1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
10 side, to decode a TLS 1.3 packet capture.
12 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
13 Previously the default library behaviour applied, sending two, each in
16 JH/03 Debug output for ACL now gives the config file name and line number for
19 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
21 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
23 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
24 buffer overrun for (non-chunking) other transports.
26 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
27 TLS1.3, means that a server rejecting a client certificate is not visible
28 to the client until the first read of encrypted data (typically the
29 response to EHLO). Add detection for that case and treat it as a failed
30 TLS connection attempt, so that the normal retry-in-clear can work (if
33 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
34 and/or domain. Found and fixed by Jason Betts.
36 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
37 configuration). If a CNAME target was not a wellformed name pattern, a
40 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
41 the OS reports them interleaved with other addresses.
43 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
44 used both for input and for a verify callout, both encrypted, SMTP
45 responses being sent by the server could be lost. This resulted in
46 dropped connections and sometimes bounces generated by a peer sending
49 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
50 string. Previously it was possible to cause undefined behaviour in a
51 library routine (usually a crash). Found by "zerons".
53 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
56 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
57 API was removed, so update to use the newer ones.
59 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
60 any timeout set, is taking a long time. Previously we would hang on to a
61 rotated logfile "forever" if the input was arriving with long gaps
62 (a previous attempt to fix addressed lack, for a long time, of initial
65 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
66 shared (NFS) environment. The length of the tempfile name is now
67 4 + 16 ("hdr.$message_exim_id") which might break on file
68 systems which restrict the file name length to lower values.
71 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
72 shared (NFS) environment.
74 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
75 did for all versions <4.90). Notably -M, -m, --invert, -I may be
78 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
79 on some platforms for bit 31.
81 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
82 to changes apparently associated with TLS1.3 handling some of the APIs
83 previously used were either nonfunctional or inappropriate. Strings
84 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
85 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
86 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
87 This affects log line X= elements, the $tls_{in,out}_cipher variables,
88 and the use of specific cipher names in the encrypted= ACL condition.
90 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
92 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
93 verification result was not updated unless hosts_require_ocsp applied.
95 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
96 queue_list_requires_admin set to false, non-admin users were denied the
99 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
100 directory-of-certs mode. Previously they were advertised despite the
103 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
104 A single TCP connection by a client will now hold a TLS connection open
105 for multiple message deliveries, by default. Previoud the default was to
108 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
109 default. If built with the facility, DANE will be used. The facility
110 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
112 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
113 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
114 must be defined and you must still, unless you define DISABLE_TLS, manage
115 the the include-dir and library-file requirements that go with that
116 choice. Non-TLS builds are still supported.
118 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
121 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
122 default. If the platform supports and has the facility enabled, it will
123 be requested on all coneections.
125 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
126 controlled by the build-time option SUPPORT_PIPE_CONNECT.
128 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
130 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
131 success-DSN messages. Previously the From: header was always the default
132 one for these; the option was ignored.
134 JH/28 Fix the timeout on smtp response to apply to the whole response.
135 Previously it was reset for every read, so a teergrubing peer sending
136 single bytes within the time limit could extend the connection for a
137 long time. Credit to Qualsys Security Advisory Team for the discovery.
139 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
140 delivery address, which leaked information of the results of local
141 forwarding. Change to the original envelope recipient address, per
144 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
145 requested. Previously not bounce was generated and a log entry of
146 error ignored was made.
148 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
150 JH/32 Introduce a general tainting mechanism for values read from the input
151 channel, and values derived from them. Refuse to expand any tainted
152 values, to catch one form of exploit.
154 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
155 was unused and the unexpanded text used for the test. Found and
156 fixed by Ruben Jenster.
158 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
159 an attempt to use a TLS library read routine dereffed a nul pointer,
162 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
163 them in our smtp response.
165 JH/36 Have the general router option retry_use_local_part default to true when
166 any of the restrictive preconditions are set (to anything). Previously it
167 was only for check_local user. The change removes one item of manual
168 configuration which is required for proper retries when a remote router
169 handles a subset of addresses for a domain.
171 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
172 link count into consideration.
174 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
175 caused the extension of big_buffer, the following lines were ignored.
177 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
178 accordance with RFC 2308. Previously there was no expiry, so a longlived
179 receive process (eg. due to ACL delays) versus a short SOA value could
182 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
184 JH/39 Promote DMARC support to mainline.
186 JH/40 Bug 2452: Add a References: header to DSNs.
188 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
189 parameters. The relevant library call is documented as "Deprecated: This
190 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
191 3.6.0, DH parameters are negotiated following RFC7919."
193 HS/06 Change the default of dnssec_request_domains to "*"
195 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
196 carried on and emitted a BDAT command, even when PIPELINING was not
199 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
200 buffer was used for the filename, resulting in a trap when tainted
201 arguments (eg. $domain) were used.
203 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
204 recommended to avoid a possible server-load attack. The feature can be
205 re-enabled via the openssl_options main cofiguration option.
211 JH/01 Remove code calling the customisable local_scan function, unless a new
212 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
214 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
215 non-signal-safe functions being used.
217 JH/03 Bug 2269: When presented with a received message having a stupidly large
218 number of DKIM-Signature headers, disable DKIM verification to avoid
219 a resource-consumption attack. The limit is set at twenty.
221 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
222 report of oldest_pass in ${authres } in consequence, and separate out
223 some descriptions of reasons for verification fail.
225 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
226 files in the spool were present and unlocked. A queue-runner could spot
227 them, resulting in a duplicate delivery. Fix that by doing the unlock
228 after the unlink. Investigation by Tim Stewart. Take the opportunity to
229 add more error-checking on spoolfile handling while that code is being
232 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
233 No known attacks, no CVE, this is defensive hardening.
235 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
236 a queue-runner could start a delivery while other operations were ongoing.
237 Cutthrough delivery was a common victim, resulting in duplicate delivery.
238 Found and investigated by Tim Stewart. Fix by using the open message data
239 file handle rather than opening another, and not locally closing it (which
240 releases a lock) for that case, while creating the temporary .eml format
241 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
243 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
244 $sender_verify_failure/$recipient_verify_failure to "random".
246 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
249 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
250 Previously this would segfault.
252 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
255 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
256 like zero, since the resolver should be doing this for us, But we need one
257 as a CNAME but no MX presence gets the CNAME returned; we need to check
258 that doesn't point to an MX to declare it "no MX returned" rather than
259 "error, loop". A new main option is added so the older capability of
260 following some limited number of chain links is maintained.
262 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
264 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
265 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
266 now-deprecated earlier definitions used only specified the range up to TLS
267 1.2 (in the older-version library docs).
269 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
271 JH/15 Rework TLS client-side context management. Stop using a global, and
272 explicitly pass a context around. This enables future use of TLS for
273 connections to service-daemons (eg. malware scanning) while a client smtp
274 connection is using TLS; with cutthrough connections this is quite likely.
276 JH/16 Fix ARC verification to do AS checks in reverse order.
278 JH/17 Support a "tls" option on the ${readsocket } expansion item.
280 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
281 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
282 Previously the "utf8" would be re-prepended for every additional message.
284 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
285 Previously thery were accepted, resulting in issues when attempting to
286 forward messages to a non-supporting MTA.
288 PP/02 Let -n work with printing macros too, not just options.
290 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
291 one parent address was copied, and bogus data was used at delivery-logging
292 time. Either a crash (after delivery) or bogus log data could result.
293 Discovery and analysis by Tim Stewart.
295 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
296 Previously if the string ended mid-character, we did not insert the
297 promised '?' replacement.
299 PP/04 Documentation: current string operators work on bytes, not codepoints.
301 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
302 should pack well giving a smaller memory footprint so better caching and
303 therefore performance. Group the declarations where this can't be done so
304 that the byte-sized flag variables are not interspersed among pointer
305 variables, giving a better chance of good packing by the compiler.
307 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
308 non-null, to avoid issues with sites running BATV. Previously reports were
309 sent with an empty envelope sender so looked like bounces.
311 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
312 The ignore_error flag wasn't being returned from the filter subprocess so
313 was not set for later routers. Investigation and fix by Matthias Kurz.
315 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
316 and a msg:complete for the whole, when a message is manually removed using
317 -Mrm. Developement by Matthias Kurz, hacked on by JH.
319 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
320 a "Gnu special" function, asprintf() in the DB utility binary builds; I
321 hope that is portable enough.
323 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
324 requiring a known-CA anchor certificate; make it now rely entirely on the
325 TLSA as an anchor. Checking the name on the leaf cert against the name
326 on the A-record for the host is still done for TA (but not for EE mode).
328 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
329 included in delivery lines for non-proxied connections, when compiled with
330 SUPPORT_SOCKS and running with proxy logging enabled.
332 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
333 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
334 move the existing event to fire before the normal logging of message
335 failure so that custom logging is bracketed by normal logging.
337 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
338 msg:fail:internal event. Developement by Matthias Kurz.
340 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
341 far too small for todays use of crypto signatures stored there. Go all
342 the way to the max DNS message size of 64kB, even though this might be
343 overmuch for IOT constrained device use.
345 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
346 copy a string over itself. The library routine is documented as not
347 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
349 JH/32 For main options check_spool_space and check_inode_space, where the
350 platform supports 64b integers, support more than the previous 2^31 kB
351 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
352 the previous G, M, k.
354 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
355 $authenticated_fail_id variable on authentication failure. Previously
358 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
359 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
360 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
361 GNUTLS_SEC_PARAM_MEDIUM.
363 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
364 side. Previously we would continue as if no SNI had been received.
