and to
4466 add a /name= option so that you can make the user's friendly name appear
4469 TF/03 Added the control = fakedefer ACL modifier.
4471 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4472 Mark Lowes for thorough testing.
4474 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4476 TK/03 Merged latest SRS patch from Miles Wilton.
4478 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4479 with the definition in sysexits.h (which is #included earlier).
4480 Fortunately, Exim does not actually use EX_OK. The code used to try to
4481 preserve the sysexits.h value, by assuming that macro definitions were
4482 scanned for macro replacements. I have been disabused of this notion,
4483 so now the code just undefines EX_OK before #including unistd.h.
4485 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4486 in the smtp transport. When a block could not be written in a single
4487 write() function, the timeout was being re-applied to each part-write.
4488 This seems wrong - if the receiver was accepting one byte at a time it
4489 would take for ever. The timeout is now adjusted when this happens. It
4490 doesn't have to be particularly precise.
4492 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4493 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4495 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4496 without a domain if the check to <postmaster@domain> fails.
4498 SC/01 Eximstats: added -xls and the ability to specify output files
4499 (patch written by Frank Heydlauf).
4501 SC/02 Eximstats: use FileHandles for outputting results.
4503 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4505 SC/04 Eximstats: fixed display of large numbers with -nvr option
4507 SC/05 Eximstats: fixed merging of reports with empty tables.
4509 SC/06 Eximstats: added the -include_original_destination flag
4511 SC/07 Eximstats: removed tabs and trailing whitespace.
4513 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4515 TK/06 MBOX spool code: Add real "From " MBOX separator line
4516 so the .eml file is really in mbox format (even though
4517 most programs do not really care). Patch from Alex Miller.
4519 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4520 The latter is generated from $received_to and is only set if the
4521 message has one envelope recipient. SA can use these headers,
4522 obviously out-of-the-box. Patch from Alex Miller.
4524 PH/08 The ${def test on a variable was returning false if the variable's
4525 value was "0", contrary to what the specification has always said!
4526 The result should be true unless the variable is empty.
4528 PH/09 The syntax error of a character other than { following "${if
4529 def:variable_name" (after optional whitespace) was not being diagnosed.
4530 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4531 accidental colon was present, for example, could give incorrect results.
4533 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4534 result is used (not including appendfile, where other changes are about
4537 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4538 This involved changing a lot of size variables from int to off_t. It
4539 should work with maildirs and everything.
4541 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4542 spamd dying while we are connected to it.
4544 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4547 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4548 function for BDB 4.3.
4550 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4552 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4553 the sources for PCRE have been re-arranged and more files are now
4556 PH/15 The code I had for printing potentially long long variables in PH/11
4557 above was not the best (it lost precision). The length of off_t variables
4558 is now inspected at build time, and an appropriate printing format (%ld
4559 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4560 to be "long long int" or "long int". This is needed for the internal
4561 formatting function string_vformat().
4563 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4564 the configuration file to be ":syslog", then the script "guesses" where
4565 the logs files are, rather than using the compiled in default. In our
4566 case the guess is not the same as the compiled default, so the script
4567 suddenly stopped working when I started to use syslog. The patch checks
4568 to see if log_file_path is "". If so, it attempts to read it from exim
4569 with no configuration file to get the compiled in version, before it
4570 falls back to the previous guessing code."
4572 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4573 implementing BATV in an Exim configuration. See NewStuff for the gory
4576 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4577 Makefile that are specific to HP-UX.
4579 PH/18 If the "use_postmaster" option was set for a recipient callout together
4580 with the "random" option, the postmaster address was used as the MAIL
4581 FROM address for the random test, but not for the subsequent recipient
4582 test. It is now used for both.
4584 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4585 patch removes a few documentation additions to RFC 3028, because the
4586 latest draft now contains them. It adds the new en;ascii-case comparator
4587 and a new error check for 8bit text in MIME parts. Comparator and
4588 require names are now matched exactly. I enabled the subaddress
4589 extension, but it is not well tested yet (read: it works for me)."
4591 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4592 rework some of the code of TK/09 above to avoid the hardwired use of
4593 "%lld" and "long long". Replaced the call to snprintf() with a call to
4596 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4597 records point to non-existent hosts", "retry timeout exceeded", and
4598 "retry time not reached for any host after a long failure period".
4600 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4601 experimental DomainKeys support:
4603 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4604 (2) On an error such as an illegally used "control", the wrong name for
4605 the control was given.
4607 These problems did NOT occur unless DomainKeys support was compiled.
4609 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4611 PH/24 Added ${if match_ip condition.
4613 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4614 databases so that it will be absolutely obvious if a crash occurs in the
4615 DB library. This is a regular occurrence (often caused by mis-matched
4618 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4619 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4620 picked up on a user's system that detects such things. There doesn't seem
4621 to be a gcc warning option for this - only an attribute that has to be
4622 put on the function's prototype. It seems that in Fedora Core 4 they have
4623 set this on a number of new functions. No doubt there will be more in due
4626 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4627 specify a transport. However, if an address that was verified by such a
4628 router was the subject of a callout, Exim crashed because it tried to
4629 read the rcpt_include_affixes from the non-existent transport. Now it
4630 just assumes that the setting of that option is false. This bug was
4631 introduced by 4.51/PH/31.
4633 PH/28 Changed -d+all to exclude +memory, because that information is very
4634 rarely of interest, but it makes the output a lot bigger. People tend to
4635 do -d+all out of habit.
4637 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4638 code in os-type was giving problems when libc.so lives in lib64, like on
4641 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4642 aren't the modern standard, and it seems that some systems' include files
4643 don't always have them. Exim was already checking for some of the newer
4644 ones like T_AAAA, and defining it itself. I've added checks for all the
4645 record types that Exim uses.
4647 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4648 not automatically generating a new one, as it is supposed to. This
4649 prevented TLS from working. If the file did exist, but contained invalid
4650 data, a new version was generated, as expected. It was only the case of a
4651 non-existent file that was broken.
4653 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4654 with a change in libdomainkeys > 0.64.
4656 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4657 from DNS. If the selector record carries the flag, it now has
4658 precedence over the domain-wide flag.
4660 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4662 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4663 the use of a port name that isn't defined in /etc/services) occurred, the
4664 message was deferred as in a normal delivery, and thus remained on the
4665 spool, instead of being failed because of the mua_wrapper setting. This
4666 is now fixed, and I tidied up some of the mua_wrapper messages at the
4669 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4670 the messages in a hash of arrays rather than using individual hashes.
4671 This is a bit cleaner and results in dramatic memory savings, albeit
4672 at a slight CPU cost.
4674 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4675 as requested by Marc Sherman.
4677 SC/10 Eximstats: added histograms for user specified patterns as requested
4680 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4682 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4683 fopen() in the content-scanning modules that did not already have it.
4689 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4690 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4692 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4694 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4696 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4697 to test Sieve filters that use "vacation".
4699 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4700 that changes the way the GnuTLS parameters are stored in the cache file.
4701 The new format can be generated externally. For backward compatibility,
4702 if the data in the cache doesn't make sense, Exim assumes it has read an
4703 old-format file, and it generates new data and writes a new file. This
4704 means that you can't go back to an older release without removing the
4707 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4708 work if there are any delivery delays because "one_time" forces the
4709 parent to be marked "delivered", so its unseen clone is never tried
4710 again. For this reason, Exim now forbids the simultaneous setting of
4713 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4714 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4715 entry if you want to know the details. The fix, however, neglected to
4716 consider the case when local delivery batching is involved. The test for
4717 "previously delivered" was not happening when checking to see if an
4718 address could be batched with a previous (undelivered) one; under
4719 certain circumstances this could lead to multiple deliveries to the same
4722 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4723 in its include files, and this causes problems building Exim.
4725 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4726 header_syntax) but Exim was just ignoring anything given after a slash.
4727 In particular, this caused confusion with an attempt to use "verify =
4728 reverse_host_lookup/defer_ok". An error is now given when options are
4729 supplied for verify items that do not have them. (Maybe reverse_host_
4730 lookup should have a defer_ok option, but that's a different point.)
4732 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4733 defined by RFC 821) to 2048, because there were problems with some AUTH
4734 commands, and RFC 1869 says the size should be increased for extended
4735 SMTP commands that take arguments.
4737 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4740 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4741 "unknown" error; now it says that the functionality isn't in the binary.
4743 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4744 an address' error message when a string expansion fails (syntax or
4745 whatever). Otherwise the password may appear in the log. Following change
4746 PH/42 below, there is no longer a chance of it appearing in a bounce
4749 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4751 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4752 of its addresses was ever tried. (Bugzilla bug #2.)
4754 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4755 the result incorrectly in the debug output. (It correctly added a newline
4756 to what was transported.)
4758 TF/01 Added $received_time.
4760 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4761 commented out examples of how to interface to a virus scanner and to
4762 SpamAssassin. Also added commented examples of av_scanner and
4763 spamd_address settings.
4765 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4766 and controls are allowed in which ACLs. There were a couple of minor
4767 errors. Some of the entries in the conditions table (which is a table of
4768 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4769 negation of where the condition IS allowed.
4771 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4773 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4774 header file does not have a version number, so I've had to invent a new
4775 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4776 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4777 radiusclient), but it was contributed by a Radius user.
4779 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4780 files or directories, for OpenSSL.
4782 PH/19 When an Exim process that is running as root has to create an Exim log
4783 file, it does so in a subprocess that runs as exim:exim so as to get the
4784 ownership right at creation (otherwise, other Exim processes might see
4785 the file with the wrong ownership). There was no test for failure of this
4786 fork() call, which would lead to the process getting stuck as it waited
4787 for a non-existent subprocess. Forks do occasionally fail when resources
4788 run out. I reviewed all the other calls to fork(); they all seem to check
4791 PH/20 When checking for unexpected SMTP input at connect time (before writing
4792 the banner), Exim was not dealing correctly with a non-positive return
4793 from the read() function. If the client had disconnected by this time,
4794 the result was a log entry for a synchronization error with an empty
4795 string after "input=" when read() returned zero. If read() returned -1
4796 (an event I could not check), uninitialized data bytes were printed.
4797 There were reports of junk text (parts of files, etc) appearing after
4800 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4802 PH/22 Added support for macro redefinition, and (re)definition in between
4803 driver and ACL definitions.
4805 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4806 forgetting to use the resulting value; it was using the unexpanded value.
4808 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4809 hadn't been configured. The fix is from Juergen Kreileder, who
4810 understands it better than I do:
4812 "Here's what I see happening with three configured cyrus_sasl
4813 authenticators configured (plain, login, cram-md5):
4815 On startup auth_cyrus_sasl_init() gets called for each of these.
4816 This means three calls to sasl_listmech() without a specified mech_list.
4817 => SASL tests which mechs of all available mechs actually work
4818 => three warnings about OTP not working
4819 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4821 With the patch, sasl_listmech() also gets called three times. But now
4822 SASL's mech_list option is set to the server_mech specified in the the
4823 authenticator. Or in other words, the answer from sasl_listmech()
4824 gets limited to just the mech you're testing for (which is different
4826 => the return list contains just 'plain' or 'login', 'cram-md5' or
4827 nothing depending on the value of ob->server_mech.
4829 I've just tested the patch: Authentication still works fine,
4830 unavailable mechs specified in the exim configuration are still
4831 caught, and the auth.log warnings about OTP are gone."
4833 PH/25 When debugging is enabled, the contents of the command line are added
4834 to the debugging output, even when log_selector=+arguments is not
4837 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4838 answer is "GNU", and only if the return is "GNU/something" is the answer
4841 PH/27 $acl_verify_message is now set immediately after the failure of a
4842 verification in an ACL, and so is available in subsequent modifiers. In
4843 particular, the message can be preserved by coding like this:
4845 warn !verify = sender
4846 set acl_m0 = $acl_verify_message
4848 Previously, $acl_verify_message was set only while expanding "message"
4849 and "log_message" when a very denied access.
4851 PH/28 Modified OS/os.c-Linux with
4853 -#ifndef OS_LOAD_AVERAGE
4854 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4856 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4857 nomenclature these days.)
4859 PH/29 Installed patch from the Sieve maintainer that adds the options
4860 sieve_useraddress and sieve_subaddress to the redirect router.
4862 PH/30 In these circumstances:
4863 . Two addresses routed to the same list of hosts;
4864 . First host does not offer TLS;
4865 . First host accepts first address;
4866 . First host gives temporary error to second address;
4867 . Second host offers TLS and a TLS session is established;
4868 . Second host accepts second address.
4869 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4870 and peerdn, if requested) that were in fact used only for the second
4873 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4874 attention to any local part prefix or suffix that was matched by the
4875 router that accepted the address. It now behaves in the same way as it
4876 does for delivery: the affixes are removed from the local part unless
4877 rcpt_include_affixes is set on the transport.
4879 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4880 timeout during the DATA phase of an incoming message.
4882 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4883 applied a patch sanctioned by the Sieve maintainer.
4885 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4886 the uid or gid is negative. A case of a negative gid caused this to be
4887 noticed. The fix allows for either to be negative.
4889 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4890 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4893 PH/36 The change PH/12 above was broken. Fixed it.
4895 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4896 the grounds that routing the same address twice would always produce the
4897 same answer. This might have been true once, but it is certainly no
4898 longer true now. Routing a child address may depend on the previous
4899 routing that produced that child. Some complicated redirection strategies
4900 went wrong when messages had multiple recipients, and made Exim's
4901 behaviour dependent on the order in which the addresses were given.