366 JH/36 Harden the handling of string-lists. When a list consisted of a sole
367 "<" character, which should be a list-separator specification, we walked
368 off past the nul-terimation.
370 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
371 causes) even when the retry time is not yet met. Previously they were
372 not, meaning that when (say) an account was over-quota and temp-rejecting,
373 and multiple senders' messages were queued, only one sender would get
374 notified on each configured delay_warning cycle.
376 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
378 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
379 cipher-suites, an error can be left on the stack even for a succeeding
380 accept; this results in impossible error messages when a later operation
383 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
384 return error codes indicating retry. Under TLS1.3 this becomes required.
386 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
387 it only wrote the new authenticators, resulting in a lack of tracking of
388 peer changes of ESMTP extensions until the next cache flush.
390 JH/41 Fix the loop reading a message header line to check for integer overflow,
391 and more-often against header_maxsize. Previously a crafted message could
392 induce a crash of the recive process; now the message is cleanly rejected.
394 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
395 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
401 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
402 When redis_servers is set to a list of > 1 element, and the Redis servers
403 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
404 case of MOVED into a DEFER case instead, thus moving the query onto the
405 next server in the list. For a cluster of N elements, all N servers must
406 be defined in redis_servers.
408 GF/02 Catch and remove uninitialized value warning in exiqsumm
409 Check for existence of @ARGV before looking at $ARGV[0]
411 JH/01 Replace the store_release() internal interface with store_newblock(),
412 which internalises the check required to safely use the old one, plus
413 the allocate and data copy operations duplicated in both (!) of the
414 extant use locations.
416 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
417 modifier. This matches the restriction on the commandline.
419 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
420 Previously only the last row was returned.
422 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
423 we assumed that tags in the header were well-formed, and parsed the
424 element content after inspecting only the first char of the tag.
425 Assumptions at that stage could crash the receive process on malformed
428 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
429 While running the DKIM ACL we operate on the Permanent memory pool so that
430 variables created with "set" persist to the DATA ACL. Also (at any time)
431 DNS lookups that fail create cache records using the Permanent pool. But
432 expansions release any allocations made on the current pool - so a dnsdb
433 lookup expansion done in the DKIM ACL releases the memory used for the
434 DNS negative-cache, and bad things result. Solution is to switch to the
435 Main pool for expansions.
436 While we're in that code, add checks on the DNS cache during store_reset,
437 active in the testsuite.
438 Problem spotted, and debugging aided, by Wolfgang Breyha.
440 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
441 When none of the hosts presented to a transport match an already-open
442 connection, close it and proceed with the list. Previously we would
443 queue the message. Spotted by Lena with Yahoo, probably involving
446 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
447 Previously a spurious "250 OK id=" response was appended to the proper
450 JH/08 The "support for" informational output now, which built with Content
451 Scanning support, has a line for the malware scanner interfaces compiled
452 in. Interface can be individually included or not at build time.
454 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
455 by the template makefile "src/EDITME". The "STREAM" support for an older
456 ClamAV interface method is removed.
458 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
459 rows affected is given instead).
461 JH/11 The runtime Berkeley DB library version is now additionally output by
462 "exim -d -bV". Previously only the compile-time version was shown.
464 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
465 SMTP connection. Previously, when one had more receipients than the
466 first, an abortive onward connection was made. Move to full support for
467 multiple onward connections in sequence, handling cutthrough connection
468 for all multi-message initiating connections.
470 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
471 routers. Previously, a multi-recipient message would fail to match the
472 onward-connection opened for the first recipient, and cause its closure.
474 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
475 a timeout on read on a GnuTLS initiating connection, resulting in the
476 initiating connection being dropped. This mattered most when the callout
477 was marked defer_ok. Fix to keep the two timeout-detection methods
480 JH/15 Relax results from ACL control request to enable cutthrough, in
481 unsupported situations, from error to silently (except under debug)
482 ignoring. This covers use with PRDR, frozen messages, queue-only and
485 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
487 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
488 metadata, resulting in a crash in free().
490 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
491 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
492 Broken also in d185889f4, with init system revamp.
494 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
495 Previously we abruptly closed the connection after reading a malware-
496 found indication; now we go on to read the "scan ok" response line,
499 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
500 ACL. Previously, a crash would result.
502 JH/19 Speed up macro lookups during configuration file read, by skipping non-
503 macro text after a replacement (previously it was only once per line) and
504 by skipping builtin macros when searching for an uppercase lead character.
506 JH/20 DANE support moved from Experimental to mainline. The Makefile control
507 for the build is renamed.
509 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
510 was allocated for every new TLS startup, meaning one per message. Fix
511 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
513 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
514 reported the original. Fix to report (as far as possible) the ACL
515 result replacing the original.
517 JH/23 Fix memory leak during multi-message connections using STARTTLS under
518 OpenSSL. Certificate information is loaded for every new TLS startup,
519 and the resources needed to be freed.
521 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
523 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
526 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
527 DATA response info to the (existing) per-recipient response info for
528 the "C=" log element. It can have useful tracking info from the
529 destination system. Patch from Simon Arlott.
531 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
532 length value. Previously this would segfault.
534 HS/02 Support Avast multiline protoocol, this allows passing flags to
535 newer versions of the scanner.
537 JH/28 Ensure that variables possibly set during message acceptance are marked
538 dead before release of memory in the daemon loop. This stops complaints
539 about them when the debug_store option is enabled. Discovered specifically
540 for sender_rate_period, but applies to a whole set of variables.
541 Do the same for the queue-runner and queue-list loops, for variables set
542 from spool message files. Do the same for the SMTP per-message loop, for
543 certain variables indirectly set in ACL operations.
545 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
546 as a multi-recipient message from a mailinglist manager). The coding had
547 an arbitrary cutoff number of characters while checking for more input;
548 enforced by writing a NUL into the buffer. This corrupted long / fast
549 input. The problem was exposed more widely when more pipelineing of SMTP
550 responses was introduced, and one Exim system was feeding another.
551 The symptom is log complaints of SMTP syntax error (NUL chars) on the
552 receiving system, and refused recipients seen by the sending system
553 (propating to people being dropped from mailing lists).
554 Discovered and pinpointed by David Carter.
556 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
557 replaced by the ${authresults } expansion.
559 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
561 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
562 allows proper process termination in container environments.
564 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
565 Previously the "final dot" had a newline after it; ensure it is CR,LF.
567 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
568 and "err_perm", deprecated since 4.83 when the RFC-defined words
569 "temperror" and "permerror" were introduced.
571 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
572 transport-filters or DKIM-signing. The restriction was lost in the
573 consolidation of verify-callout and delivery SMTP handling.
574 Extend the restriction to also cover ARC-signing.
576 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
577 in defer=pass mode supply a 450 to the initiator. Previously the message
580 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
581 tls_require_ciphers is used as before.
583 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
584 "pass_unscanned". Only tmpfails from the scanner are written to
585 the paniclog, as they may require admin intervention (permission
586 denied, license issues). Other scanner errors (like decompression
587 bombs) do not cause a paniclog entry.
589 JH/36 Fix reinitialisation of DKIM logging variable between messages.
590 Previously it was possible to log spurious information in receive log
593 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
594 triggered odd behaviour from Outlook Express clients.
596 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
599 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
600 since the IETF WG has not yet settled on that versus the original
601 "bare" representation.
603 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
604 Previously the millisecond value corrupted the output.
605 Fix also for syslog_pid=no and log_selector +pid, for which the pid
606 corrupted the output.
612 JH/01 Rework error string handling in TLS interface so that the caller in
613 more cases is responsible for logging. This permits library-sourced
614 string to be attached to addresses during delivery, and collapses
615 pairs of long lines into single ones.
617 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
618 during configuration. Wildcards are allowed and expanded.
620 JH/02 Rework error string handling in DKIM to pass more info back to callers.
621 This permits better logging.
623 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
624 do not close it down and have the child transport start it up again on
625 the passed-on TCP connection. Instead, proxy the child (and any
626 subsequent ones) for TLS via a unix-domain socket channel. Logging is
627 affected: the continued delivery log lines do not have any DNSSEC, TLS
628 Certificate or OCSP information. TLS cipher information is still logged.
630 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
631 identical IP addresses on different listening ports. Will also affect
634 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
635 add noisy ifdef guards to special-case this sillyness.
636 Patch from Bernd Kuhls.
638 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
639 than 255 are no longer allowed.
641 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
642 Disable the session-cache too, which might reduce our load. Since we
643 currrectly use a new context for every connection, both as server and
644 client, there is no benefit for these.
645 GnuTLS appears to not support tickets server-side by default (we don't
646 call gnutls_session_ticket_enable_server()) but client side is enabled
647 by default on recent versions (3.1.3 +) unless the PFS priority string
650 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
651 <https://reproducible-builds.org/specs/source-date-epoch/>.
653 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
654 the check for any unsuccessful recipients did not notice the limit, and
655 erroneously found still-pending ones.
657 JH/08 Pipeline CHUNKING command and data together, on kernels that support
658 MSG_MORE. Only in-clear (not on TLS connections).
660 JH/09 Avoid using a temporary file during transport using dkim. Unless a
661 transport-filter is involved we can buffer the headers in memory for
662 creating the signature, and read the spool data file once for the
663 signature and again for transmission.
665 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
666 4.77 as the kernel support then wasn't solid, having issues in 64bit
667 mode. Now, it's been long enough. Add support for FreeBSD also.
669 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
670 case where the routing stage had gathered several addresses to send to
671 a host before calling the transport for the first, we previously failed
672 to close down TLS in the old transport process before passing the TCP
673 connection to the new process. The new one sent a STARTTLS command
674 which naturally failed, giving a failed delivery and bloating the retry
675 database. Investigation and fix prototype from Wolfgang Breyha.
677 JH/12 Fix check on SMTP command input synchronisation. Previously there were
678 false-negatives in the check that the sender had not preempted a response
679 or prompt from Exim (running as a server), due to that code's lack of
680 awareness of the SMTP input buffering.
682 PP/04 Add commandline_checks_require_admin option.
683 Exim drops privileges sanely, various checks such as -be aren't a
684 security problem, as long as you trust local users with access to their
685 own account. When invoked by services which pass untrusted data to
686 Exim, this might be an issue. Set this option in main configuration
687 AND make fixes to the calling application, such as using `--` to stop
690 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
691 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
692 responses to those, into a single TLS record each way (this usually means
693 a single packet). As a side issue, smtp_enforce_sync now works on TLS
696 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
697 affects you only if you're dancing at the edge of the param size limits.
698 If you are, and this message makes sense to you, then: raise the
699 configured limit or use OpenSSL 1.1. Nothing we can do for older
702 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
703 cmdline element to get the documented default one. Previously it was
706 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
707 are unsuable for non-message-specific reasons - eg. connection timeout,
708 banner-time rejection.
710 JH/16 Fix logging of delivery remote port, when specified by router, under
713 PP/06 Repair manualroute's ability to take options in any order, even if one
714 is the name of a transport.
717 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
719 JH/17 Change the list-building routines interface to use the expanding-string
720 triplet model, for better allocation and copying behaviour.
722 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
723 Previously it was constructed the first time a possibly-matching string
724 was met in the configuration file input during startup; now it is done
727 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
728 compatible one, to avoid the (poorly documented) possibility of a config
729 file in the working directory redirecting the DB files, possibly correpting
730 some existing file. CVE-2017-10140 assigned for BDB.
732 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
733 cache-hot. Previously, although the result was properly cached, the
734 initial verify call returned a defer.
736 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
737 the main verify for receipient in uncached-mode.
739 JH/22 Retire historical build files to an "unsupported" subdir. These are
740 defined as "ones for which we have no current evidence of testing".
742 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
743 if present. Previously it was ignored.
745 JH/24 Start using specified-initialisers in C structure init coding. This is
746 a C99 feature (it's 2017, so now considered safe).
748 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
749 if was a fixed-sized field and bitmask ops via macros; it is now more
752 PP/07 GitHub PR 56: Apply MariaDB build fix.
753 Patch provided by Jaroslav Škarvada.
755 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
756 during Coverity cleanups [4.87 JH/47]
757 Diagnosis and fix provided by Michael Fischer v. Mollard.
759 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
760 the right size to place the terminating semicolon on its own folded
761 line, the header hash was calculated to an incorrect value thanks to
762 the (relaxed) space the fold became.
764 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
765 and confused the parent.
767 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
768 which could crash as a result. This could lead to undeliverable messages.
770 JH/28 Logging: "next input sent too soon" now shows where input was truncated
773 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
774 matters on fast-turnover and PID-randomising systems, which were getting
775 out-of-order delivery.
777 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
778 a possibly-overlapping copy. The symptom was that "Remote host closed
779 connection in response to HELO" was logged instead of the actual 4xx
782 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
783 Previously only that bufferd was discarded, resulting in SYMTP command
786 JH/32 DKIM: when a message has multiple signatures matching an identity given
787 in dkim_verify_signers, run the dkim acl once for each. Previously only
788 one run was done. Bug 2189.
790 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
791 "panic the current process" to "deliberately defer". The panic log is
792 still written with the problem list name; the mail and reject logs now
793 get a temp-reject line for the message that was being handled, saying
794 something like "domains check lookup or other defer". The SMTP 451
795 message is still "Temporary local problem".
797 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
798 A crafted sequence of BDAT commands could result in in-use memory beeing
799 freed. CVE-2017-16943.
801 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
802 from SMTP input. Previously it was always done; now only done for DATA
803 and not BDAT commands. CVE-2017-16944.
805 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
806 to the message (such as an overlong header line). Previously this was
807 not done and we did not exit BDAT mode. Followon from the previous item
808 though a different problem.
814 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
815 than -2003 did; needs libidn2 in addition to libidn.
817 JH/02 The path option on a pipe transport is now expanded before use.
819 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
820 Patch provided by "Björn", documentation fix added too.
822 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
823 missing a wire-to-host endian conversion.
825 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
826 close after a BDAT command line could be taken as a following command,
827 giving a synch failure. Fix by only checking for synch immediately
828 before acknowledging the chunk.
830 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
831 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
832 macro. Patches provided by Josh Soref.
834 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
835 Previously we did not; the RFC seems ambiguous and VRFY is not listed
836 by IANA as a service extension. However, John Klensin suggests that we
839 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
840 the dkim code may be unix-mode line endings rather than smtp wire-format
841 CRLF, so prepend a CR to any bare LF.
843 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
844 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
846 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
847 the first were themselves being wrongly included in the feed into dkim
848 processing; with most chunk sizes in use this resulted in an incorrect
849 body hash calculated value.
851 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
852 DKIM signature block, for verification. Although advised against by
853 standards it is specifically not ruled illegal.
855 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
857 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
858 missing a body hash (the bh= tag).
860 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
861 It seems that HAProxy sends the Proxy Protocol information in clear and
862 only then does a TLS startup, so do the same.
864 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
865 TCP connections (such as for Spamd) unless the daemon successfully set
866 Fast Open mode on its listening sockets. This fixes breakage seen on
867 too-old kernels or those not configured for Fast Open, at the cost of
868 requiring both directions being enabled for TFO, and TFO never being used
869 by non-daemon-related Exim processes.
871 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
872 endings, at least on the first header line. Try to canonify any that get
873 past that check, despite the cost.
875 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
876 now limited to an arbitrary five deep, while parsing addresses with the
877 strip_excess_angle_brackets option enabled.
879 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
880 instead leave the unprompted TLS handshake in socket buffer for the
881 TLS library to consume.
883 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
885 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
887 JH/16 Drop variables when they go out of scope. Memory management drops a whole
888 region in one operation, for speed, and this leaves assigned pointers
889 dangling. Add checks run only under the testsuite which checks all
890 variables at a store-reset and panics on a dangling pointer; add code
891 explicitly nulling out all the variables discovered. Fixes one known
892 bug: a transport crash, where a dangling pointer for $sending_ip_address
893 originally assigned in a verify callout, is re-used.
895 PP/06 Drop '.' from @INC in various Perl scripts.
897 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
899 PP/08 Reduce a number of compilation warnings under clang; building with
900 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
901 should be warning-free.
903 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
905 HS/01 Fix portability problems introduced by PP/08 for platforms where
906 realloc(NULL) is not equivalent to malloc() [SunOS et al].
908 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
909 chunk. This allows us to accept broken chunked messages. We need a more
910 general solution here.
912 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
913 already-broken messages in the queue.
915 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
917 JH/19 Fix reference counting bug in routing-generated-address tracking.
923 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
924 supports it and a size is available (ie. the sending peer gave us one).
926 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
927 years of being deprecated). The replacements are the ACLs
928 acl_smtp_mime and acl_not_smtp_mime.
930 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
931 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
932 or even an in-clear connection were permitted. Now, if the host lookup
933 was dnssec and dane was requested then the host is only used if the
934 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
935 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
936 if one fails this test.
937 This means that a poorly-configured remote DNS will make it incommunicado;
938 but it protects against a DNS-interception attack on it.
940 JH/04 Bug 1810: make continued-use of an open smtp transport connection
941 non-noisy when a race steals the message being considered.
943 JH/05 If main configuration option tls_certificate is unset, generate a
944 self-signed certificate for inbound TLS connections.
946 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
947 in rewrites and routers.
949 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
950 and logged a warning sing 4.83; now they are a configuration file error.
952 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
953 (lacking @domain). Apply the same qualification processing as RCPT.
955 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
957 JH/10 Support ${sha256:} applied to a string (as well as the previous
960 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
961 a cutthrough deliver is pending, as we always want to make a connection.
962 This also avoids re-routing the message when later placing the cutthrough
963 connection after a verify cache hit.
964 Do not update it with the verify result either.
966 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
967 when routing results in more than one destination address.
969 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
970 signing (which inhibits the cutthrough capability). Previously only
971 the presence of an option was tested; now an expansion evaluating as
972 empty is permissible (obviously it should depend only on data available
973 when the cutthrough connection is made).
975 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
976 the relevant preceding SMTP command did not note the pipelining mode.
978 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
979 Previously they were not counted.
981 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
982 as one having no matching records. Previously we deferred the message
983 that needed the lookup.
985 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
986 distinguished as "(=".
988 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
989 for missing MX records. Previously it only worked for missing A records.
991 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
993 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
994 after the data-go-ahead and data-ack. Patch from Jason Betts.
996 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
997 even for a "none" policy. Patch from Tony Meyer.
999 JH/22 Fix continued use of a connection for further deliveries. If a port was
1000 specified by a router, it must also match for the delivery to be
1003 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1004 When one of the recipients of a message was unsuitable for the connection
1005 (has no matching addresses), we lost track of needing to mark it
1006 deferred. As a result mail would be lost.
1008 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1010 JH/25 Decoding ACL controls is now done using a binary search; the source code
1011 takes up less space and should be simpler to maintain. Merge the ACL
1012 condition decode tables also, with similar effect.