4903 I have moved the duplicate checking until after the routing is complete.
4904 Exim scans the addresses that are assigned to local and remote
4905 transports, and removes any duplicates. This means that more work will be
4906 done, as duplicates will always all be routed, but duplicates are
4907 presumably rare, so I don't expect this is of any significance.
4909 For deliveries to pipes, files, and autoreplies, the duplicate checking
4910 still happens during the routing process, since they are not going to be
4913 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4914 It corrects a timeout issue with spamd. This is Ian's comment: "The
4915 background is that sometimes spamd either never reads data from a
4916 connection it has accepted, or it never writes response data. The exiscan
4917 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4918 blindly assumes that writes won't block so it may never time out."
4920 PH/39 Allow G after quota size as well as K and M.
4922 PH/40 The value set for $authenticated_id in an authenticator may not contain
4923 binary zeroes or newlines because the value is written to log lines and
4924 to spool files. There was no check on this. Now the value is run through
4925 the string_printing() function so that such characters are converted to
4926 printable escape sequences.
4928 PH/41 $message_linecount is a new variable that contains the total number of
4929 lines in the message. Compare $body_linecount, which is the count for the
4932 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4933 bounce and delay warning messages, except in certain special cases, which
4936 (a) An SMTP error message from a remote host;
4937 (b) A message specified in a :fail: redirection;
4938 (c) A message specified in a "fail" command in a system filter;
4939 (d) A message specified in a FAIL return from the queryprogram router;
4940 (e) A message specified by the cannot_route_message router option.
4942 In these cases only, Exim does include the error details in bounce and
4943 warning messages. There are also a few cases where bland messages such
4944 as "unrouteable address" or "local delivery error" are given.
4946 PH/43 $value is now also set for the "else" part of a ${run expansion.
4948 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4949 being worked on, but at least Exim now implements the latest version to
4952 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4953 process to complete was treated as a delivery failure, a timeout while
4954 writing the message to the pipe was logged, but erroneously treated as a
4955 successful delivery. Such timeouts include transport filter timeouts. For
4956 consistency with the overall process timeout, these timeouts are now
4957 treated as errors, giving rise to delivery failures by default. However,
4958 there is now a new Boolean option for the pipe transport called
4959 timeout_defer, which, if set TRUE, converts the failures into defers for
4960 both kinds of timeout. A transport filter timeout is now identified in
4963 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4964 systems where "make" and "gmake" are different, calling "gmake" at top
4965 level broke things. I've arranged for the value of $(MAKE) to be passed
4966 from the Makefile to this script so that it can call the same version of
4970 A note about Exim versions 4.44 and 4.50
4971 ----------------------------------------
4973 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4974 changes of various kinds. As a consequence, a big documentation update was
4975 needed. This delayed the release for rather longer than seemed good, especially
4976 in the light of a couple of (minor) security issues. Therefore, the changes
4977 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4978 release. So 4.44 and 4.50 are in effect two different branches that both start
4981 I have left the 4.50 change log unchanged; it contains all the changes since
4982 4.43. The change log for 4.44 is below; many of its items are identical to
4983 those for 4.50. This seems to be the most sensible way to preserve the
4984 historical information.
4990 1. Minor wording change to the doc/README.SIEVE file.
4992 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4993 computation of the current number of files was incorrect.
4995 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4996 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4999 4. Give more explanation in the error message when the command for a transport
5000 filter fails to execute.
5002 5. There are several places where Exim runs a non-Exim command in a
5003 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5004 was being done only for the command run by the queryprogram router. It is
5005 now done for all such subprocesses. The other cases are: ${run, transport
5006 filters, and the commands run by the lmtp and pipe transports.
5008 6. Added CONFIGURE_GROUP build-time option.
5010 7. Some older OS have a limit of 256 on the maximum number of file
5011 descriptors. Exim was using setrlimit() to set 1000 as a large value
5012 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5013 systems. I've change it so that if it can't get 1000, it tries for 256.
5015 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5016 was an oversight, and furthermore, ever since the addition of extra
5017 controls (e.g. 4.43/32), the checks on when to allow different forms of
5018 "control" were broken. There should now be diagnostics for all cases when a
5019 control that does not make sense is encountered.
5021 9. Added the /retain_sender option to "control=submission".
5023 10. $recipients is now available in the predata ACL (oversight).
5025 11. Tidy the search cache before the fork to do a delivery from a message
5026 received from the command line. Otherwise the child will trigger a lookup
5027 failure and thereby defer the delivery if it tries to use (for example) a
5028 cached ldap connection that the parent has called unbind on.
5030 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5031 of $address_data from the recipient verification was clobbered by the
5032 sender verification.
5034 13. The value of address_data from a sender verification is now available in
5035 $sender_address_data in subsequent conditions in the ACL statement.
5037 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5039 15. Added a new option "connect=<time>" to callout options, to set a different
5042 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5043 was its contents. (It was OK if the option was not defined at all.)
5045 17. A "Completed" log line is now written for messages that are removed from
5046 the spool by the -Mrm option.
5048 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5049 information about exactly what failed.
5051 19. Added -dd to debug only the daemon process.
5053 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5054 handles timeouts, both on the server side and network timeouts. Renamed the
5055 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5057 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5058 It is now set to "smtps".
5060 22. $host_address is now set to the target address during the checking of
5061 ignore_target_hosts.
5063 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5064 being passed; this would have caused $sender_host_name to have been used if
5065 matching the list had actually called for a host name (not very likely,
5066 since this list is usually IP addresses). A host name is now passed as
5069 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5070 code that specifies a non-restarting handler (typically sigaction() in
5071 modern systems) in an attempt to fix a rare and obscure crash bug.
5073 25. Narrowed the window for a race in the daemon that could cause it to ignore
5074 SIGCHLD signals. This is not a major problem, because they are used only to
5075 wake it up if nothing else does.
5077 26. A malformed maildirsize file could cause Exim to calculate negative values
5078 for the mailbox size or file count. Odd effects could occur as a result.
5079 The maildirsize information is now recalculated if the size or filecount
5082 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5083 support for a long time. Removed HAVE_SYS_VFS_H.
5085 28. Installed the latest version of exipick from John Jetmore.
5087 29. In an address list, if the pattern was not a regular expression, an empty
5088 subject address (from a bounce message) matched only if the pattern was an
5089 empty string. Non-empty patterns were not even tested. This was the wrong
5090 because it is perfectly reasonable to use an empty address as part of a
5091 database query. An empty address is now tested by patterns that are
5092 lookups. However, all the other forms of pattern expect the subject to
5093 contain a local part and a domain, and therefore, for them, an empty
5094 address still always fails if the pattern is not itself empty.
5096 30. Exim went into a mad DNS loop when attempting to do a callout where the
5097 host was specified on an smtp transport, and looking it up yielded more
5098 than one IP address.
5100 31. Re-factored the code for checking spool and log partition space into a
5101 function that finds that data and another that does the check. The former
5102 is then used to implement four new variables: $spool_space, $log_space,
5103 $spool_inodes, and $log_inodes.
5105 32. The RFC2047 encoding function was originally intended for short strings
5106 such as real names; it was not keeping to the 75-character limit for
5107 encoded words that the RFC imposes. It now respects the limit, and
5108 generates multiple encoded words if necessary. To be on the safe side, I
5109 have increased the buffer size for the ${rfc2047: expansion operator from
5112 33. It is now permitted to omit both strings after an "if" condition; if the
5113 condition is true, the result is "true". As before, when the second string
5114 is omitted, a false condition yields an empty string. This makes it less
5115 cumbersome to write custom ACL and router conditions.
5117 34. Failure to deliver a bounce message always caused it to be frozen, even if
5118 there was an errors_to setting on the router. The errors_to setting is now
5121 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5122 canonical form (fully expanded) before being placed in
5123 $sender_host_address.
5125 36. The table in the code that translates DNS record types into text (T_A to
5126 "A" for instance) was missing entries for NS and CNAME. It is just possible
5127 that this could have caused confusion if both these types were looked up
5128 for the same domain, because the text type is used as part of Exim's
5129 per-process caching. But the chance of anyone hitting this buglet seems
5132 37. The dnsdb lookup has been extended in a number of ways.
5134 (1) There is a new type, "zns", which walks up the domain tree until it
5135 finds some nameserver records. It should be used with care.
5137 (2) There is a new type, "mxh", which is like "mx" except that it returns
5138 just the host names, not the priorities.
5140 (3) It is now possible to give a list of domains (or IP addresses) to be
5141 looked up. The behaviour when one of the lookups defers can be
5142 controlled by a keyword.
5144 (4) It is now possible to specify the separator character for use when
5145 multiple records are returned.
5147 38. The dnslists ACL condition has been extended: it is now possible to supply
5148 a list of IP addresses and/or domains to be looked up in a particular DNS
5151 39. Added log_selector=+queue_time_overall.
5153 40. When running the queue in the test harness, wait just a tad after forking a
5154 delivery process, to get repeatability of debugging output.
5156 41. Include certificate and key file names in error message when GnuTLS fails
5157 to set them up, because the GnuTLS error message doesn't include the name
5158 of the failing file when there is a problem reading it.
5160 42. Allow both -bf and -bF in the same test run.
5162 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5164 44. The "Exiscan patch" is now merged into the mainline Exim source.
5166 45. Sometimes the final signoff response after QUIT could fail to get
5167 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5168 < 0 before doing a fflush(). This bug looks as though it goes back to the
5169 introduction of TLS in release 3.20, but "sometimes" must have been rare
5170 because the tests only now provoked it.
5172 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5173 (this can affect the format of dates).
5175 47. exim_tidydb, when checking for the continued existence of a message for
5176 which it has found a message-specific retry record, was not finding
5177 messages that were in split spool directories. Consequently, it was
5178 deleting retry records that should have stayed in existence.
5180 48. Steve fixed some bugs in eximstats.
5182 49. The SPA authentication driver was not abandoning authentication and moving
5183 on to the next authenticator when an expansion was forced to fail,
5184 contradicting the general specification for all authenticators. Instead it
5185 was generating a temporary error. It now behaves as specified.
5187 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5188 (the order specifies the preference for clients). The order is now AES256,
5189 AES128, 3DES, ARCFOUR128.
5191 51. Small patch to Sieve code - explicitly set From: when generating an
5194 52. Exim crashed if a remote delivery caused a very long error message to be
5195 recorded - for instance if somebody sent an entire SpamAssassin report back
5196 as a large number of 550 error lines. This bug was coincidentally fixed by
5197 increasing the size of one of Exim's internal buffers (big_buffer) that
5198 happened as part of the Exiscan merge. However, to be on the safe side, I
5199 have made the code more robust (and fixed the comments that describe what
5202 53. Now that there can be additional text after "Completed" in log lines (if
5203 the queue_time_overall log selector is set), a one-byte patch to exigrep
5204 was needed to allow it to recognize "Completed" as not the last thing in
5207 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5208 patch that reportedly fixes this has been added. I am not expert enough to
5209 create a test for it. This is what the patch creator wrote:
5211 "I found a little strange behaviour of ldap code when working with
5212 Windows 2003 AD Domain, where users was placed in more than one
5213 Organization Units. When I tried to give exim partial DN, the exit code
5214 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5215 But simultaneously result of request was absolutely normal ldap result,
5216 so I produce this patch..."
5218 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5219 REFERENCE, so I have modified the code to exclude the patch when that macro
5222 55. Some experimental protocols are using DNS PTR records for new purposes. The
5223 keys for these records are domain names, not reversed IP addresses. The
5224 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5225 leaves it alone. Component reversal etc. now happens only for IP addresses.
5228 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5230 57. Double the size of the debug message buffer (to 2048) so that more of very
5231 long debug lines gets shown.
5233 58. The exicyclog utility now does better if the number of log files to keep
5234 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5236 59. Two changes related to the smtp_active_hostname option:
5238 (1) $smtp_active_hostname is now available as a variable.
5239 (2) The default for smtp_banner uses $smtp_active_hostname instead
5240 of $primary_hostname.
5242 60. The host_aton() function is supposed to be passed a string that is known
5243 to be a valid IP address. However, in the case of IPv6 addresses, it was
5244 not checking this. This is a hostage to fortune. Exim now panics and dies
5245 if the condition is not met. A case was found where this could be provoked
5246 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5247 components; fortuitously, this particular loophole had already been fixed
5248 by change 4.50/55 above.
5250 If there are any other similar loopholes, the new check in host_aton()
5251 itself should stop them being exploited. The report I received stated that
5252 data on the command line could provoke the exploit when Exim was running as
5253 exim, but did not say which command line option was involved. All I could
5254 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5255 running as the user.
5258 61. There was a buffer overflow vulnerability in the SPA authentication code
5259 (which came originally from the Samba project). I have added a test to the
5260 spa_base64_to_bits() function which I hope fixes it.
5263 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5264 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5266 63. The daemon start-up calls getloadavg() while still root for those OS that
5267 need the first call to be done as root, but it missed one case: when
5268 deliver_queue_load_max is set with deliver_drop_privilege. This is
5269 necessary for the benefit of the queue runner, because there is no re-exec
5270 when deliver_drop_privilege is set.
5272 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5273 This has been fixed.