1014 JH/26 Fix problem with one_time used on a redirect router which returned the
1015 parent address unchanged. A retry would see the parent address marked as
1016 delivered, so not attempt the (identical) child. As a result mail would
1019 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1020 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1021 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1024 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1025 Main config options check_{log,spool}_{inodes,space} are now
1026 100 inodes, 10MB unless set otherwise in the configuration.
1028 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1029 Previously it only applied to the main-section connection policy
1032 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1034 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1035 by me. Added RFC7919 DH primes as an alternative.
1037 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1038 are not in the system include path.
1040 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1041 GnuTLS, when a session startup failed (eg because the client disconnected)
1042 Exim did stdio operations after fclose. This was exposed by a recent
1043 change which nulled out the file handle after the fclose.
1045 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1046 signed directly by the cert-signing cert, rather than an intermediate
1047 OCSP-signing cert. This is the model used by LetsEncrypt.
1049 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1051 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1052 an incoming connection.
1054 HS/02 Bug 1802: Do not half-close the connection after sending a request
1057 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1058 fallback to "prime256v1".
1060 JH/34 SECURITY: Use proper copy of DATA command in error message.
1061 Could leak key material. Remotely exploitable. CVE-2016-9963.
1067 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1068 and 3.4.4 - once the server is enabled to respond to an OCSP request
1069 it does even when not requested, resulting in a stapling non-aware
1070 client dropping the TLS connection.
1072 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1073 support variable-length bit vectors. No functional change.
1075 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1076 The I= interface field on outgoing lines is now after the H= remote
1077 host field, same as incoming lines. There is a separate
1078 outgoing_interface log selector which allows you to disable the
1081 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1082 If not running log_selector +smtp_connection the mainlog would be held
1083 open indefinitely after a "too many connections" event, including to a
1084 deleted file after a log rotate. Leave the per net connection logging
1085 leaving it open for efficiency as that will be quickly detected by the
1086 check on the next write.
1088 HS/01 Bug 1671: Fix post transport crash.
1089 Processing the wait-<transport> messages could crash the delivery
1090 process if the message IDs didn't exist for some reason. When
1091 using 'split_spool_directory=yes' the construction of the spool
1092 file name failed already, exposing the same netto behaviour.
1094 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1095 mime_regex ACL conditions.
1097 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1098 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1099 to HELO, local diagnostic string.
1101 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1102 log line to debug. Even when configured with a tls authenticator many
1103 client connections are expected to not authenticate in this way, so
1104 an authenticate fail is not an error.
1106 HS/02 Add the Exim version string to the process info. This way exiwhat
1107 gives some more detail about the running daemon.
1109 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1110 matter for fast-change records such as DNSBLs.
1112 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1113 retry record, even if constant. There may be multiple transports with
1114 different interface settings and the retry behaviour needs to be kept
1117 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1119 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1121 JH/10 Bug 840: fix log_defer_output option of pipe transport
1123 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1124 hosts_randomize. This matters a lot when combined with mua_wrapper.
1126 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1127 ${quote_pgsql:<string>} operator.
1129 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1131 JH/14 Change header file name for freeradius-client. Relevant if compiling
1132 with Radius support; from the Gentoo tree and checked under Fedora.
1134 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1136 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1137 option transport_filter, meaning no filtering.
1139 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1141 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1142 defaults to "*" (all hosts). The variable is now available when not built
1143 with TLS, default unset, mainly to enable keeping the testsuite sane.
1144 If a server certificate is not supplied (via tls_certificate) an error is
1145 logged, and clients will find TLS connections fail on startup. Presumably
1146 they will retry in-clear.
1147 Packagers of Exim are strongly encouraged to create a server certificate
1148 at installation time.
1150 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1151 with the $config_file variable.
1153 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1154 in transport context, after the attempt, and per-recipient. The latter type
1155 is per host attempted. The event data is the error message, and the errno
1156 information encodes the lookup type (A vs. MX) used for the (first) host,
1157 and the trailing two digits of the smtp 4xx response.
1159 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1160 to write to mainlog (or rejectlog, paniclog) in the window between file
1161 creation and permissions/ownership being changed. Particularly affects
1162 installations where exicyclog is run as root, rather than exim user;
1163 result is that the running daemon panics and dies.
1165 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1167 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1168 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1169 "pri" and "weight". Note that the previous implicit priority given by the
1170 list order is no longer honoured.
1172 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1173 for DKIM processing.
1175 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1176 by defining SUPPORT_SOCKS.
1178 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1179 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1180 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1181 variables are renamed to proxy_{local,external}_{address,port}.
1183 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1184 for a build by defining SUPPORT_I18N
1186 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1187 of the query string, and make ${quote_redis:} do that quoting.
1189 JH/29 Move Events support from Experimental to mainline, enabled by default
1190 and removable for a build by defining DISABLE_EVENT.
1192 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1194 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1195 cached by the daemon.
1197 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1198 by defining LOOKUP_REDIS. The libhiredis library is required.
1200 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1201 keys are given for lookup.
1203 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1204 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1205 only supported when built with TLS support. The PolarSSL SHA routines
1206 are still used when the TLS library is too old for convenient support.
1208 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1209 openssl_options), for security. OpenSSL forces this from version 1.1.0
1210 server-side so match that on older versions.
1212 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1213 allocation for $value could be released as the expansion processing
1214 concluded, but leaving the global pointer active for it.
1216 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1217 and to use the domains and local_parts ACL conditions.
1219 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1220 incorrectly not doubled on cutthrough transmission, hence seen as a
1221 body-termination at the receiving system - resulting in truncated mails.
1222 Commonly the sender saw a TCP-level error, and retransmitted the message
1223 via the normal store-and-forward channel. This could result in duplicates
1224 received - but deduplicating mailstores were liable to retain only the
1225 initial truncated version.
1227 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1229 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1231 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1232 we're in there, support oversigning also; bug 1309.
1234 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1236 HS/04 Add support for keep_environment and add_environment options.
1238 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1239 either intentional arithmetic overflow during PRNG, or testing config-
1242 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1243 delivery resulted in actual delivery. Cancel cutthrough before DATA
1246 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1247 rejects a recipient: pass the reject to the originator.
1249 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1250 Many were false-positives and ignorable, but it's worth fixing the
1253 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1254 for the new environment-manipulation done at startup. Move the routines
1255 from being local to tls.c to being global via the os.c file.
1257 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1258 an extract embedded as result-arg for a map, the first arg for extract
1259 is unavailable so we cannot tell if this is a numbered or keyed
1260 extraction. Accept either.
1266 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1269 JH/02 The smtp transport option "multi_domain" is now expanded.
1271 JH/03 The smtp transport now requests PRDR by default, if the server offers
1274 JH/04 Certificate name checking on server certificates, when exim is a client,
1275 is now done by default. The transport option tls_verify_cert_hostnames
1276 can be used to disable this per-host. The build option
1277 EXPERIMENTAL_CERTNAMES is withdrawn.
1279 JH/05 The value of the tls_verify_certificates smtp transport and main options
1280 default to the word "system" to access the system default CA bundle.
1281 For GnuTLS, only version 3.0.20 or later.
1283 JH/06 Verification of the server certificate for a TLS connection is now tried
1284 (but not required) by default. The verification status is now logged by
1285 default, for both outbound TLS and client-certificate supplying inbound
1288 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1291 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1292 Status Notification (bounce) messages are now MIME format per RFC 3464.
1293 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1294 under the control of the dsn_advertise_hosts option, and routers may
1295 have a dsn_lasthop option.
1297 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1298 default, modifiable by a malware= option. The list separator for
1299 the options can now be changed in the usual way. Bug 68.
1301 JH/10 The smtp_receive_timeout main option is now expanded before use.
1303 JH/11 The incoming_interface log option now also enables logging of the
1304 local interface on delivery outgoing connections.
1306 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1307 if the interface and destination host and port all match.
1309 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1312 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1313 Patch from Andrew Lewis.
1315 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1316 now supports optional time-restrictions, weighting, and priority
1317 modifiers per server. Patch originally by <rommer@active.by>.
1319 JH/16 The spamd_address main option now supports a mixed list of local
1320 and remote servers. Remote servers can be IPv6 addresses, and
1321 specify a port-range.
1323 JH/17 Bug 68: The spamd_address main option now supports an optional
1324 timeout value per server.
1326 JH/18 Bug 1581: Router and transport options headers_add/remove can
1327 now have the list separator specified.
1329 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1332 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1335 JH/21 Support for the A6 type of dns record is withdrawn.
1337 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1338 rather than the verbs used.
1340 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1341 from 255 to 1024 chars.
1343 JH/24 Verification callouts now attempt to use TLS by default.
1345 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1346 are generic router options now. The defaults didn't change.
1348 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1349 Original patch from Alexander Shikoff, worked over by JH.
1351 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1352 files if ZCAT_COMMAND is not executable.
1354 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1356 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1358 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1359 Normally benign, it bites when the pair was led to by a CNAME;
1360 modern usage is to not canonicalize the domain to a CNAME target
1361 (and we were inconsistent anyway for A-only vs AAAA+A).
1363 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1365 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1366 when evaluating $sender_host_dnssec.
1368 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1369 $sender_helo_dnssec variable.
1371 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1373 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1375 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1377 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1378 documented as working, but never had. Support all but $spam_report.
1380 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1381 added for tls authenticator.
1383 HS/03 Add perl_taintmode main config option
1389 TL/01 When running the test suite, the README says that variables such as
1390 no_msglog_check are global and can be placed anywhere in a specific
1391 test's script, however it was observed that placement needed to be near
1392 the beginning for it to behave that way. Changed the runtest perl
1393 script to read through the entire script once to detect and set these
1394 variables, reset to the beginning of the script, and then run through
1395 the script parsing/test process like normal.
1397 TL/02 The BSD's have an arc4random API. One of the functions to induce
1398 adding randomness was arc4random_stir(), but it has been removed in
1399 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1400 function when detected.
1402 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1403 cause callback expansion.
1405 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1406 syntax errors in an expansion can be treated as a string instead of
1407 logging or causing an error, due to the internal use of bool_lax
1408 instead of bool when processing it.
1410 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1411 server certificates when making smtp deliveries.
1413 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1415 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1417 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1419 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1420 Merged patch from Sebastian Wiedenroth.
1422 JH/05 Fix results-pipe from transport process. Several recipients, combined
1423 with certificate use, exposed issues where response data items split
1424 over buffer boundaries were not parsed properly. This eventually
1425 resulted in duplicates being sent. This issue only became common enough
1426 to notice due to the introduction of connection certificate information,
1427 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1429 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1430 size buffer was used, resulting in syntax errors when an expansion
1433 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1434 version 3.3.6 or later.
1436 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1437 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1438 both become "event_action", the variables become $event_name, $event_data
1439 and $event_defer_errno. There is a new variable $verify_mode, usable in
1440 routers, transports and related events. The tls:cert event is now also
1441 raised for inbound connections, if the main configuration event_action
1444 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1445 early OCSP support, but no stapling (appears to be less than 1.0.0).
1447 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1448 server certificate names available under the smtp transport option
1449 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1452 JH/10 Time-related extraction expansions from certificates now use the main
1453 option "timezone" setting for output formatting, and are consistent
1454 between OpenSSL and GnuTLS compilations. Bug 1541.
1456 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1457 encoded parameter in the incoming message. Bug 1558.
1459 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1460 include certificate info, eximon was claiming there were spoolfile
1463 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1465 JH/14 Log delivery-related information more consistently, using the sequence
1466 "H=<name> [<ip>]" wherever possible.
1468 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1469 are problematic for Debian distribution, omit them from the release
1472 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1474 JH/16 Fix string representation of time values on 64bit time_t architectures.
1477 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1478 output list separator was used.
1483 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1484 checkers that were complaining about end of non-void function with no
1487 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1488 This was a regression introduced in 4.83 by another bugfix.
1490 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1492 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1493 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1499 TF/01 Correctly close the server side of TLS when forking for delivery.
1501 When a message was received over SMTP with TLS, Exim failed to clear up
1502 the incoming connection properly after forking off the child process to
1503 deliver the message. In some situations the subsequent outgoing
1504 delivery connection happened to have the same fd number as the incoming
1505 connection previously had. Exim would try to use TLS and fail, logging
1506 a "Bad file descriptor" error.
1508 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1509 utilities have not been installed.
1511 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1512 temporary space as the ACL may create new global variables.
1514 TL/01 LDAP support uses per connection or global context settings, depending
1515 upon the detected version of the libraries at build time.
1517 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1518 to extract and use the src ip:port in logging and expansions as if it
1519 were a direct connection from the outside internet. PPv2 support was
1520 updated based on HAProxy spec change in May 2014.
1522 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1524 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1525 Properly escape header and check for NULL return.
1527 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1530 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1532 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1533 characters in header names, implemented as a verify condition.
1534 Contributed by Michael Fischer v. Mollard.
1536 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1537 results permerror and temperror. Previous values are deprecated but
1538 still accepted. In a future release, err_perm and err_temp will be
1539 completely removed, which will be a backward incompatibility if the
1540 ACL tests for either of these two old results. Patch contributed by
1541 user bes-internal on the mailing list.
1543 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1545 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1546 selectors, in both main and reject logs.
1548 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1551 JH/07 Add malware type "sock" for talking to simple daemon.
1553 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1555 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1556 routers/transports under cutthrough routing.
1558 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1559 numbers. Touch up "bool" conditional to keep the same definition.
1561 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1563 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1565 JH/12 Expand items in router/transport headers_add or headers_remove lists
1566 individually rather than the list as a whole. Bug 1452.
1568 Required for reasonable handling of multiple headers_ options when
1569 they may be empty; requires that headers_remove items with embedded
1570 colons must have them doubled (or the list-separator changed).
1572 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1573 view the policy declared in the DMARC record. Currently, $dmarc_status
1574 is a combined value of both the record presence and the result of the
1577 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1579 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1580 dnslookup router and the smtp transport (applying to the forward
1583 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1584 of ldap servers used for a specific lookup. Patch provided by Heiko
1587 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1588 New variable $lookup_dnssec_authenticated for observability.
1590 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1591 Patch submitted by Lars Timman.
1593 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1595 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1596 Requires trusted mode and valid format message id, aborts otherwise.
1597 Patch contributed by Heiko Schlichting.
1599 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1600 certextract with support for various fields. Bug 1358.
1602 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1603 is requested by default, modifiable by smtp transport option
1606 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1607 operate on certificate variables to give certificate fingerprints
1608 Also new ${sha256:cert_variable}.
1610 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1612 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1613 Christian Aistleitner.
1615 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1617 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1618 file. Patch from Wolfgang Breyha.
1620 JH/25 Expand the coverage of the delivery $host and $host_address to
1621 client authenticators run in verify callout. Bug 1476.
1623 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1624 align with daemon_smtp_ports. Bug 72.
1626 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1627 support and error reporting did not work properly.
1629 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1630 and is readable. Patch from Andrew Colin Kissa.
1632 TL/14 Enhance documentation of ${run expansion and how it parses the
1633 commandline after expansion, particularly in the case when an
1634 unquoted variable expansion results in an empty value.
1636 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1638 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1639 by renaming away from it. Addresses GH issue 15, reported by
1642 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1644 TL/15 SECURITY: prevent double expansion in math comparison functions
1645 (can expand unsanitized data). Not remotely exploitable.
1652 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1654 PP/02 Make -n do something, by making it not do something.
1655 When combined with -bP, the name of an option is not output.
1657 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1660 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1661 $sender_host_name and config options to manage this, and basic check
1664 PP/05 DSCP support for outbound connections and control modifier for inbound.
1666 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1667 (Only plugin which currently uses this is kerberos4, which nobody should
1668 be using, but we should make it available and other future plugins might
1669 conceivably use it, even though it would break NAT; stuff *should* be
1670 using channel bindings instead).
1672 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1673 name; added for Sendmail compatibility; requires admin caller.
1674 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1675 just ignore it); requires trusted caller.
1676 Also parse but ignore: -Ac -Am -X<logfile>
1679 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1681 TL/02 Add +smtp_confirmation as a default logging option.
1683 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1684 Patch by Magnus Holmgren from 2007-02-20.
1686 TL/04 Bugzilla 1281 - Spec typo.
1687 Bugzilla 1283 - Spec typo.
1688 Bugzilla 1290 - Spec grammar fixes.
1690 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1692 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1694 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1695 mailing list by by Dmitry Isaikin.
1697 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1699 JH/02 Support "G" suffix to numbers in ${if comparisons.
1701 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1703 NM/01 Bugzilla 1197 - Spec typo
1704 Bugzilla 1196 - Spec examples corrections
1706 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1708 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1709 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1712 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1713 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1715 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1716 "acl {{name}{arg}...}", and optional args on acl condition
1719 JH/05 Permit multiple router/transport headers_add/remove lines.
1721 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1723 JH/07 Avoid using a waiting database for a single-message-only transport.
1724 Performance patch from Paul Fisher. Bugzilla 1262.
1726 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1729 JH/09 Add $headers_added variable, with content from use of ACL modifier
1730 add_header (but not yet added to the message). Bugzilla 199.
1732 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1733 Pulled from Bugzilla 817 by Wolfgang Breyha.
1735 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1737 (nb: this is the same fix as in Exim 4.80.1)
1739 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1742 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1745 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1747 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1748 advertise SMTP AUTH mechanism to us, instead of a generic
1749 protocol violation error. Also, make Exim more robust to bad
1750 data from the Dovecot auth socket.
1752 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1754 When a queue runner is handling a message, Exim first routes the
1755 recipient addresses, during which it prunes them based on the retry
1756 hints database. After that it attempts to deliver the message to
1757 any remaining recipients. It then updates the hints database using
1760 So if a recipient address works intermittently, it can get repeatedly
1761 deferred at routing time. The retry hints record remains fresh so the
1762 address never reaches the final cutoff time.
1764 This is a fairly common occurrence when a user is bumping up against
1765 their storage quota. Exim had some logic in its local delivery code
1766 to deal with this. However it did not apply to per-recipient defers
1767 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1769 This change adds a proper retry rule check during routing so that the
1770 final cutoff time is checked against the message's age. We only do
1771 this check if there is an address retry record and there is not a
1772 domain retry record; this implies that previous attempts to handle
1773 the address had the retry_use_local_parts option turned on. We use
1774 this as an approximation for the destination being like a local
1775 delivery, as in LMTP.
1777 I suspect this new check makes the old local delivery cutoff check
1778 redundant, but I have not verified this so I left the code in place.
1780 TF/02 Correct gecos expansion when From: is a prefix of the username.
1782 Test 0254 submits a message to Exim with the header
1786 When I ran the test suite under the user fanf2, Exim expanded
1787 the header to contain my full name, whereas it should have added
1788 a Resent-Sender: header. It erroneously treats any prefix of the
1789 username as equal to the username.
1791 This change corrects that bug.
1793 GF/01 DCC debug and logging tidyup
1794 Error conditions log to paniclog rather than rejectlog.
1795 Debug lines prefixed by "DCC: " to remove any ambiguity.
1797 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1799 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1800 Bug spotted by Jeremy Harris; was flawed since initial commit.