5275 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5276 was in use, was not putting the data itself into the right store pool;
5277 consequently, it could be overwritten for a subsequent message in the same
5278 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5281 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5283 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5284 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5285 the function were treating the return as a boolean value, which happened to
5286 work because 0=false and not-0=true, but is not correct code.
5288 68. The host_aton() function was not handling scoped IPv6 addresses (those
5289 with, for example, "%eth0" on the end) correctly.
5291 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5292 negated items (that is, ~something) in unsigned ints. Some compilers
5293 apparently mutter when there is no cast.
5295 70. If an address verification called from an ACL failed, and did not produce a
5296 user-specific message (i.e. there was only a "system" message), nothing was
5297 put in $acl_verify_message. In this situation, it now puts the system
5300 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5301 session; change 4.31/43 added the unwanted input to the log line - except
5302 that it did not do this in the start of session case. It now does.
5304 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5305 This is wrong and can cause the other end to generate a synchronization
5306 error if it is another Exim or anything else that does the synchronization
5307 check. A QUIT command is no longer sent after a timeout.
5309 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5310 during host lookups.
5312 74. The defer_ok option of callout verification was not working if it was used
5313 when verifying addresses in header lines, that is, for this case:
5315 verify = header_sender/callout=defer_ok
5317 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5318 those file descriptors could be used for SMTP connections. If anything
5319 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5320 could be sent to the SMTP client, causing chaos. The daemon now opens
5321 stdin, stdout, and stderr to /dev/null when it puts itself into the
5324 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5325 log by default. The user can override this with suitable Perl magic.
5327 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5328 the log message when discard triggers, was not working for the DATA ACL or
5329 for the non-SMTP ACL.
5331 78. Error message wording change in sieve.c.
5333 79. If smtp_accept_max_per_host was set, the number of connections could be
5334 restricted to fewer than expected, because the daemon was trying to set up
5335 a new connection before checking whether the processes handling previous
5336 connections had finished. The check for completed processes is now done
5337 earlier. On busy systems, this bug wouldn't be noticed because something
5338 else would have woken the daemon, and it would have reaped the completed
5341 80. If a message was submitted locally by a user whose login name contained one
5342 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5343 It caused a spool format error. I have fixed the spool reading code. A
5344 related problem was that the "from" clause in the Received: line became
5345 illegal because of the space(s). It is now covered by ${quote_local_part.
5347 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5350 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5352 83. Patch from Sieve maintainer to fix unterminated string problem in
5353 "vacation" handling.
5355 84. Some minor changes to the Linux configuration files to help with other
5356 OS variants using glibc.
5358 85. One more patch for Sieve to update vacation handling to latest spec.
5361 ----------------------------------------------------
5362 See the note above about the 4.44 and 4.50 releases.
5363 ----------------------------------------------------
5369 1. Change 4.43/35 introduced a bug that caused file counts to be
5370 incorrectly computed when quota_filecount was set in an appendfile
5373 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5374 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5377 3. Give more explanation in the error message when the command for a transport
5378 filter fails to execute.
5380 4. There are several places where Exim runs a non-Exim command in a
5381 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5382 was being done only for the command run by the queryprogram router. It is
5383 now done for all such subprocesses. The other cases are: ${run, transport
5384 filters, and the commands run by the lmtp and pipe transports.
5386 5. Some older OS have a limit of 256 on the maximum number of file
5387 descriptors. Exim was using setrlimit() to set 1000 as a large value
5388 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5389 systems. I've change it so that if it can't get 1000, it tries for 256.
5391 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5392 was an oversight, and furthermore, ever since the addition of extra
5393 controls (e.g. 4.43/32), the checks on when to allow different forms of
5394 "control" were broken. There should now be diagnostics for all cases when a
5395 control that does not make sense is encountered.
5397 7. $recipients is now available in the predata ACL (oversight).
5399 8. Tidy the search cache before the fork to do a delivery from a message
5400 received from the command line. Otherwise the child will trigger a lookup
5401 failure and thereby defer the delivery if it tries to use (for example) a
5402 cached ldap connection that the parent has called unbind on.
5404 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5405 of $address_data from the recipient verification was clobbered by the
5406 sender verification.
5408 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5409 was its contents. (It was OK if the option was not defined at all.)
5411 11. A "Completed" log line is now written for messages that are removed from
5412 the spool by the -Mrm option.
5414 12. $host_address is now set to the target address during the checking of
5415 ignore_target_hosts.
5417 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5418 being passed; this would have caused $sender_host_name to have been used if
5419 matching the list had actually called for a host name (not very likely,
5420 since this list is usually IP addresses). A host name is now passed as
5423 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5424 code that specifies a non-restarting handler (typically sigaction() in
5425 modern systems) in an attempt to fix a rare and obscure crash bug.
5427 15. Narrowed the window for a race in the daemon that could cause it to ignore
5428 SIGCHLD signals. This is not a major problem, because they are used only to
5429 wake it up if nothing else does.
5431 16. A malformed maildirsize file could cause Exim to calculate negative values
5432 for the mailbox size or file count. Odd effects could occur as a result.
5433 The maildirsize information is now recalculated if the size or filecount
5436 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5437 support for a long time. Removed HAVE_SYS_VFS_H.
5439 18. Updated exipick to current release from John Jetmore.
5441 19. Allow an empty sender to be matched against a lookup in an address list.
5442 Previously the only cases considered were a regular expression, or an
5445 20. Exim went into a mad DNS lookup loop when doing a callout where the
5446 host was specified on the transport, if the DNS lookup yielded more than
5449 21. The RFC2047 encoding function was originally intended for short strings
5450 such as real names; it was not keeping to the 75-character limit for
5451 encoded words that the RFC imposes. It now respects the limit, and
5452 generates multiple encoded words if necessary. To be on the safe side, I
5453 have increased the buffer size for the ${rfc2047: expansion operator from
5456 22. Failure to deliver a bounce message always caused it to be frozen, even if
5457 there was an errors_to setting on the router. The errors_to setting is now
5460 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5461 canonical form (fully expanded) before being placed in
5462 $sender_host_address.
5464 24. Updated eximstats to version 1.33
5466 25. Include certificate and key file names in error message when GnuTLS fails
5467 to set them up, because the GnuTLS error message doesn't include the name
5468 of the failing file when there is a problem reading it.
5470 26. Expand error message when OpenSSL has problems setting up cert/key files.
5473 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5474 (this can affect the format of dates).
5476 28. exim_tidydb, when checking for the continued existence of a message for
5477 which it has found a message-specific retry record, was not finding
5478 messages that were in split spool directories. Consequently, it was
5479 deleting retry records that should have stayed in existence.
5481 29. eximstats updated to version 1.35
5482 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5483 1.35 - bugfix such that pie charts by volume are generated correctly
5485 30. The SPA authentication driver was not abandoning authentication and moving
5486 on to the next authenticator when an expansion was forced to fail,
5487 contradicting the general specification for all authenticators. Instead it
5488 was generating a temporary error. It now behaves as specified.
5490 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5491 (the order specifies the preference for clients). The order is now AES256,
5492 AES128, 3DES, ARCFOUR128.
5494 31. Small patch to Sieve code - explicitly set From: when generating an
5497 32. Exim crashed if a remote delivery caused a very long error message to be
5498 recorded - for instance if somebody sent an entire SpamAssassin report back
5499 as a large number of 550 error lines. This bug was coincidentally fixed by
5500 increasing the size of one of Exim's internal buffers (big_buffer) that
5501 happened as part of the Exiscan merge. However, to be on the safe side, I
5502 have made the code more robust (and fixed the comments that describe what
5505 33. Some experimental protocols are using DNS PTR records for new purposes. The
5506 keys for these records are domain names, not reversed IP addresses. The
5507 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5508 leaves it alone. Component reversal etc. now happens only for IP addresses.
5511 34. The host_aton() function is supposed to be passed a string that is known
5512 to be a valid IP address. However, in the case of IPv6 addresses, it was
5513 not checking this. This is a hostage to fortune. Exim now panics and dies
5514 if the condition is not met. A case was found where this could be provoked
5515 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5516 components; fortuitously, this particular loophole had already been fixed
5517 by change 4.50/55 or 4.44/33 above.
5519 If there are any other similar loopholes, the new check in host_aton()
5520 itself should stop them being exploited. The report I received stated that
5521 data on the command line could provoke the exploit when Exim was running as
5522 exim, but did not say which command line option was involved. All I could
5523 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5524 running as the user.
5527 35. There was a buffer overflow vulnerability in the SPA authentication code
5528 (which came originally from the Samba project). I have added a test to the
5529 spa_base64_to_bits() function which I hope fixes it.
5532 36. The daemon start-up calls getloadavg() while still root for those OS that
5533 need the first call to be done as root, but it missed one case: when
5534 deliver_queue_load_max is set with deliver_drop_privilege. This is
5535 necessary for the benefit of the queue runner, because there is no re-exec
5536 when deliver_drop_privilege is set.
5538 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5539 was in use, was not putting the data itself into the right store pool;
5540 consequently, it could be overwritten for a subsequent message in the same
5541 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5544 38. Sometimes the final signoff response after QUIT could fail to get
5545 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5546 < 0 before doing a fflush(). This bug looks as though it goes back to the
5547 introduction of TLS in release 3.20, but "sometimes" must have been rare
5548 because the tests only now provoked it.
5554 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5555 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5556 Now it returns an integer. A number of calls were still expecting a T/F
5557 return. Fortuitously, in all cases, the tests worked in OK situations,
5558 which is the norm. However, things would have gone wrong on any write
5559 failures on the smtp file descriptor. This function is used when sending
5560 messages over SMTP and also when doing verify callouts.
5562 2. When Exim is called to do synchronous delivery of a locally submitted
5563 message (the -odf or -odi options), it no longer closes stderr before doing
5566 3. Implemented the mua_wrapper option.
5568 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5570 5. Implemented the functions header_remove(), header_testname(),
5571 header_add_at_position(), and receive_remove_recipient(), and exported them
5574 6. If an ACL "warn" statement specified the addition of headers, Exim already
5575 inserted X-ACL-Warn: at the start if there was no header name. However, it
5576 was not making this test for the second and subsequent header lines if
5577 there were newlines in the string. This meant that an invalid header could
5578 be inserted if Exim was badly configured.
5580 7. Allow an ACL "warn" statement to add header lines at the start or after all
5581 the Received: headers, as well as at the end.
5583 8. Added the rcpt_4xx retry error code.
5585 9. Added postmaster_mailfrom=xxx to callout verification option.
5587 10. Added mailfrom=xxxx to the callout verification option, for verify=
5590 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5591 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5593 12. Inserted some casts to stop certain compilers warning when using pointer
5594 differences as field lengths or precisions in printf-type calls (mostly
5595 affecting debugging statements).
5597 13. Added optional readline() support for -be (dynamically loaded).
5599 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5600 same clock tick as a message's arrival, so that its received time was the
5601 same as the "first fail" time on the retry record, and that message
5602 remained on the queue past the ultimate address timeout, every queue runner
5603 would try a delivery (because it was past the ultimate address timeout) but
5604 after another failure, the ultimate address timeout, which should have then
5605 bounced the address, did not kick in. This was a "< instead of <=" error;
5606 in most cases the first failure would have been in the next clock tick
5607 after the received time, and all would be well.
5609 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5610 being recognized when the domain list was tested by the match_domain
5611 condition in an expansion string.
5613 16. Added the ${str2b64: operator.
5615 17. Exim was always calling setrlimit() to set a large limit for the number of
5616 processes, without checking whether the existing limit was already
5617 adequate. (It did check for the limit on file descriptors.) Furthermore,
5618 errors from getrlimit() and setrlimit() were being ignored. Now they are
5619 logged to the main and panic logs, but Exim does carry on, to try to do its
5620 job under whatever limits there are.
5622 18. Imported PCRE 5.0.
5624 19. Trivial typo in log message " temporarily refused connection" (the leading
5627 20. If the log selector return_path_on_delivery was set and an address was
5628 redirected to /dev/null, the delivery process crashed because it assumed
5629 that a return path would always be set for a "successful" delivery. In this
5630 case, the whole delivery is bypassed as an optimization, and therefore no
5633 21. Internal re-arrangement: the function for sending a challenge and reading
5634 a response while authentication was assuming a zero-terminated challenge
5635 string. It's now changed to take a pointer and a length, to allow for
5636 binary data in such strings.
5638 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5640 23. Exim was not respecting finduser_retries when seeking the login of the
5641 uid under which it was called; it was always trying 10 times. (The default
5642 setting of finduser_retries is zero.) Also, it was sleeping after the final
5643 failure, which is pointless.
5645 24. Implemented tls_on_connect_ports.
5647 25. Implemented acl_smtp_predata.
5649 26. If the domain in control=submission is set empty, Exim assumes that the
5650 authenticated id is a complete email address when it generates From: or
5651 Sender: header lines.
5653 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5654 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5655 chown and chgrp in /bin and hostname in /usr/bin.
5657 28. Exim was keeping the "process log" file open after each use, just as it
5658 does for the main log. This opens the possibility of it remaining open for
5659 long periods when the USR1 signal hits a daemon. Occasional processlog
5660 errors were reported, that could have been caused by this. Anyway, it seems
5661 much more sensible not to leave this file open at all, so that is what now
5664 29. The long-running daemon process does not normally write to the log once it
5665 has entered its main loop, and it closes the log before doing so. This is
5666 so that log files can straightforwardly be renamed and moved. However,
5667 there are a couple of unusual error situations where the daemon does write
5668 log entries, and I had neglected to close the log afterwards.