1801 Would have resulted in OCSP responses post-SNI triggering an Exim
1802 NULL dereference and crash.
1804 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1806 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1807 Bug detection, analysis and fix by Samuel Thibault.
1808 Bugzilla 1331, Debian bug #698092.
1810 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1812 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1813 Server implementation by Todd Lyons, client by JH.
1814 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1815 config variable "prdr_enable" controls whether the server
1816 advertises the facility. If the client requests PRDR a new
1817 acl_data_smtp_prdr ACL is called once for each recipient, after
1818 the body content is received and before the acl_smtp_data ACL.
1819 The client is controlled by both of: a hosts_try_prdr option
1820 on the smtp transport, and the server advertisement.
1821 Default client logging of deliveries and rejections involving
1822 PRDR are flagged with the string "PRDR".
1824 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1825 fclose(). Diagnosis by Todd Lyons.
1827 PP/17 Update configure.default to handle IPv6 localhost better.
1828 Patch by Alain Williams (plus minor tweaks).
1831 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1832 This is now consistent with GnuTLS, and is now documented: the
1833 previous undocumented portable approach to treating the option as
1834 unset was to force an expansion failure. That still works, and
1835 an empty string is now equivalent.
1837 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1838 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1839 not performing validation itself.
1841 PP/20 Added force_command boolean option to pipe transport.
1842 Patch from Nick Koston, of cPanel Inc.
1844 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1847 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1849 PP/21 Fix eximon continuous updating with timestamped log-files.
1850 Broken in a format-string cleanup in 4.80, missed when I repaired the
1851 other false fix of the same issue.
1852 Report and fix from Heiko Schlichting.
1855 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1856 Report from Prashanth Katuri.
1858 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1859 It's SecureTransport, so affects any MacOS clients which use the
1860 system-integrated TLS libraries, including email clients.
1862 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1863 using a MIME ACL for non-SMTP local injection.
1864 Report and assistance in diagnosis by Warren Baker.
1866 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1868 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1870 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1871 last id that failed so it may be referenced in subsequent ACL's.
1873 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1876 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1877 ldap library initialization, allowing self-signed CA's to be
1878 used. Also properly sets require_cert option later in code by
1879 using NULL (global ldap config) instead of ldap handle (per
1880 session). Bug diagnosis and testing by alxgomz.
1882 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1883 the src/util/ subdirectory.
1885 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1886 renamed to Transport Post Delivery Action by Jeremy Harris, as
1889 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1890 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1891 redis_servers = needs to be configured which will be used by the redis
1892 lookup. Patch from Warren Baker, of The Packet Hub.
1894 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1896 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1897 hostname or reverse DNS when processing a host list. Used suggestions
1898 from multiple comments on this bug.
1900 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1902 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1903 Missed a few lines, added it to make the runtest require no keyboard
1906 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1907 contains upper case chars. Make router use caseful_local_part.
1909 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1910 support when GnuTLS has been built with p11-kit.
1916 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1918 This, or similar/improved, will also be change PP/11 of 4.82.
1924 PP/01 Handle short writes when writing local log-files.
1925 In practice, only affects FreeBSD (8 onwards).
1926 Bugzilla 1053, with thanks to Dmitry Isaikin.
1928 NM/01 Bugzilla 949 - Documentation tweak
1930 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1933 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1935 PP/02 Implemented gsasl authenticator.
1937 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1939 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1940 `pkg-config foo` for cflags/libs.
1942 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1943 with rest of GSASL and with heimdal_gssapi.
1945 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1946 `pkg-config foo` for cflags/libs for the TLS implementation.
1948 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1949 properties get this fed in as external SSF. A number of robustness
1950 and debugging improvements to the cyrus_sasl authenticator.
1952 PP/08 cyrus_sasl server now expands the server_realm option.
1954 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1955 Patch by Jeremy Harris.
1957 PP/10 Added dbmjz lookup type.
1959 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1961 PP/12 MAIL args handles TAB as well as SP, for better interop with
1962 non-compliant senders.
1963 Analysis and variant patch by Todd Lyons.
1965 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1966 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1967 Patch from Dirk Mueller <dmueller@suse.com>
1969 PP/13 tls_peerdn now print-escaped for spool files.
1970 Observed some $tls_peerdn in wild which contained \n, which resulted
1971 in spool file corruption.
1973 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1974 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1975 or write after TLS renegotiation, which otherwise led to messages
1978 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1979 as a tracking header (ie: a signed header comes before the signature).
1980 Patch from Wolfgang Breyha.
1982 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1983 comma-sep list; embedded commas doubled.
1985 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1987 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1989 Report and patch from Dmitry Banschikov.
1991 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1992 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1993 failures. We appear to now support TLS1.1+ with Exim.
1995 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1996 lets Exim select keys and certificates based upon TLS SNI from client.
1997 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1998 before an outbound SMTP session. New log_selector, +tls_sni.
2000 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2001 NULL dereference. Report and patch from Alun Jones.
2003 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2004 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2005 Not seeing resolver debug output on NetBSD, but suspect this is a
2006 resolver implementation change.
2008 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2009 Left warnings. Added "eximon gdb" invocation mode.
2011 PP/21 Defaulting "accept_8bitmime" to true, not false.
2013 PP/22 Added -bw for inetd wait mode support.
2015 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2016 locate the relevant includes and libraries. Made this the default.
2018 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2019 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2021 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2022 This may cause build issues on older platforms.
2024 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2025 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2026 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2027 Added SNI support via GnuTLS too.
2028 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2030 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2032 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2033 Applied second patch from Janne, implementing suggestion to default
2034 multiple-strings-in-record handling to match SPF spec.
2036 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2038 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2039 read-only, out of scope).
2040 Patch from Wolfgang Breyha, report from Stuart Northfield.
2042 PP/29 Fix three issues highlighted by clang analyser static analysis.
2043 Only crash-plausible issue would require the Cambridge-specific
2044 iplookup router and a misconfiguration.
2045 Report from Marcin Mirosław.
2047 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2049 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2050 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2051 As part of this, removing so much warning spew let me fix some minor
2052 real issues in debug logging.
2054 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2055 assignment on my part. Fixed.
2057 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2058 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2059 Janne Snabb (who went above and beyond: thank you).
2061 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2062 string otherwise requires a connection and a bunch more work and it's
2063 relatively easy to get wrong. Should also expose TLS library linkage
2066 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2067 64-bit ${eval} (JH/03).
2069 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2070 GNU libc to support some of the 64-bit stuff, should not lead to
2071 conflicts. Defined before os.h is pulled in, so if a given platform
2072 needs to override this, it can.
2074 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2075 protection layer was required, which is not implemented.
2076 Bugzilla 1254, patch from Wolfgang Breyha.
2078 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2079 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2080 tls_dhparam take prime identifiers. Also unbreak combination of
2081 OpenSSL+DH_params+TLSSNI.
2083 PP/39 Disable SSLv2 by default in OpenSSL support.
2089 PP/01 Solaris build fix for Oracle's LDAP libraries.
2090 Bugzilla 1109, patch from Stephen Usher.
2092 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2094 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2097 TF/02 Fix a couple more cases where we did not log the error message
2098 when unlink() failed. See also change 4.74-TF/03.
2100 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2101 lock up or crash if it happened to be inside a call to libc when it
2102 got a SIGUSR1 from exiwhat.
2104 The SIGUSR1 handler appends the current process status to the process
2105 log which is later printed by exiwhat. It used to use the general
2106 purpose logging code to do this, but several functions it calls are
2107 not safe for signals.
2109 The new output code in the SIGUSR1 handler is specific to the process
2110 log, and simple enough that it's easy to inspect for signal safety.
2111 Removing some special cases also simplifies the general logging code.
2112 Removing the spurious timestamps from the process log simplifies
2115 TF/04 Improved ratelimit ACL condition.
2117 The /noupdate option has been deprecated in favour of /readonly which
2118 has clearer semantics. The /leaky, /strict, and /readonly update modes
2119 are mutually exclusive. The update mode is no longer included in the
2120 database key; it just determines when the database is updated. (This
2121 means that when you upgrade Exim will forget old rate measurements.)
2123 Exim now checks that the per_* options are used with an update mode that
2124 makes sense for the current ACL. For example, when Exim is processing a
2125 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2126 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2127 must specify per_mail/readonly. If you omit the update mode it defaults to
2128 /leaky where that makes sense (as before) or /readonly where required.
2130 The /noupdate option is now undocumented but still supported for
2131 backwards compatibility. It is equivalent to /readonly except that in
2132 ACLs where /readonly is required you may specify /leaky/noupdate or
2133 /strict/noupdate which are treated the same as /readonly.
2135 A useful new feature is the /count= option. This is a generalization
2136 of the per_byte option, so that you can measure the throughput of other
2137 aggregate values. For example, the per_byte option is now equivalent
2138 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2140 The per_rcpt option has been generalized using the /count= mechanism
2141 (though it's more complicated than the per_byte equivalence). When it is
2142 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2143 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2144 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2145 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2146 Note that using per_rcpt with a non-readonly update mode in more than
2147 one ACL will cause the recipients to be double-counted. (The per_mail
2148 and per_byte options don't have this problem.)
2150 The handling of very low rates has changed slightly. If the computed rate
2151 is less than the event's count (usually one) then this event is the first
2152 after a long gap. In this case the rate is set to the same as this event's
2153 count, so that the first message of a spam run is counted properly.