5670 30. The text of an SMTP error response that was received during a remote
5671 delivery was being truncated at 512 bytes. This is too short for some of
5672 the long messages that one sometimes sees. I've increased the limit to
5675 31. It is now possible to make retry rules that apply only when a message has a
5676 specific sender, in particular, an empty sender.
5678 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5679 it possible to be selective about when SMTP synchronization is enforced.
5681 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5683 32. Implemented hosts_connection_nolog.
5685 33. Added an ACL for QUIT.
5687 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5690 35. Added mailbox_size and mailbox_filecount to appendfile.
5692 36. Added control = no_multiline_responses to ACLs.
5694 37. There was a bug in the logic of the code that waits for the clock to tick
5695 in the case where the clock went backwards by a substantial amount such
5696 that the microsecond fraction of "now" was more than the microsecond
5697 fraction of "then" (but the whole seconds number was less).
5699 38. Added support for the libradius Radius client library this is found on
5700 FreeBSD (previously only the radiusclient library was supported).
5706 1. When certain lookups returned multiple values in the form name=value, the
5707 quoting of the values was not always being done properly. Specifically:
5708 (a) If the value started with a double quote, but contained no whitespace,
5710 (b) If the value contained whitespace other than a space character (i.e.
5711 tabs or newlines or carriage returns) it was not quoted.
5712 This fix has been applied to the mysql and pgsql lookups by writing a
5713 separate quoting function and calling it from the lookup code. The fix
5714 should probably also be applied to nisplus, ibase and oracle lookups, but
5715 since I cannot test any of those, I have not disturbed their existing code.
5717 2. A hit in the callout cache for a specific address caused a log line with no
5718 reason for rejecting RCPT. Now it says "Previous (cached) callout
5719 verification failure".
5721 3. There was an off-by-one bug in the queryprogram router. An over-long
5722 return line was truncated at 256 instead of 255 characters, thereby
5723 overflowing its buffer with the terminating zero. As well as fixing this, I
5724 have increased the buffer size to 1024 (and made a note to document this).
5726 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5727 when Exim is waiting for an SMTP response from a remote server, Exim
5728 restarts its select() call on the socket, thereby resetting its timeout.
5729 This is not a problem when such interrupts are rare. Somebody set up a cron
5730 job to run exiwhat every 2 minutes, which is less than the normal select()
5731 timeout (5 or 10 minutes). This meant that the select() timeout never
5732 kicked in because it was always reset. I have fixed this by comparing the
5733 time when an interrupt arrives with the time at the start of the first call
5734 to select(). If more time than the timeout has elapsed, the interrupt is
5735 treated as a timeout.
5737 5. Some internal re-factoring in preparation for the addition of Sieve
5738 extensions (by MH). In particular, the "personal" test is moved to a
5739 separate function, and given an option for scanning Cc: and Bcc: (which is
5740 not set for Exim filters).
5742 6. When Exim created an email address using the login of the caller as the
5743 local part (e.g. when creating a From: or Sender: header line), it was not
5744 quoting the local part when it contained special characters such as @.
5746 7. Installed new OpenBSD configuration files.
5748 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5749 try to make them clearer.
5751 9. Callout options, other than the timeout value, were being ignored when
5752 verifying sender addresses in header lines. For example, when using
5754 verify = header_sender/callout=no_cache
5756 the cache was (incorrectly) being used.
5758 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5759 only the Cygwin environment.
5761 11. When return_path_on_delivery was set as a log selector, if different remote
5762 addresses in the same message used different return paths and parallel
5763 remote delivery occurred, the wrong values would sometimes be logged.
5764 (Whenever a remote delivery process finished, the return path value from
5765 the most recently started remote delivery process was logged.)
5767 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5768 lines when AUTH and/or TLS are in use. This is the "received protocol"
5769 field. Exim used to use "asmtp" for authenticated SMTP, without any
5770 indication (in the protocol name) for TLS use. Now it follows the RFC and
5771 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5772 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5773 names appear in log lines as well as in Received: header lines.
5775 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5776 extensions, and comparison tests, and to fix some bugs.
5778 14. Changes to the "personal" filter test:
5780 (1) The test was buggy in that it was just doing the equivalent of
5781 "contains" tests on header lines. For example, if a user's address was
5782 anne@some.where, the "personal" test would incorrectly be true for
5784 To: susanne@some.where
5786 This test is now done by extracting each address from the header in turn,
5787 and checking the entire address. Other tests that are part of "personal"
5788 are now done using regular expressions (for example, to check local parts
5789 of addresses in From: header lines).
5791 (2) The list of non-personal local parts in From: addresses has been
5792 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5793 taken from the Sieve specification recommendations.
5795 (3) If the message contains any header line starting with "List-" it is
5796 treated as non-personal.
5798 (4) The test for "circular" in the Subject: header line has been removed
5799 because it now seems ill-conceived.
5801 15. Minor typos in src/EDITME comments corrected.
5803 16. Installed latest exipick from John Jetmore.
5805 17. If headers_add on a router specified a text string that was too long for
5806 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5807 of string_sprintf() is now avoided.
5809 18. $message_body_size was not set (it was always zero) when running the DATA
5810 ACL and the local_scan() function.
5812 19. For the "mail" command in an Exim filter, no default was being set for
5813 the once_repeat time, causing a random time value to be used if "once" was
5814 specified. (If the value happened to be <= 0, no repeat happened.) The
5815 default is now 0s, meaning "never repeat". The "vacation" command was OK
5816 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5817 (I found it when inspecting the code).
5819 20. There is now an overall timeout for performing a callout verification. It
5820 defaults to 4 times the callout timeout, which applies to individual SMTP
5821 commands during the callout. The overall timeout applies when there is more
5822 than one host that can be tried. The timeout is checked before trying the
5823 next host. This prevents very long delays if there are a large number of
5824 hosts and all are timing out (e.g. when the network connections are timing
5825 out). The value of the overall timeout can be changed by specifying an
5826 additional sub-option for "callout", called "maxwait". For example:
5828 verify = sender/callout=5s,maxwait=20s
5830 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5831 to the end before writing, but this should make it even safer).
5833 22. Exim was forgetting that it had advertised PIPELINING for the second and
5834 subsequent messages on an SMTP connection. It was also not resetting its
5835 memory on STARTTLS and an internal HELO.
5837 23. When Exim logs an SMTP synchronization error within a session, it now
5838 records whether PIPELINING has been advertised or not.
5840 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5841 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5842 rather than long int.
5844 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5846 26. Added the never_mail option to autoreply.
5852 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5853 crash if the getsockname() call failed; this can happen if a connection is
5854 closed very soon after it is established. The problem was simply in the
5855 order in which certain operations were done, causing Exim to try to write
5856 to the SMTP stream before it had set up the file descriptor. The bug has
5857 been fixed by making things happen in the correct order.
5863 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5864 before the connection was closed, thus losing the rejection response.
5866 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5867 some early Solaris releases, but causes trouble in current releases where
5868 socklen_t is defined.
5870 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5873 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5876 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5877 that releasing the top of it at the end releases what was used for sub-
5878 expansions (unless the block got too big). However, discard this block if
5879 the first thing is a variable or header, so that we can use its block when
5880 it is dynamic (useful for very large $message_headers, for example).
5882 6. Lookups now cache *every* query, not just the most recent. A new, separate
5883 store pool is used for this. It can be recovered when all lookup caches are
5884 flushed. Lookups now release memory at the end of their result strings.
5885 This has involved some general refactoring of the lookup sources.
5887 7. Some code has been added to the store_xxx() functions to reduce the amount
5888 of flapping under certain conditions.
5890 8. log_incoming_interface used to affect only the <= reception log lines. Now
5891 it causes the local interface and port to be added to several more SMTP log
5892 lines, for example "SMTP connection from", and rejection lines.
5894 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5896 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5898 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5899 could be overwritten at the end of the current message (or the start of a
5900 new message if it was set in a HELO ACL). The value is now preserved for
5901 the duration of the SMTP connection.
5903 12. If a transport had a headers_rewrite setting, and a matching header line
5904 contained an unqualified address, that address was qualified, even if it
5905 did not match any rewriting rules. The underlying bug was that the values
5906 of the flags that permit the existence of unqualified sender and recipient
5907 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5908 non-local messages, and by -bnq for local messages) were not being
5909 preserved with the message after it was received.
5911 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5912 "next input=" as part of the text comprising the host identity instead of
5913 the correct text. The code was using the same buffer for two different
5914 strings. However, depending on which order the printing function evaluated
5915 its arguments, the bug did not always show up. Under Linux, for example, my
5916 test suite worked just fine.
5918 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5919 doesn't work with some older versions of Perl. It has been changed to "my",
5920 which in any case is probably the better facility to use.
5922 15. A really picky compiler found some instances of statements for creating
5923 error messages that either had too many or two few arguments for the format
5926 16. The size of the buffer for calls to the DNS resolver has been increased
5927 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5928 for addresses that have a lot of PTR records. This alleviates a problem; it
5929 does not fully solve it.
5931 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5932 buffer now truncates the list and logs the incident, which is the same
5933 action as happens when Exim is looking up a host name and its aliases.
5934 Previously in this situation something unpredictable would happen;
5935 sometimes it was "internal error: store_reset failed".
5937 18. If a server dropped the connection unexpectedly when an Exim client was
5938 using GnuTLS and trying to read a response, the client delivery process
5939 crashed while trying to generate an error log message.
5941 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5942 string, for example:
5944 warn message = H1: something\nH2: something
5946 the text was added as a single header line from Exim's point of view
5947 though it ended up OK in the delivered message. However, searching for the
5948 second and subsequent header lines using $h_h2: did not work. This has been
5949 fixed. Similarly, if a system filter added multiple headers in this way,
5950 the routers could not see them.
5952 20. Expanded the error message when iplsearch is called with an invalid key to
5953 suggest using net-iplsearch in a host list.
5955 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5956 ACLs are no longer actually imposed (and a message to that effect is
5959 22. If a "gecos" field in a passwd entry contained escaped characters, in
5960 particular, if it contained a \" sequence, Exim got it wrong when building
5961 a From: or a Sender: header from that name. A second bug also caused
5962 incorrect handling when an unquoted " was present following a character
5963 that needed quoting.
5965 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5966 was not being matched caselessly.
5968 24. Arranged for all hyphens in the exim.8 source to be escaped with
5971 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5972 properly. Recipient callout cache records were still being keyed to include
5973 the sender, even when use_sender was set false. This led to far more
5974 callouts that were necessary. The sender is no longer included in the key
5975 when use_sender is false.
5977 26. Added "control = submission" modifier to ACLs.
5979 27. Added the ${base62d: operator to decode base 62 numbers.
5981 28. dnsdb lookups can now access SRV records.
5983 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5984 the configuration file.
5986 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5987 (-v) mode. This makes the output for a verbose queue run more intelligible.
5989 31. Added a use_postmaster feature to recipient callouts.
5991 32. Added the $body_zerocount variable, containing the number of binary zero
5992 bytes in the message body.
5994 33. The time of last modification of the "new" subdirectory is now used as the
5995 "mailbox time last read" when there is a quota error for a maildir
5998 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6000 35. Added +ignore_unknown as a special item in host lists.
6002 36. Code for decoding IPv6 addresses in host lists is now included, even if
6003 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6004 address was recognized as an IP address, but was then not correctly decoded
6005 into binary, causing unexpected and incorrect effects when compared with
6012 1. Very minor rewording of debugging text in manualroute to say "list of
6013 hosts" instead of "hostlist".
6015 2. If verify=header_syntax was set, and a header line with an unqualified
6016 address (no domain) and a large number of spaces between the end of the
6017 name and the colon was received, the reception process suffered a buffer
6018 overflow, and (when I tested it) crashed. This was caused by some obsolete
6019 code that should have been removed. The fix is to remove it!
6021 3. When running in the test harness, delay a bit after writing a bounce
6022 message to get a bit more predictability in the log output.
6024 4. Added a call to search_tidyup() just before forking a reception process. In
6025 theory, someone could use a lookup in the expansion of smtp_accept_max_
6026 per_host which, without the tidyup, could leave open a database connection.
6028 5. Added the variables $recipient_data and $sender_data which get set from a
6029 lookup success in an ACL "recipients" or "senders" condition, or a router
6030 "senders" option, similar to $domain_data and $local_part_data.
6032 6. Moved the writing of debug_print from before to after the "senders" test
6035 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6036 problems for message scanning, either using a data ACL, or using
6037 local_scan() because the Received: header was not generated till after they
6038 were called (in order to set the time as the time of reception completion).
6039 I have revised the way this works. The header is now generated after the
6040 body is received, but before the ACL or local_scan() are called. After they
6041 are run, the timestamp in the header is updated.
6047 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6048 before starting a queue runner without re-exec. This happened only when
6049 deliver_drop_privilege was set or when the Exim user was set to root. The
6050 effect of the bug was that timeouts during subsequent deliveries caused
6051 crashes instead of being properly handled. The handler is now left at its
6052 default (and expected) setting.
6054 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6055 message, again when deliver_drop_privilege is set or Exim is run as root.
6056 The bug described in (1) was not present in this case, but the tidying up
6057 of the other signals was missing. I have made the two cases consistent.