2155 The major new feature is a mechanism for counting the rate of unique
2156 events. The new per_addr option counts the number of different
2157 recipients that someone has sent messages to in the last time period. It
2158 behaves like per_rcpt if all the recipient addresses are different, but
2159 duplicate recipient addresses do not increase the measured rate. Like
2160 the /count= option this is a general mechanism, so the per_addr option
2161 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2162 example, measure the rate that a client uses different sender addresses
2163 with the options per_mail/unique=$sender_address. There are further
2164 details in the main documentation.
2166 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2168 TF/06 Removed a few PCRE remnants.
2170 TF/07 Automatically extract Exim's version number from tags in the git
2171 repository when doing development or release builds.
2173 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2174 Bugzilla 879. Patch from Paul Fisher.
2176 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2177 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2180 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2182 PP/05 Variable $av_failed, true if the AV scanner deferred.
2183 Bugzilla 1078. Patch from John Horne.
2185 PP/06 Stop make process more reliably on build failure.
2186 Bugzilla 1087. Patch from Heiko Schlittermann.
2188 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2189 Bugzilla 1089. Patch from Heiko Schlittermann.
2191 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2192 Bugzilla 1131. Patch from Holger Weiß.
2194 PP/09 Handle IPv6 addresses with SPF.
2195 Bugzilla 860. Patch from Wolfgang Breyha.
2197 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2199 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2202 PP/11 match_* no longer expand right-hand-side by default.
2203 New compile-time build option, EXPAND_LISTMATCH_RHS.
2204 New expansion conditions, "inlist", "inlisti".
2206 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2208 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2210 PP/14 fix log_write() format string regression from TF/03.
2211 Bugzilla 1152. Patch from Dmitry Isaikin.
2217 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2219 PP/02 Harmonised TLS library version reporting; only show if debugging.
2220 Layout now matches that introduced for other libraries in 4.74 PP/03.
2222 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2224 PP/04 New "dns_use_edns0" global option.
2226 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2229 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2230 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2232 TK/01 Updated PolarSSL code to 0.14.2.
2233 Bugzilla 1097. Patch from Andreas Metzler.
2235 PP/07 Catch divide-by-zero in ${eval:...}.
2236 Fixes bugzilla 1102.
2238 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2241 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2242 format-string attack -- SECURITY: remote arbitrary code execution.
2244 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2245 time unintentionally subject to list matching rules, letting the header
2246 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2247 arbitrary string expansion). This allowed for information disclosure.
2249 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2250 INT_MIN/-1 -- value coerced to INT_MAX.
2256 NM/01 Workaround for PCRE version dependency in version reporting
2259 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2260 This fixes portability to compilers other than gcc, notably
2261 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2263 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2264 makefiles for portability to HP-UX and POSIX correctness.
2266 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2267 Also via indented variable definition in the Makefile.
2268 (Debugging by Oliver Heesakkers).
2270 PP/02 Restore caching of spamd results with expanded spamd_address.
2271 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2273 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2274 Improves build reliability. Fix from: Frank Elsner
2276 NM/02 Fix wide character breakage in the rfc2047 coding
2277 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2279 NM/03 Allow underscore in dnslist lookups
2280 Fixes bug 1026. Patch from Graeme Fowler
2282 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2283 Code patches from Adam Ciarcinski of NetBSD.
2285 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2288 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2289 is logged, to avoid truncation. Patch from John Horne.
2291 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2292 Patch from Jakob Hirsch.
2294 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2295 SQL string expansion failure details.
2296 Patch from Andrey Oktyabrski.
2298 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2299 Patch from Simon Arlott.
2301 PP/09 New lookups functionality failed to compile on old gcc which rejects
2302 extern declarations in function scope.
2303 Patch from Oliver Fleischmann
2305 PP/10 Use sig_atomic_t for flags set from signal handlers.
2306 Check getgroups() return and improve debugging.
2307 Fixed developed for diagnosis in bug 927 (which turned out to be
2310 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2311 Patch from Mark Zealey.
2313 PP/12 Bugzilla 1056: Improved spamd server selection.
2314 Patch from Mark Zealey.
2316 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2317 Based on patch from Heiko Schlittermann.
2319 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2320 Patch from Uwe Doering, sign-off by Michael Haardt.
2322 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2323 variable declaration deep within a block. Bug and patch from
2326 PP/15 lookups-Makefile IRIX compatibility coercion.
2328 PP/16 Make DISABLE_DKIM build knob functional.
2330 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2331 Patch by Simon Arlott
2333 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2334 variable argument macros. Our copy now differs from upstream.
2340 TF/01 Failure to get a lock on a hints database can have serious
2341 consequences so log it to the panic log.
2343 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2344 controlled using the smtp_confirmation log selector.
2346 TF/03 Include the error message when we fail to unlink a spool file.
2348 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2349 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2350 for maintaining out-of-tree patches for some time.
2352 PP/01 Bugzilla 139: Documentation and portability issues.
2353 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2354 Handle per-OS dynamic-module compilation flags.
2356 PP/02 Let /dev/null have normal permissions.
2357 The 4.73 fixes were a little too stringent and complained about the
2358 permissions on /dev/null. Exempt it from some checks.
2359 Reported by Andreas M. Kirchwitz.
2361 PP/03 Report version information for many libraries, including
2362 Exim version information for dynamically loaded libraries. Created
2363 version.h, now support a version extension string for distributors
2364 who patch heavily. Dynamic module ABI change.
2366 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2367 privilege escalation vulnerability whereby the Exim run-time user
2368 can cause root to append content of the attacker's choosing to
2371 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2374 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2375 If dropping privileges for untrusted macros, we disabled normal logging
2376 on the basis that it would fail; for the Exim run-time user, this is not
2377 the case, and it resulted in successful deliveries going unlogged.
2378 Fixed. Reported by Andreas Metzler.
2384 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2385 only prepend for the Resent-* case. Fixes regression introduced in
2386 Exim 4.70 by NM/22 for Bugzilla 607.
2388 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2389 increasing numbers of administrators be bitten by this.
2391 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2393 PP/03 Bugzilla 994: added openssl_options main configuration option.
2395 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2397 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2399 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2401 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2402 without a peer certificate, leading to a segfault because of an
2403 assumption that peers always have certificates. Be a little more
2404 paranoid. Problem reported by Martin Tscholak.
2406 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2407 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2408 NB: ClamAV planning to remove STREAM in "middle of 2010".
2409 CL also introduces -bmalware, various -d+acl logging additions and
2410 more caution in buffer sizes.
2412 PP/09 Implemented reverse_ip expansion operator.
2414 PP/10 Bugzilla 937: provide a "debug" ACL control.
2416 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2418 PP/12 Bugzilla 973: Implement --version.
2420 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2422 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2424 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2426 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2427 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2428 ignore trailing whitespace.
2430 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2432 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2435 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2436 Notification from Dr Andrew Aitchison.
2438 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2439 ExtendedDetectionInfo response format.
2440 Notification from John Horne.
2442 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2445 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2446 XSL and documented dependency on system catalogs, with examples of how
2449 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2452 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2453 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2454 configuration file which is writeable by the Exim user or group.
2456 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2457 of configuration files to cover files specified with the -C option if
2458 they are going to be used with root privileges, not just the default
2461 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2462 option (effectively making it always true).
2464 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2465 files to be used while preserving root privileges.
2467 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2468 that rogue child processes cannot use them.
2470 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2471 run-time user, instead of root.
2473 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2474 Exim run-time user without dropping privileges.
2476 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2477 result string, instead of calling string_vformat() twice with the same
2480 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2481 for other users. Others should always drop root privileges if they use
2482 -C on the command line, even for a whitelisted configure file.
2484 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2486 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2492 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2493 $data_path, and $header_path variables; fixed documentation bugs and
2496 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2497 exipick to access non-standard spools, including the "frozen" queue
2500 NM/01 Bugzilla 965: Support mysql stored procedures.
2501 Patch from Alain Williams
2503 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2505 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2506 Patch from Andreas Metzler
2508 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2509 Patch from Kirill Miazine
2511 NM/05 Bugzilla 671: Added umask to procmail example.
2513 JJ/03 installed exipick 20100323.0, fixing doc bug
2515 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2516 directory. Notification and patch from Dan Rosenberg.
2518 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2520 TK/02 Improve log output when DKIM signing operation fails.
2522 MH/01 Treat the transport option dkim_domain as a colon separated
2523 list, not as a single string, and sign the message with each element,
2524 omitting multiple occurences of the same signer.
2526 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2527 Bugzilla 985, 986. Patch by Simon Arlott
2529 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2530 Patch by Simon Arlott
2532 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2533 MBX locking. Notification from Dan Rosenberg.
2539 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2541 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2543 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2545 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2547 NM/04 Bugzilla 915: Flag broken perl installation during build.
2553 TK/01 Added patch by Johannes Berg that expands the main option
2554 "spamd_address" if it starts with a dollar sign.
2556 TK/02 Write list of recipients to X-Envelope-Sender header when building
2557 the mbox-format spool file for content scanning (suggested by Jakob
2560 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2561 (http://www.dcc-servers.net/) support via dccifd. Activated by
2562 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2564 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2565 by Mark Daniel Reidel <mr@df.eu>.
2567 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2568 When building exim an external PCRE library is now needed -
2569 PCRE is a system library on the majority of modern systems.
2570 See entry on PCRE_LIBS in EDITME file.
2572 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2573 conversation. Added nologin parameter to request.
2574 Patch contributed by Kirill Miazine.
2576 TF/01 Do not log submission mode rewrites if they do not change the address.
2578 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2580 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2581 log files in place. Contributed by Roberto Lima.