6059 3. The ignore_target_hosts setting on a manualroute router was being ignored
6060 for hosts that were looked up using the /MX notation.
6062 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6065 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6066 operated on the sender address. After changing the $sender_address to <>
6067 for the sender address verify, Exim was re-instated it as the original
6068 (before rewriting) address, but remembering that it had rewritten it, so it
6069 wasn't rewriting it again. This bug also had the effect of breaking the
6070 sender address verification caching when the sender address was rewritten.
6072 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6073 This has been changed so that if the ip literal address matches
6074 ignore_target_hosts, the router declines.
6076 7. Added expansion conditions match_domain, match_address, and match_local_
6077 part (NOT match_host).
6079 8. The placeholder for the Received: header didn't have a length field set.
6081 9. Added code to Exim itself and to exim_lock to test for a specific race
6082 condition that could lead to file corruption when using MBX delivery. The
6083 issue is with the lockfile that is created in /tmp. If this file is removed
6084 after a process has opened it but before that process has acquired a lock,
6085 there is the potential for a second process to recreate the file and also
6086 acquire a lock. This could lead to two Exim processes writing to the file
6087 at the same time. The added code performs the same test as UW imapd; it
6088 checks after acquiring the lock that its file descriptor still refers to
6089 the same named file.
6091 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6092 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6095 11. Added the smtp_active_hostname option. If used, this will typically be made
6096 to depend on the incoming interface address. Because $interface_address is
6097 not set up until the daemon has forked a reception process, error responses
6098 that can happen earlier (such as "too many connections") no longer contain
6101 12. If an expansion in a condition on a "warn" statement fails because a lookup
6102 defers, the "warn" statement is abandoned, and the next ACL statement is
6103 processed. Previously this caused the whole ACL to be aborted.
6105 13. Added the iplsearch lookup type.
6107 14. Added ident_timeout as a log selector.
6109 15. Added tls_certificate_verified as a log selector.
6111 16. Added a global option tls_require_ciphers (compare the smtp transport
6112 option of the same name). This controls incoming TLS connections.
6114 17. I finally figured out how to make tls_require_ciphers do a similar thing
6115 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6116 before starting the TLS session.
6118 18. Tabs are now shown as \t in -bP output.
6120 19. If the log selector return_path_on_delivery was set, Exim crashed when
6121 bouncing a message because it had too many Received: header lines.
6123 20. If two routers both had headers_remove settings, and the first one included
6124 a superfluous trailing colon, the final name in the first list and the
6125 first name in the second list were incorrectly joined into one item (with a
6126 colon in the middle).
6132 1. Added -C and -D options to the exinext utility, mainly to make it easier
6133 to include in the automated testing, but these could be helpful when
6134 multiple configurations are in use.
6136 2. The exinext utility was not formatting the output nicely when there was
6137 an alternate port involved in the retry record key, nor when there was a
6138 message id as well (for retries that were specific to a specific message
6139 and a specific host). It was also confused by IPv6 addresses, because of
6140 the additional colons they contain. I have fixed the IPv4 problem, and
6141 patched it up to do a reasonable job for IPv6.
6143 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6144 delivery, the log line now contains "pipelined" if PIPELINING was used.
6146 4. An SMTP transport process used to panic and die if the bind() call to set
6147 an explicit outgoing interface failed. This has been changed; it is now
6148 treated in the same way as a connect() failure.
6150 5. A reference to $sender_host_name in the part of a conditional expansion
6151 that was being skipped was still causing a DNS lookup. This no longer
6154 6. The def: expansion condition was not recognizing references to header lines
6155 that used bh_ and bheader_.
6157 7. Added the _cache feature to named lists.
6159 8. The code for checking quota_filecount in the appendfile transport was
6160 allowing one more file than it should have been.
6162 9. For compatibility with Sendmail, the command line option
6170 and sets the incoming protocol and host name (for trusted callers). The
6171 host name and its colon can be omitted when only the protocol is to be set.
6172 Note the Exim already has two private options, -pd and -ps, that refer to
6173 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6174 "s", but I don't think that's a major issue.
6176 10. A number of refactoring changes to the code, none of which should affect
6179 (a) The number of logging options was getting close to filling up the
6180 32-bit word that was used as a bit map. I have split them into two classes:
6181 those that are passed in the argument to log_write(), and those that are
6182 only ever tested independently outside of that function. These are now in
6183 separate 32-bit words, so there is plenty of room for expansion again.
6184 There is no change in the user interface or the logging behaviour.
6186 (b) When building, for example, log lines, the code previously used a
6187 macro that called string_cat() twice, in order to add two strings. This is
6188 not really sufficiently general. Furthermore, there was one instance where
6189 it was actually wrong because one of the argument was used twice, and in
6190 one call a function was used. (As it happened, calling the function twice
6191 did not affect the overall behaviour.) The macro has been replaced by a
6192 function that can join an arbitrary number of extra strings onto a growing
6195 (c) The code for expansion conditions now uses a table and a binary chop
6196 instead of a serial search (which was left over from when there were very
6197 few conditions). Also, it now recognizes conditions like "pam" even when
6198 the relevant support is not compiled in: a suitably worded error message is
6199 given if an attempt is made to use such a condition.
6201 11. Added ${time_interval:xxxxx}.
6203 12. A bug was causing one of the ddress fields not to be passed back correctly
6204 from remote delivery subprocesses. The field in question was not being
6205 subsequently used, so this caused to problems in practice.
6207 13. Added new log selectors queue_time and deliver_time.
6209 14. Might have fixed a bug in maildirsizefile handling that threw up
6210 "unexpected character" debug warnings, and recalculated the data
6211 unnecessarily. In any case, I expanded the warning message to give more
6214 15. Added the message "Restricted characters in address" to the statements in
6215 the default ACL that block characters like @ and % in local parts.
6217 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6218 Three changes have been made:
6220 (a) There was a serious bug; a negative response to MAIL caused the whole
6221 recipient domain to be cached as invalid, thereby blocking all messages
6222 to all local parts at the same domain, from all senders. This bug has
6223 been fixed. The domain is no longer cached after a negative response to
6224 MAIL if the sender used is not empty.
6226 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6229 (c) A new callout option, "use_sender" has been added for people who want
6230 the modified behaviour.
6236 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6239 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6240 indeed breaks things for older releases.
6242 3. Added additional logging to the case where there is a problem reading data
6243 from a filter that is running in a subprocess using a pipe, in order to
6244 try to track down a specific problem.
6246 4. Testing facility fudge: when running in the test harness and attempting
6247 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6248 getting "No route to host". Convert this to a timeout.
6250 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6253 6. Some OS don't have socklen_t but use size_t instead. This affects the
6254 fifth argument of getsockopt() amongst other things. This is now
6255 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6256 can be set for individual OS. I have set it for SunOS5, OSF1, and
6257 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6258 some earlier ones do not.
6260 7. Change 4.30/15 was not doing the test caselessly.
6262 8. The standard form for an IPv6 address literal was being rejected by address
6263 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6264 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6265 this, as well as the form without the "IPv6" on the front (but only when
6266 address literals are enabled, of course).
6268 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6270 10. Exim crashed if a message with an empty sender address specified by -f
6271 encountered a router with an errors_to setting. This could be provoked only
6272 by a command such as
6276 where an empty string was supplied; "<>" did not hit this bug.
6278 11. Installed PCRE release 4.5.
6280 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6281 remained set. It is now erased.
6283 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6284 times from message ids (which are base 36 rather than the normal 62).
6286 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6287 were being counted as actual protocol errors, and logged if the log
6288 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6289 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6290 and DATA following a set of rejected RCPTs do not count as protocol errors.
6291 In other words, Exim assumes they were pipelined, though this may not
6292 actually be the case. Of course, in all cases the client gets an
6293 appropriate error code.
6295 15. If a lookup fails in an ACL condition, a message about the failure may
6296 be available; it is used if testing the ACL cannot continue, because most
6297 such messages specify what the cause of the deferral is. However, some
6298 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6299 that caused an old message to be retained and used if a later statement
6300 caused a defer, replacing the real cause of the deferral.
6302 16. If an IP address had so many PTR records that the DNS lookup buffer
6303 was not large enough to hold them, Exim could crash while trying to process
6304 the truncated data. It now detects and logs this case.
6306 17. Further to 4.21/58, another change has been made: if (and only if) the
6307 first line of a message (the first header line) ends with CRLF, a bare LF
6308 in a subsequent header line has a space inserted after it, so as not to
6309 terminate the header.
6311 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6312 unnecessarily, used atoi() instead of strtol(), and didn't check the
6313 termination when getting file sizes from file names by regex.
6315 19. Completely re-implemented the support for maildirsize files, in the light
6316 of a number of problems with the previous contributed implementation
6317 (4.30/29). In particular:
6319 . If the quota is zero, the maildirsize file is maintained, but no quota is
6322 . If the maildir directory does not exist, it is created before any attempt
6323 to write a maildirsize file.
6325 . The quota value in the file is just a cache; if the quota is changed in
6326 the transport, the new value overrides.
6328 . A regular expression is available for excluding directories from the
6331 20. The autoreply transport checks the characters in options that define the
6332 message's headers; it allows continued headers, but it was checking with
6333 isspace() after an embedded newline instead of explicitly looking for a
6336 21. If all the "regular" hosts to which an address was routed had passed their
6337 expiry times, and had not reached their retry times, the address was
6338 bounced, even if fallback hosts were defined. Now Exim should go on to try
6341 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6342 equivalent code in the SMTP transport. Some hosts send humungous responses
6343 to HELO/EHLO, more than 1024 it seems.
6345 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6346 gives compiler warnings in some environments. I've now done it "properly",
6349 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6350 (because of problems with the built-in one) was declared to return uschar *
6351 instead of char *, causing compiler failure.
6353 25. Fixed a file descriptor leak when processing alias/forward files.
6355 26. Fixed a minor format string issue in dbfn.c.
6357 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6359 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6360 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6363 29. When a sender address is verified, it is cached, to save repeating the test
6364 when there is more than one recipient in a message. However, when the
6365 verification involves a callout, it is possible for different callout
6366 options to be set for different recipients. It is too complicated to keep
6367 track of this in the cache, so now Exim always runs a verification when a
6368 callout is required, relying on the callout cache for the optimization.
6369 The overhead is duplication of the address routing, but this should not be
6372 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6373 to be verified with callout=postmaster, and the main callout worked but the
6374 postmaster check failed, the verification correctly failed. However, if a
6375 subsequent RCPT command asked for sender verification *without* the
6376 postmaster check, incorrect caching caused this verification also to fail,
6379 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6380 it was not caching the DNS options (qualify_single, search_parents) that
6381 were used when the lookup failed. A subsequent lookup with different
6382 options therefore always gave the same answer, though there were cases
6383 where it should not have. (Example: a "domains = !$mx_any" option on a
6384 dnslookup router: the "domains" option is always processed without any
6385 widening, but the router might have qualify_single set.) Now Exim uses the
6386 cached value only when the same options are set.
6388 32. Added John Jetmore's "exipick" utility to the distribution.
6390 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6391 than a timeout (e.g. a certificate is required, and is not provided), an
6392 Exim server now closes the connection immediately. Previously it waited for
6393 the client to close - but if the client is SSL, it seems that they each
6394 wait for each other, leading to a delay before one of them times out.
6396 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6397 maintained 0.8.x compatibility because I don't think many are using it, and
6398 it is clearly obsolete.
6400 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6403 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6404 client certificate was expired. A simple patch fixes this, though I don't
6405 understand the full logic of why the verify callback is called multiple
6408 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6409 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6410 which causes problems with some clients (such as the Certicom SSL Plus
6411 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6412 disables the coutermeasure allowing Eudora to connect."
6414 38. Exim was not checking that a write() to a log file succeeded. This could
6415 lead to Bad Things if a log got too big, in particular if it hit a file
6416 size limit. Exim now panics and dies if it cannot write to a log file, just
6417 as it does if it cannot open a log file.
6419 39. Modified OS/Makefile-Linux so that it now contains
6421 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6423 The two -D definitions ensure that Exim is compiled with large file
6424 support, which makes it possible to handle log files that are bigger than
6427 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6428 instance) a domain was checked against a named list that involved a lookup,
6429 causing $domain_data to be set, then another domain was checked against the
6430 same list, then the first domain was re-checked, the value of $domain_data
6431 after the final check could be wrong. In particular, if the second check
6432 failed, it could be set empty. This bug probably also applied to
6435 41. The strip_trailing_dot option was not being applied to the address given
6436 with the -f command-line option.
6438 42. The code for reading a message's header from the spool was incrementing
6439 $received_count, but never initializing it. This meant that the value was
6440 incorrect (doubled) while delivering a message in the same process in which
6441 it was received. In the most common configuration of Exim, this never
6442 happens - a fresh exec is done - but it can happen when
6443 deliver_drop_privilege is set.
6445 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6446 it now includes up to 150 characters of the unexpected data in the log
6449 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6450 and building data strings. The size of both of these buffers was 10 000
6451 bytes - far larger than anybody would *ever* want, thought I. Needless to
6452 say, somebody hit the limit. I have increased the maximum line length to
6453 20 000 and the maximum data length of concatenated lines to 100 000. I have
6454 also fixed two bugs, because there was no checking on these buffers. Tsk,
6455 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6456 buffer is too small.
6458 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6459 lsearch lookups. Now it does.
6461 46. When parsing a route_list item in a manualroute router, a fixed-length
6462 buffer was used for the list of hosts. I made this 1024 bytes long,
6463 thinking that nobody would ever have a list of hosts that long. Wrong.
6464 Somebody had a whole pile of complicated expansion conditions, and the
6465 string was silently truncated, leading to an expansion error. It turns out
6466 that it is easier to change to an unlimited length (owing to other changes
6467 that have happened since this code was originally written) than to build
6468 structure for giving a limitation error. The length of the item that
6469 expands into the list of hosts is now unlimited.
6471 47. The lsearch lookup could not handle data where the length of text line was
6472 more than 4095 characters. Such lines were truncated, leading to shortened
6473 data being returned. It should now handle lines of any length.
6475 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6476 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6479 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6480 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6481 no longer happens in comment lines. A list of replacements is now placed
6482 at the head of all of the source files, except those whose only change is
6483 to replace PERL_COMMAND in the very first #! line.
6485 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6486 messages on the queue, with a bottom-up merge sort, using code contributed
6487 by Michael Haardt. This should make operations like -bp somewhat faster on
6488 large queues. It won't affect queue runners, except when queue_run_in_order
6491 51. Installed eximstats 1.31 in the distribution.
6493 52. Added support for SRV lookups to the dnslookup router.
6495 53. If an ACL referred to $message_body or $message_body_end, the value was not
6496 reset for any messages that followed in the same SMTP session.
6498 54. The store-handling optimization for building very long strings was not
6499 differentiating between the different store pools. I don't think this
6500 actually made any difference in practice, but I've tidied it.
6502 55. While running the routers to verify a sender address, $sender_address
6503 was still set to the sender address. This is wrong, because when routing to
6504 send a bounce to the sender, it would be empty. Therefore, I have changed
6505 it so that, while verifying a sender address, $sender_address is set to <>.
6506 (There is no change to what happens when verifying a recipient address.)
6508 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6509 target A or AAAA records (if not already returned) without resetting the
6510 qualify_single or search_parents options of the DNS resolver. These are
6511 inappropriate in this case because the targets of MX and SRV records must
6512 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6513 target that, when qualified, matched something in the local domain. These
6514 two options are now turned off when doing these lookups.
6516 57. It seems that at least some releases of Reiserfs (which does not have the
6517 concept of a fixed number of inodes) returns zero and not -1 for the
6518 number of available inodes. This interacted badly with check_spool_inodes,
6519 which assumed that -1 was the "no such thing" setting. What I have done is
6520 to check that the total number of inodes is greater than zero before doing
6521 the test of how many are available.
6523 58. When a "warn" ACL statement has a log_message modifier, the message is
6524 remembered, and not repeated. This is to avoid a lot of repetition when a
6525 message has many recipients that cause the same warning to be written.
6526 However, Exim was preserving the list of already written lines for an
6527 entire SMTP session, which doesn't seem right. The memory is now reset if a
6528 new message is started.
6530 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6531 debugging output unless log_rewrite was also set.
6533 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6534 of (int)(handle) when we know that handle contains (void *)(-1).
6536 61. The Exim daemon panic-logs an error return when it closes the incoming
6537 connection. However "connection reset by peer" seems to be common, and
6538 isn't really an error worthy of noting specially, so that particular error
6541 62. When Exim is trying to find all the local interfaces, it used to panic and
6542 die if the ioctl to get the interface flags failed. However, it seems that
6543 on at least one OS (Solaris 9) it is possible to have an interface that is
6544 included in the list of interfaces, but for which you get a failure error
6545 for this call. This happens when the interface is not "plumbed" into a
6546 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6547 failure of the "get flags" call assumes that the interface is down.
6549 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6550 makes life easier for people who are doing arithmetic on fields extracted
6551 from dates, where you often get leading zeros that should not be
6552 interpreted as octal.
6554 64. Added qualify_domain to the redirect router, to override the global
6557 65. If a pathologically long header line contained very many addresses (the
6558 report of this problem mentioned 10 000) and each of them was rewritten,
6559 Exim could use up a very large amount of memory. (It kept on making new
6560 copies of the header line as it rewrote, and never released the old ones.)
6561 At the expense of a bit more processing, the header rewriting function has
6562 been changed so that it no longer eats memory in this way.
6564 66. The generation of the Received: header has been moved from the time that a
6565 message starts to be received, to the time that it finishes. The timestamp
6566 in the Received: header should now be very close to that of the <= log
6567 line. There are two side-effects of this change:
6569 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6570 logged header lines no longer include the local Received: line, because
6571 it has not yet been created. The same applies to a copy of the message
6572 that is returned to a non-SMTP sender when a message is rejected.
6574 (b) When a filter file is tested using -bf, no additional Received: header
6575 is added to the test message. After some thought, I decided that this
6578 This change does not affect the value of $received_for. It is still set
6579 after address rewriting, but before local_scan() is called.
6581 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6583 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6584 gave an unhelpful panic error message, and a defer error. I have managed to
6585 change this behaviour so that it now rejects any supplied certificate,
6586 which seems right, as the list of acceptable certificates is empty.
6588 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6589 gave an unhelpful defer error. I have not managed to make this reject any
6590 supplied certificates, but the error message it gives is "no certificate
6591 supplied", which is not helpful.
6593 70. exigrep's output now also includes lines that are not associated with any
6594 message, but which match the given pattern. Implemented by a patch from
6595 Martin Sluka, which also tidied up the Perl a bit.
6597 71. Recipient callout verification, like sender verification, was using <> in
6598 the MAIL FROM command. This isn't really the right thing, since the actual
6599 sender may affect whether the remote host accepts the recipient or not. I
6600 have changed it to use the actual sender in the callout; this means that
6601 the cache record is now keyed on a recipient/sender pair, not just the
6602 recipient address. There doesn't seem to be a real danger of callout loops,
6603 since a callout by the remote host to check the sender would use <>.
6604 [SEE ABOVE: changed after hitting problems.]
6606 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6607 temporary errors. However, in the case of such a code being given after
6608 the end of a data transmission (i.e. after ".") Exim was failing to write
6609 a retry record for the message. (Yes, there was some broken host that was
6610 actually sending 8xx at this point.)
6612 73. An unknown lookup type in a host list could cause Exim to panic-die when
6613 the list was checked. (An example that provoked this was putting <; in the
6614 middle of a list instead of at the start.) If this happened during a DATA
6615 ACL check, a -D file could be left lying around. This kind of configuration
6616 error no longer causes Exim to die; instead it causes a defer error. The
6617 incident is still logged to the main and panic logs.
6619 74. Buglet left over from Exim 3 conversion. The message "too many messages
6620 in one connection" was written to the rejectlog but not the mainlog, except
6621 when address rewriting (yes!) was being logged.
6623 75. Added write_rejectlog option.
6625 76. When a system filter was run not as root (that is, when system_filter_user
6626 was set), the values of the $n variables were not being returned to the
6627 main process; thus, they were not subsequently available in the $sn
6630 77. Added +return_path_on_delivery log selector.
6632 78. A connection timeout was being treated differently from recipients deferred
6633 when testing hosts_max_try with a message that was older than the host's
6634 retry timeout. (The host should not be counted, thus allowing all hosts to
6635 be tried at least once before bouncing.) This may have been the cause of an
6636 occasionally reported bug whereby a message would remain on the queue
6637 longer than the retry timeout, but would be bounced if a delivery was
6638 forced. I say "may" because I never totally pinned down the problem;
6639 setting up timeout/retry tests is difficult. See also the next item.
6641 79. The ultimate address timeout was not being applied to errors that involved
6642 a combination of host plus message (for example, a timeout on a MAIL
6643 command). When an address resolved to a number of possible hosts, and they
6644 were not all tried for each delivery (e.g. because of hosts_max_try), a
6645 message could remain on the queue longer than the retry timeout.
6647 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6650 81. Fixed an obscure SMTP outgoing bug which required at least the following
6651 conditions: (a) there was another message waiting for the same server;
6652 (b) the server returned 5xx to all RCPT commands in the first message so
6653 that the message was not completed; (c) the server dropped the connection
6654 or gave a negative response to the RSET that Exim sends to abort the
6655 transaction. The observed case was a dropped connection after DATA that had
6656 been sent in pipelining mode. That is, the server had advertised PIPELINING
6657 but was not implementing it correctly. The effect of the bug was incorrect
6658 behaviour, such as trying another host, and this could lead to a crash.
6664 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6665 and daemon.c were passed as pointers to ints; they should have been
6666 pointers to socklen_t variables (which are typically unsigned ints).
6668 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6671 3. Fixed a really odd bug that affected only the testing scheme; patching a
6672 certain fixed string in the binary changed the value of another string that
6673 happened to be identical to the end of the original first string.
6675 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6676 name", it returns that address as the IP address. On some operating
6677 systems (e.g. Solaris), it also passes back the IP address string as the
6678 "host name". However, on others (e.g. Linux), it passes back an empty
6679 string. Exim wasn't checking for this, and was changing the host name to an
6680 empty string, assuming it had been canonicalized.
6682 5. Although rare, it is permitted to have more than one PTR record for a given
6683 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6684 all the names associated with an address, because they do in Solaris.
6685 However, it seems that they do not in Linux for data that comes from the
6686 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6687 I found this out when I moved to a new Linux workstation and tried to run
6688 the Exim test suite.
6690 To get round this problem I have changed the code so that it now does its
6691 own call to the DNS to look up PTR records when searching for a host name.
6692 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6693 addresses that are only in /etc/hosts are still found.
6695 This behaviour is, however, controlled by an option called host_lookup_
6696 order, which defaults to "bydns:byaddr". If people want to use the other
6697 order, or indeed, just use one or the other means of lookup, they can
6698 specify it in this variable.
6700 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6701 some operating systems, this comes back from gethostbyaddr() as an empty
6702 string, and this is what Exim used to test for. However, it seems that in
6703 other systems, "." is yielded. Exim now tests for this case too.
6705 7. The values of check_spool_space and check_log_space are now held internally
6706 as a number of kilobytes instead of an absolute number of bytes. If a
6707 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6708 kilobyte. This means that much larger values can be stored.
6710 8. Exim monitor: an attempt to get the action menu when not actually pointing
6711 at a message produces an empty menu entitled "No message selected". This
6712 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6713 no entries in it ("Shell widget menu has zero width and/or height"). So I
6714 have added a single, blank menu entry in this case.
6716 9. Added ${quote_local_part.
6718 10. MIME decoding is now applied to the contents of Subject: header lines when
6721 11. Now that a reference to $sender_host_address automatically causes a reverse
6722 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6723 host lookup before query-style lookups in lists that might use this
6724 variable. This has therefore been abolished, and the "net-" prefix is no
6725 longer necessary for query-style lookups.
6727 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6728 have been a typo for LFLAGS, so it has been changed.
6730 13. The install script calls Exim with "-C /dev/null" in order to find the
6731 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6732 to be output. However, since Exim outputs its version number before the
6733 error, it didn't break the script. It just looked ugly. I fixed this by
6734 always allowing "-C /dev/null" if the caller is root.
6736 14. Ignore overlarge ACL variable number when reading spool file - insurance
6737 against a later release with more variables having written the file.
6739 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6740 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6741 this, as well as the form without the "IPv6" on the front.
6743 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6744 OS/Makefile-Darwin file.
6746 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6747 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6749 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6750 variable. However, some OS use ints for inodes. I've added cast to long int
6751 to get rid of the compiler warning.
6753 19. I had forgotten to lock out "/../" in configuration file names when
6754 ALT_CONFIG_PREFIX was set.
6756 20. Routers used for verification do not need to specify transports. However,
6757 if such a router generated a host list, and callout was configured, Exim
6758 crashed, because it could not find a port number from the (non-existent)
6759 transport. It now assumes port 25 in this circumstance.
6761 21. Added the -t option to exigrep.
6763 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6764 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6765 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6766 them is set, LOOKUP_LSEARCH is forced.
6768 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6770 24. Added sender and host information to the "rejected by local_scan()" log
6771 line; previously there was no indication of these.
6773 25. Added .include_if_exists.
6775 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6776 Linux. It turns out that not all file systems support this. Apparently some
6777 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6778 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6779 not supported on the file descriptor, is now ignored when Exim is trying to
6780 sync a directory. This applies only to Linux.
6782 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6784 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6785 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6788 29. Added support for maildirsize files from supplied patch (modified a bit).
6790 30. The use of :fail: followed by an empty string could lead Exim to respond to
6791 sender verification failures with (e.g.):
6793 550 Verification failed for <xxx>
6794 550 Sender verify failed
6796 where the first response line was missing the '-' that indicates it is not
6797 the final line of the response.
6799 31. The loop for finding the name of the user that called Exim had a hardwired
6800 limit of 10; it now uses the value of finduser_retries, which is used for
6801 all other user lookups.
6803 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6806 33. Exim was neglecting to zero errno before one call of strtol() when
6807 expanding a string and expecting an integer value. On some systems this
6808 resulted in spurious "integer overflow" errors. Also, it was casting the
6809 result into an int without checking.
6811 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6812 did not work. The code looks as if it has *never* worked, though it appears
6813 to have been documented since at least release 1.62. I have made it work.
6815 35. The "timeout_DNS" error in retry rules, also documented since at least
6816 1.62, also never worked. As it isn't clear exactly what this means, and
6817 clearly it isn't a major issue, I have abolished the feature by treating it
6818 as "timeout", and writing a warning to the main and panic logs.