2583 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2585 TF/03 Bugzilla 615: When checking the local_parts router precondition
2586 after a local_part_suffix or local_part_prefix option, Exim now
2587 does not use the address's named list lookup cache, since this
2588 contains cached lookups for the whole local part.
2590 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2591 Robert Millan. Documentation is in experimental-spec.txt.
2593 TF/04 Bugzilla 668: Fix parallel build (make -j).
2595 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2597 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2598 Patch provided by Jan Srzednicki.
2600 TF/05 Leading white space used to be stripped from $spam_report which
2601 wrecked the formatting. Now it is preserved.
2603 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2604 that they are available at delivery time.
2606 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2608 TF/08 TLS error reporting now respects the incoming_interface and
2609 incoming_port log selectors.
2611 TF/09 Produce a more useful error message if an SMTP transport's hosts
2612 setting expands to an empty string.
2614 NM/06 Bugzilla 744: EXPN did not work under TLS.
2615 Patch provided by Phil Pennock.
2617 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2618 Patch provided by Richard Godbee.
2620 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2621 acl_smtp_notquit, added index entry.
2623 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2624 Patch provided by Eugene Bujak.
2626 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2627 accept(). Patch provided by Maxim Dounin.
2629 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2630 Patch provided by Phil Pennock.
2632 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2634 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2635 Patch provided by Brad "anomie" Jorsch.
2637 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2638 Patch provided by Dean Brooks.
2640 TK/05 Add native DKIM support (does not depend on external libraries).
2642 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2643 Patch provided by Graeme Fowler.
2645 NM/16 Bugzilla 851: Documentation example syntax fix.
2647 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2649 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2652 NM/19 Bugzilla 745: TLS version reporting.
2653 Patch provided by Phil Pennock.
2655 NM/20 Bugzilla 167: bool: condition support.
2656 Patch provided by Phil Pennock.
2658 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2659 clients. Patch provided by Phil Pennock.
2661 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2662 Patch provided by Brad "anomie" Jorsch.
2664 NM/23 Bugzilla 687: Fix misparses in eximstats.
2665 Patch provided by Heiko Schlittermann.
2667 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2668 Patch provided by Heiko Schlittermann.
2670 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2671 plus update to original patch.
2673 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2675 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2676 Patch provided by David Brownlee.
2678 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2680 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2682 NM/30 Bugzilla 888: TLS documentation bugfixes.
2684 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2686 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2687 Unlike the original bugzilla I have changed all shell scripts in src tree.
2689 NM/33 Bugzilla 898: Transport filter timeout fix.
2690 Patch by Todd Rinaldo.
2692 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2693 Patch by Serge Demonchaux.
2695 NM/35 Bugzilla 39: Base64 decode bug fixes.
2696 Patch by Jakob Hirsch.
2698 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2700 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2702 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2704 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2710 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2711 ALT-N's libdkim that I have put here:
2712 http://duncanthrax.net/exim-experimental/
2714 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2715 were called 'true' and it seems that C99 defines that as a reserved
2716 keyword to be used with 'bool' variable types. That means you could
2717 not include C99-style headers which use bools without triggering
2718 build errors in sieve.c.
2720 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2721 as mailq or other aliases. Changed the --help handling significantly
2722 to do whats expected. exim_usage() emits usage/help information.
2724 SC/01 Added the -bylocaldomain option to eximstats.
2726 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2728 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2730 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2736 PH/01 Another patch from the Sieve maintainer.
2738 PH/02 When an IPv6 address is converted to a string for single-key lookup
2739 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2740 dots are used instead of colons so that keys in lsearch files need not
2741 contain colons. This was done some time before quoting was made available
2742 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2743 (notated using the quote facility) so as to distinguish them from IPv4
2744 keys. This meant that lookups for IP addresses in host lists did not work
2745 for iplsearch lookups.
2747 This has been fixed by arranging for IPv6 addresses to be expressed with
2748 colons if the lookup type is iplsearch. This is not incompatible, because
2749 previously such lookups could never work.
2751 The situation is now rather anomalous, since one *can* have colons in
2752 ordinary lsearch keys. However, making the change in all cases is
2753 incompatible and would probably break a number of configurations.
2755 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2758 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2759 conversion specification without a maximum field width, thereby enabling
2760 a rogue spamd server to cause a buffer overflow. While nobody in their
2761 right mind would setup Exim to query an untrusted spamd server, an
2762 attacker that gains access to a server running spamd could potentially
2763 exploit this vulnerability to run arbitrary code as the Exim user.
2765 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2766 $primary_hostname instead of what libspf2 thinks the hosts name is.
2768 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2769 a directory entry by the name of the lookup key. Previously, if a
2770 symlink pointed to a non-existing file or a file in a directory that
2771 Exim lacked permissions to read, a lookup for a key matching that
2772 symlink would fail. Now it is enough that a matching directory entry
2773 exists, symlink or not. (Bugzilla 503.)
2775 PH/03 The body_linecount and body_zerocount variables are now exported in the
2778 PH/04 Added the $dnslist_matched variable.
2780 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2781 This means they are set thereafter only if the connection becomes
2784 PH/06 Added the client_condition to authenticators so that some can be skipped
2785 by clients under certain conditions.
2787 PH/07 The error message for a badly-placed control=no_multiline_responses left
2788 "_responses" off the end of the name.
2790 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2792 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2793 (without spaces) instead of just copying the configuration text.
2795 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2797 PH/11 Added $max_received_linelength.
2799 PH/12 Added +ignore_defer and +include_defer to host lists.
2801 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2802 way in which PCRE > 7.0 is built.
2804 PH/14 Implemented queue_only_load_latch.
2806 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2807 MAIL command. The effect was to mangle the value on 64-bit systems.
2809 PH/16 Another patch from the Sieve maintainer.
2811 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2813 PH/18 If a system quota error occurred while trying to create the file for
2814 a maildir delivery, the message "Mailbox is full" was not appended to the
2815 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2816 only to a quota excession during the actual writing of the file.
2818 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2819 characters?) which causes problems in log lines. The DN values are now
2820 passed through string_printing() before being added to log lines.
2822 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2823 and InterBase are left for another time.)
2825 PH/21 Added message_body_newlines option.
2827 PH/22 Guard against possible overflow in moan_check_errorcopy().
2829 PH/23 POSIX allows open() to be a macro; guard against that.
2831 PH/24 If the recipient of an error message contained an @ in the local part
2832 (suitably quoted, of course), incorrect values were put in $domain and
2833 $local_part during the evaluation of errors_copy.
2839 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2840 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2843 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2844 issue a MAIL command.
2846 PH/02 In an ACL statement such as
2848 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2850 if a client was not listed at all, or was listed with a value other than
2851 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2852 the condition was not true (as it should be), so access was not denied.
2853 The bug was that the ! inversion was incorrectly passed on to the second
2854 item. This has been fixed.
2856 PH/03 Added additional dnslists conditions == and =& which are different from
2857 = and & when the dns lookup returns more than one IP address.
2859 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2860 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2862 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2863 FSYNC, which compiles an option called disable_fsync that allows for
2864 bypassing fsync(). The documentation is heavily laced with warnings.
2866 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2868 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2869 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2870 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2871 including adding "make clean"; (3) Added -fPIC when compiling the test
2872 dynamically loaded module, to get rid of a warning.
2874 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2875 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2876 = 0s. The bug is otherwise harmless.
2878 PH/07 There was a bug in the dovecot authenticator such that the value of
2879 $auth1 could be overwritten, and so not correctly preserved, after a
2880 successful authentication. This usually meant that the value preserved by
2881 the server_setid option was incorrect.
2883 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2885 PH/09 Installed PCRE release 7.0.
2887 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2888 run for batched SMTP input. It is now run at the start of every message
2889 in the batch. While fixing this I discovered that the process information
2890 (output by running exiwhat) was not always getting set for -bs and -bS
2891 input. This is fixed, and it now also says "batched" for BSMTP.
2893 PH/11 Added control=no_pipelining.
2895 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2896 patch, slightly modified), and move the expansion of helo_data till after
2897 the connection is made in the smtp transport (so it can use these
2900 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2902 PH/14 Added log_selector = +pid.
2904 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2906 PH/16 Add ${if forany and ${if forall.
2908 PH/17 Added dsn_from option to vary the From: line in DSNs.
2910 PH/18 Flush SMTP output before performing a callout, unless control =
2911 no_callout_flush is set.
2913 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2914 was true (the default) a successful delivery failed to delete the retry
2915 item, thus causing premature timeout of the address. The bug is now
2918 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2920 PH/21 Long custom messages for fakedefer and fakereject are now split up
2921 into multiline responses in the same way that messages for "deny" and
2922 other ACL rejections are.
2924 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2925 with slight modification.
2927 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2928 draft, changing the syntax and factoring some duplicate code".
2930 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2931 for deliveries of the second and subsequent messages over the same SMTP
2934 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2935 ${reduce, with only minor "tidies".
2937 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2939 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2940 expansion side effects.
2942 PH/27 When a message times out after an over-quota error from an Exim-imposed
2943 quota, the bounce message says "mailbox is full". This message was not
2944 being given when it was a system quota that was exceeded. It now should
2947 MH/03 Made $recipients available in local_scan(). local_scan() already has
2948 better access to the recipient list through recipients_list[], but
2949 $recipients can be useful in postmaster-provided expansion strings.
2951 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2952 in the case of a MAIL command with additional options following the
2953 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2954 were accidentally chopped off.
2956 PH/29 SMTP synchronization checks are implemented when a command is read -
2957 there is a check that no more input