6820 36. The display of retry rules for -brt wasn't always showing the error code
6823 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6824 timeout_connect_A, timeout_connect_MX.
6826 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6827 to the empty sender.
6829 39. The daemon was not analysing the content of -oX till after it had closed
6830 stderr and disconnected from the controlling terminal. This meant that any
6831 syntax errors were only noted on the panic log, and the return code from
6832 the command was 0. By re-arranging the code a little, I've made the
6833 decoding happen first, so such errors now appear on stderr, and the return
6834 code is 1. However, the actual setting up of the sockets still happens in
6835 the disconnected process, so errors there are still only recorded on the
6838 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6839 connections (as happens on some IP stacks) was logged at start up time as
6840 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6841 it from "IPv6 and IPv4", which means that two separate sockets are being
6844 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6845 says whether AF_INET or AF_INET6 was passed as an argument.
6847 42. Exiwhat output was messed up when time zones were included in log
6850 43. Exiwhat now gives more information about the daemon's listening ports,
6851 and whether -tls-on-connect was used.
6853 44. The "port" option of the smtp transport is now expanded.
6855 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6856 silently ignored. Now an error message is written to the main and panic
6859 46. There's a new ACL modifier called "logwrite" which writes to a log file
6860 as soon as it is encountered.
6862 47. Added $local_user_uid and $local_user_gid at routing time.
6864 48. Exim crashed when trying to verify a sender address that was being
6867 49. Exim was recognizing only a space character after ".include". It now also
6868 recognizes a tab character.
6870 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6871 extracting the relevant information from the specification. The man page no
6872 longer contains scrambled data for the -d option, and I've added a section
6873 at the front about calling Exim under different names.
6875 51. Added "extra_headers" argument to the "mail" command in filter files.
6877 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6880 53. Installed eximstats 1.29.
6882 54. Added transport_filter_timeout as a generic transport option.
6884 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6885 Cc: header lines. This was required by RFC 822, but it not required by RFC
6888 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6889 incoming messages that did not have them. Now it does so only if the
6890 message originates locally, that is, if there is no associated remote host
6891 address. When Resent- header lines are present, this applies to the Resent-
6892 lines rather than the non-Resent- lines.
6894 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6895 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6897 58. Messages for configuration errors now include the name of the main
6898 configuration file - useful now that there may be more than one file in a
6899 list (.included file names were always shown).
6901 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6902 for those rare installations that do not start the daemon as root or run it
6903 setuid root. I've cut out the call to initgroups() if the daemon is not
6906 60. The Exim user and group can now be bound into the binary as text strings
6907 that are looked up at the start of Exim's processing.
6909 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6911 62. Added $mailstore_basename variable.
6913 63. Installed patch to sieve.c from Michael Haardt.
6915 64. When Exim failed to open the panic log after failing to open the main log,
6916 the original message it was trying to log was written to stderr and debug
6917 output, but if they were not available (the usual case in production), it
6918 was lost. Now it is written to syslog before the two lines that record the
6919 failures to open the logs.
6921 65. Users' Exim filters run in subprocesses under the user's uid. It is
6922 possible for a "deliver" command or an alias in a "personal" command to
6923 provoke an address rewrite. If logging of address rewriting is configured,
6924 this fails because the process is not running as root or exim. There may be
6925 a better way of dealing with this, but for the moment (because 4.30 needs
6926 to be released), I have disabled address rewrite logging when running a
6927 filter in a non-root, non-exim process.
6933 1. The buildconfig auxiliary program wasn't quoting the value set for
6934 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6935 not defined. This bug was masked in 4.22 by the effect that was fixed in
6938 2. Some messages that were rejected after a message id was allocated were
6939 shown as "incomplete" by exigrep. It no longer does this for messages that
6940 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6942 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6943 have allow_domain_literals set, the ID did not get logged in the <= line.
6944 Domain literals are now always recognized in Message-ID: header lines.
6946 4. The first argument for a ${extract expansion item is the key name or field
6947 number. Leading and trailing spaces in this item were not being ignored,
6948 causing some misleading effects.
6950 5. When deliver_drop_privilege was set, single queue runner processes started
6951 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6952 same command in the process it spins off) were not dropping privilege.
6954 6. When the daemon running as "exim" started a queue runner, it always
6955 re-executed Exim in the spun-off process. This is a waste of effort when
6956 deliver_drop_privilege is set. The new process now just calls the
6957 queue-runner function directly.
6963 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6966 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6967 ignored. Though the use of -oP was forcing the writing of a pid file, it
6968 was always written to the default place.
6970 3. If the message "no IP address found for host xxxx" is generated during
6971 incoming verification, it is now followed by identification of the incoming
6972 connection (so you can more easily find what provoked it).
6974 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6976 5. Added some features to "harden" Exim a bit more against certain attacks:
6978 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6979 be put in Local/Makefile. This is like the never_users runtime option,
6980 but it cannot be overridden. The default setting is "root".
6982 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6983 prefix string with which any file named in a -C command line option
6986 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6987 is retained for -C and -D only if the caller of Exim is root. Without
6988 it, the exim user may also use -C and -D and retain privilege.
6990 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6991 command line option is disabled.
6993 6. Macro names set by the -D option must start with an upper case letter, just
6994 like macro names defined in the configuration file.
6996 7. Added "dereference=" facility to LDAP.
6998 8. Two instances of the typo "uknown" in the source files are fixed.
7000 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7001 the Configure-Makefile script screwed up while processing it.
7003 10. Incorporated PCRE 4.4.
7005 11. The SMTP synchronization check was not operating right at the start of an
7006 SMTP session. For example, it could not catch a HELO sent before the client
7007 waited for the greeting. There is now a check for outstanding input at the
7008 point when the greeting is written. Because of the duplex, asynchronous
7009 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7010 way, but not yet received, when the check is performed.
7012 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7013 on TCP/IP sockets, because this apparently causes some broken clients to
7016 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7017 unchanged) from the Cygwin maintainer.
7019 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7020 was supported instead of testing for mbx. Effectively a typo.
7022 15. The spa authenticator server code was not checking that the input it
7023 received was valid base64.
7025 16. The debug output line for the "set" modifier in ACLs was not showing the
7026 name of the variable that was being set.
7028 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7030 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7031 lookup on its own. Something else was needed to trigger the lookup. For
7032 example, a match in host_lookup or the need for a host name in a host list.
7033 Now, if $sender_host_name is referenced and the host name has not yet been
7034 looked up, a lookup is performed. If the lookup fails, the variable remains
7035 empty, and $host_lookup_failed is set to "1".
7037 19. Added "eqi" as a case-independent comparison operator.
7039 20. The saslauthd authentication condition could segfault if neither service
7040 nor realm was specified.
7042 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7043 error message that was logged was misleading, and incoming SMTP
7044 connections were dropped. The message is now more accurate, and temporary
7045 errors are given to SMTP connections.
7047 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7048 (except RSET) with a 421 error, until QUIT is received. However, it was
7049 failing to send a response to QUIT.
7051 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7052 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7053 ACL is now run after the helo_try_verify_hosts code.
7055 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7056 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7057 case-independent, so other case variants are also recognized). Apparently
7058 some systems use these upper case variants.
7060 25. If more than two messages were waiting for the same host, and a transport
7061 filter was specified for the transport, Exim sent two messages over the
7062 same TCP/IP connection, and then failed with "socket operation on non-
7063 socket" when it tried to send the third.
7065 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7067 27. The extern definition of crypt16() in expand.c was not being excluded when
7068 the OS had its own crypt16() function.
7070 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7071 as a preferred synonym for return_size_limit, both as an option and as an
7074 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7076 30. Changed the default configuration ACL to relax the local part checking rule
7077 for addresses that are not in any local domains. For these addresses,
7078 slashes and pipe symbols are allowed within local parts, but the sequence
7079 /../ is explicitly forbidden.
7081 31. SPA server authentication was not clearing the challenge buffer before
7084 32. log_message in a "warn" ACL statement was writing to the reject log as
7085 well as to the main log, which contradicts the documentation and doesn't
7086 seem right (because no rejection is happening). So I have stopped it.
7088 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7089 However, I am unable to do any testing of this.
7091 34. Fixed an infelicity in the appendfile transport. When checking directories
7092 for a mailbox, to see if any needed to be created, it was accidentally
7093 using path names with one or more superfluous leading slashes; tracing
7094 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7096 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7097 contents are added to the log line that is written for every discarded
7098 recipient. (Previously a log_message setting was ignored.)
7100 36. The ${quote: operator now quotes the string if it is empty.
7102 37. The install script runs exim in order to find its version number. If for
7103 some reason other than non-existence or emptiness, which it checks, it
7104 could not run './exim', it was installing it with an empty version number,
7105 i.e. as "exim-". This error state is now caught, and the installation is
7108 38. An argument was missing from the function that creates an error message
7109 when Exim fails to connect to the socket for saslauthd authentication.
7110 This could cause Exim to crash, or give a corrupted message.
7112 39. Added isip, isip4, and isip6 to ${if conditions.
7114 40. The ACL variables $acl_xx are now saved with the message, and can be
7115 accessed later in routers, transports, and filters.
7117 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7118 strings in the file are not string-expanded.
7120 42. If a MAIL command specified a SIZE value that was too large to fit into an
7121 int variable, the check against message_size_limit failed. Such values are
7122 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7123 day this will have to be increased, but I don't think I want to be around
7124 when emails are that large.
7131 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7132 iconv() is not standard in FreeBSD.
7134 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7135 IPv6 enabled. The observed symptom was a segmentation fault on return from
7136 the function os_common_find_running_interfaces() in src/os.c.
7138 3. In the check_special_case() function in daemon.c I had used "errno" as an
7139 argument name, which causes warnings on some systems. This was basically a
7140 typo, since it was named "eno" in the comments!
7142 4. The code that waits for the clock to tick (at a resolution of some fraction
7143 of a second) so as to ensure message-id uniqueness was always waiting for
7144 at least one whole tick, when it could have waited for less. [This is
7145 almost certainly not relevant at current processor speeds, where it is
7146 unlikely to ever wait at all. But we try to future-proof.]
7148 5. The function that sleeps for a time interval that includes fractions of a
7149 second contained a race. It did not block SIGALRM between setting the
7150 timer, and suspending (a couple of lines later). If the interval was short
7151 and the sigsuspend() was delayed until after it had expired, the suspension
7152 never ended. On busy systems this could lead to processes getting stuck for
7155 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7156 process, before it forks any delivery processes. The open lookup caching
7157 mechanism meant that the open file or database connection was passed into
7158 the delivery process. The problem was that delivery processes always tidy
7159 up cached lookup data. This could cause a problem for the next delivery
7160 process started by the queue runner, because the external queue runner
7161 process does not know about the closure. So the next delivery process
7162 still has data in the lookup cache. In the case of a file lookup, there was
7163 no problem because closing a file descriptor in a subprocess doesn't affect
7164 the parent. However, if the lookup was caching a connection to a database,
7165 the connection was closed, and the second delivery process was likely to
7166 see errors such as "PGSQL: query failed: server closed the connection
7167 unexpectedly". The problem has been fixed by closing all cached lookups
7168 in a queue runner before running a delivery process.
7170 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7171 seem to have the "const" qualifier which it has on other OS. I've
7174 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7175 *of the same type* (client or server) with the same public name that an
7176 error should be diagnosed.
7178 9. When Exim looked up a host name for an IP address, but failed to find the
7179 original IP address when looking up the host name (a safety check), it
7180 output the message "<ip address> does not match any IP for NULL", which was
7181 confusing, to say the least. The bug was that the host name should have
7182 appeared instead of "NULL".
7184 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7185 user that is built into the binary, and the -C or -D options is used, root
7186 privilege is dropped before the configuration file is read. In addition,
7187 logging is switched to stderr instead of the normal log files. If the
7188 configuration then re-defines the Exim user, the unprivileged environment
7189 is probably not what is expected, so Exim logs a panic warning message (but
7192 However, if deliver_drop_privilege is set, the unprivileged state may well
7193 be exactly what is intended, so the warning has been cut out in that case,
7194 and Exim is allowed to try to write to its normal log files.
7200 1. smtp_return_error_details was not giving details for temporary sender
7201 or receiver verification errors.
7203 2. Diagnose a configuration error if two authenticators have the same public
7206 3. Exim used not to create the message log file for a message until the first
7207 delivery attempt. This could be confusing when incoming messages were held
7208 for policy or load reasons. The message log file is now created at the time
7209 the message is received, and an initial "Received" line is written to it.
7211 4. The automatically generated man page for command line options had a minor
7212 bug that caused no ill effects; however, a more serious problem was that
7213 the procedure for building the man page automatically didn't always
7214 operate. Consequently, release 4.20 contains an out-of-date version. This
7215 shouldn't happen again.
7217 5. When building Exim with embedded Perl support, the script that builds the
7218 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7219 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7221 6. The freeze_tell option was not being used for messages that were frozen on
7222 arrival, either by an ACL or by local_scan().
7224 7. Added the smtp_incomplete_transaction log selector.
7226 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7227 was accepting AUTH without a new EHLO.
7229 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7230 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7233 10. Logging of TCP/IP connections (when configured) now happens in the main
7234 daemon process instead of the child process, so that the TCP/IP connection
7235 count is more accurate (but it can never be perfect).
7237 11. The use of "drop" in a nested ACL was not being handled correctly in the
7238 outer ACL. Now, if condition failure induced by the nested "drop" causes
7239 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7240 or "require"), the connection is dropped.
7242 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7243 that yield "discard" can now be used with an "accept" or a "discard" verb,
7244 but an error is generated for any others (because I can't see a useful way
7245 to define what should happen).
7247 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7248 are now processed in the same way as lines in the Exim configuration file.
7249 In particular, continuation lines are supported.
7251 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7253 15. Added -ti meaning -t -i.
7255 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7256 domains, and warn by logging if others are found.
7258 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7259 returned from GIFCONF when Exim is trying to find the list of interfaces on
7260 a host. The code in os.c has been modified to copy each ifreq to an aligned
7261 structure in all cases.
7263 Also, in some cases, the returned ifreq's were being copied to a 'struct
7264 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7265 means the last couple of bytes of an IPv6 address could be chopped if the
7266 ifreq contained only a normal sockaddr (14 bytes storage).
7268 18. Named domain lists were not supported in the hosts_treat_as_local option.
7269 An entry such as +xxxx was not recognized, and was treated as a literal
7272 19. Ensure that header lines added by a DATA ACL are included in the reject log
7273 if the ACL subsequently rejects the message.
7275 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7276 MD5 (which is deprecated).
7278 21. When testing a filter file using -bf, Exim was writing a message when it
7279 took the sender from a "From " line in the message, but it was not doing so
7280 when it took $return_path from a Return-Path: header line. It now does.
7282 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7283 with a valid header line field name (a series of printing characters
7284 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7286 23. Changed "disc" in the source to "disk" to conform to the documentation and
7287 the book and for uniformity.
7289 24. Ignore Sendmail's -Ooption=value command line item.
7291 25. When execve() failed while trying to run a command in a pipe transport,
7292 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7293 could be confused with a return value of 69 from the command itself. This
7294 has been changed to 127, the value the shell returns if it is asked to run
7295 a non-existent command. The wording for the related log line suggests a
7296 non-existent command as the problem.
7298 26. If received_header_text expands to an empty string, do not add a Received:
7299 header line to the message. (Well, it adds a token one on the spool, but
7300 marks it "old" so that it doesn't get used or transmitted.)
7302 27. Installed eximstats 1.28 (addition of -nt option).
7304 28. There was no check for failure on the call to getsockname() in the daemon
7305 code. This can fail if there is a shortage of resources on the system, with
7306 ENOMEM, for example. A temporary error is now given on failure.
7308 29. Contrary to the C standard, it seems that in some environments, the
7309 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7310 program. Exim now does this explicitly; it affects the formatting of
7311 timestamps using strftime().
7313 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7314 complaints. I've now initialized all the variables, to avoid this.
7316 32. Header lines added by a system filter were not being "seen" during
7317 transport-time rewrites.
7319 33. The info_callback() function passed to OpenSSL is set up with type void
7320 (*)(SSL *, int, int), as described somewhere. However, when calling the
7321 function (actually a macro) that sets it up, the type void(*)() is
7322 expected. I've put in a cast to prevent warnings from picky compilers.
7324 34. If a DNS black list lookup found a CNAME record, but there were no A
7325 records associated with the domain it pointed at, Exim crashed.
7327 35. If a DNS black list lookup returned more than one A record, Exim ignored
7328 all but the first. It now scans all returned addresses if a particular IP
7329 value is being sought. In this situation, the contents of the
7330 $dnslist_value variable are a list of all the addresses, separated by a
7333 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7334 to accept a host name and all its aliases if the forward lookup for any of
7335 them yielded the IP address of the incoming connection. Now it accepts only
7336 those names whose forward lookup yields the correct IP address. Any other
7337 names are discarded. This closes a loophole whereby a rogue DNS
7338 administrator could create reverse DNS records to break through a
7339 wildcarded host restriction in an ACL.
7341 37. If a user filter or a system filter that ran in a subprocess used any of
7342 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7343 the wrong values were passed to the pipe command ($thisaddress had the
7344 value of $0, $0 had the value of $1, etc). This bug was introduced by
7345 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7347 38. Improved the line breaking for long SMTP error messages from ACLs.
7348 Previously, if there was no break point between 40 and 75 characters, Exim
7349 left the rest of the message alone. Two changes have been made: (a) I've
7350 reduced the minimum length to 35 characters; (b) if it can't find a break
7351 point between 35 and 75 characters, it looks ahead and uses the first one
7352 that it finds. This may give the occasional overlong line, but at least the
7353 remaining text gets split now.
7355 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7356 file descriptors might be low, and that setting 1000 would always raise it.
7357 It turns out that in some environments, the limit is already over 1000 and
7358 that lowering it causes trouble. So now Exim takes care not to decrease it.
7360 40. When delivering a message, the value of $return_path is set to $sender_
7361 address at the start of routing (routers may change the value). By an
7362 oversight, this default was not being set up when an address was tested by
7363 -bt or -bv, which affected the outcome if any router or filter referred to
7366 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7367 log only when "message" or "log_message" are set. However, if one of the
7368 conditions was an address verification, or a call to a nested ACL, the
7369 messages generated by the underlying test were being passed through. This
7370 no longer happens. The underlying message is available in $acl_verify_
7371 message for both "message" and "log_message" expansions, so it can be
7372 passed through if needed.
7374 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7375 new expansion $bh_ to give the encoded byte string without charset
7376 translation. Translation happens only if iconv() is available; HAVE_ICONV
7377 indicates this at build time. HEADERS_CHARSET gives the charset to
7378 translate to; headers_charset can change it in the configuration, and
7379 "headers charset" can change it in an individual filter file.
7381 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7382 that creates RFC 2047 encoded "words" labels them as that charset instead
7383 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7384 expansion operator; (ii) when Exim creates a From: line for a local
7385 message; (iii) when a header line is rewritten to include a "phrase" part.
7387 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7388 buggy, causing it to skip the first lines of messages whose message ID
7389 ended in 'D'. This would not have bitten before Exim release 4.14, because
7390 message IDs were unlikely to end in 'D' before then. The effect was to have
7391 incorrect size information for certain domains.
7393 45. #include "config.h" was missing at the start of the crypt16.c module. This
7394 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7397 46. If there was a timeout during a "random" callout check, Exim treated it as
7398 a failure of the random address, and carried on sending RSET and the real
7399 address. If the delay was just some slowness somewhere, the response to the
7400 original RCPT would be taken as a response to RSET and so on, causing
7401 mayhem of various kinds.
7403 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7404 when I implemented it. It didn't allow for the fact that some option values
7405 may legitimately be negative (e.g. size_addition), and it didn't even do
7406 the right test for positive values.
7408 48. Domain names in DNS records are case-independent. Exim always looks them up
7409 in lower case. Some resolvers return domain names in exactly the case they
7410 appear in the zone file, that is, they may contain uppercase letters. Not
7411 all resolvers do this - some return always lower case. Exim was treating a
7412 change of case by a resolver as a change of domain, similar to a widening
7413 of a domain abbreviation. This triggered its re-routing code and so it was
7414 trying to route what was effectively the same domain again. This normally
7415 caused routing to fail (because the router wouldn't handle the domain
7416 twice). Now Exim checks for this case specially, and just changes the
7417 casing of the domain that it ultimately uses when it transmits the message
7420 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7423 50. If a filter generated a file delivery with a non-absolute name (possible if
7424 no home directory exists for the router), the forbid_file option was not
7427 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7428 the existing equality matching.
7430 52. Exim was using ints instead of ino_t variables in some places where it was
7431 dealing with inode numbers.
7433 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7434 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7435 called TMPDIR, and if it finds it is different, it changes its value.
7437 54. The smtp_printf() function is now made available to local_scan() so
7438 additional output lines can be written before returning. There is also an
7439 smtp_fflush() function to enable the detection of a dropped connection.
7440 The variables smtp_input and smtp_batched_input are exported to
7443 55. Changed the default runtime configuration: the message "Unknown user"
7444 has been removed from the ACL, and instead placed on the localuser router,
7445 using the cannot_route_message feature. This means that any verification
7446 failures that generate their own messages won't get overridden. Similarly,
7447 the "Unrouteable address" message that was in the ACL for unverifiable
7448 relay addresses has also been removed.
7450 56. Added hosts_avoid_esmtp to the smtp transport.
7452 57. The exicyclog script was not checking for the esoteric option
7453 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7454 will work only if exicyclog is run under the appropriate euid.
7456 58. Following a discussion on the list, the rules by which Exim recognises line
7457 endings on incoming messages have been changed. The -dropcr and drop_cr
7458 options are now no-ops, retained only for backwards compatibility. The
7459 following line terminators are recognized: LF CRLF CR. However, special
7460 processing applies to CR:
7462 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7463 nor a local message in the state where . is a terminator.
7465 (ii) If a bare CR is encountered in a header line, an extra space is added
7466 after the line terminator so as not to end the header. The reasoning
7467 behind this is that bare CRs in header lines are most likely either
7468 to be mistakes, or people trying to play silly games.
7470 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7471 was being incorrectly given as 18 bytes larger than it should have been.
7472 This is a VOB (very old bug).
7474 60. This may never have affected anything current, but just in case it has:
7475 When the local host is found other than at the start of a list of hosts,
7476 the local host, those with the same MX, and any that follow, are discarded.
7477 When the list in question was part of a longer list of hosts, the following
7478 hosts (not currently being processed) were also being discarded. This no
7479 longer happens. I'm not sure if this situation could ever has previously
7482 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7485 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7486 header. This is something that is recommended in a new Internet Draft, and
7487 is something that is documented as being done by Sendmail. There are two
7488 possible values. For messages generated by the autoreply transport, Exim
7491 Auto-Submitted: auto-replied
7493 whereas for all other generated messages (e.g. bounces) it adds
7495 Auto-Submitted: auto-generated
7497 63. The "personal" condition in filters now includes a test for the
7498 Auto-Submitted: header. If it contains the string "auto-" the message it
7499 not considered personal.
7501 64. Added rcpt_include_affixes as a generic transport option.
7503 65. Added queue_only_override (default true).
7505 66. Added the syslog_duplication option.
7507 67. If what should have been the first header line of a message consisted of
7508 a space followed by a colon, Exim was mis-interpreting it as a header line.
7509 It isn't of course - it is syntactically invalid and should therefore be
7510 treated as the start of the message body. The misbehaviour could have
7511 caused a number of strange effects, including loss of data in subsequent
7512 header lines, and spool format errors.
7514 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7515 client host had authenticated. This control can now be exercised by an ACL
7516 for more flexibility.
7518 69. By default, callouts do not happen when testing with -bh. There is now a
7519 variant, -bhc, which does actually run the callout code, including
7520 consulting and updating the callout cache.
7522 70. Added support for saslauthd authentication, courtesy of Alexander
7525 71. If statvfs() failed on the spool or log directories while checking their
7526 size for availability, Exim confusingly gave the error "space shortage".
7527 Furthermore, in debugging mode it crashed with a floating point exception.
7528 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7529 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7530 serious problem, Exim now writes to the main and panic logs when this
7531 happens, with details of the failure. It then refuses to accept the
7532 incoming message, giving the message "spool directory problem" or "log
7533 directory problem" with a 421 code for SMTP messages.
7535 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7536 0, 1, and 2 exist, because some OS complain for execs without them (see
7537 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7538 descriptors. However, the code omitted to check that the open succeeded,
7539 causing mysterious errors if for some reason the permissions on /dev/null
7540 got screwed. Now Exim writes a message to the main and panic logs, and
7541 bombs out if it can't open /dev/null.
7543 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7544 interact so that it is all more flexible. It is supposed to remain
7545 backwards compatible. Also added extra_local_interfaces.
7547 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7548 to bomb out with an assertion failure - to the client this appears as a
7549 connection drop. This problem occurs in the part of the code that was taken
7550 from the Samba project. Fortunately, the assertion is in a very simple
7551 function, so I have fixed this by reproducing the function inline in the
7552 one place where it is called, and arranging for authentication to fail
7553 instead of killing the process with assert().
7555 75. The SPA client code was not working when the server requested OEM rather
7556 than Unicode encoding.
7558 76. Added code to make require_files with a specific uid setting more usable in
7559 the case where statting the file as root fails - usually a non-root-mounted
7560 NFS file system. When this happens and the failure is EACCES, Exim now
7561 forks a subprocess and does the per-uid checking as the relevant uid.
7563 77. Added process_log_path.
7565 78. If log_file_path was not explicitly set, a setting of check_log_space or
7566 check_log_inodes was ignored.
7568 79. If a space check for the spool or log partitions fails, the incident is now
7569 logged. Of course, in the latter case the data may get lost...
7571 80. Added the %p formatting code to string_format() so that it can be used to
7572 print addresses in debug_print(). Adjusted all the address printing in the
7573 debugging in store.c to use %p rather than %d.
7575 81. There was a concern that a line of code in smtp_in.c could overflow a
7576 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7577 initially expressed, the concern was not well-founded, because trailing
7578 spaces are removed early. However, if the trailing spaces were followed by
7579 a NULL, they did not get removed, so the overflow was possible. Two fixes
7582 (a) I re-wrote the offending code in a cleaner fashion.
7583 (b) If an incoming SMTP command contains a NULL character, it is rejected
7586 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7587 runs initgroups(), so that if the Exim user is in any additional groups,
7588 they will be used during message reception.
7594 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob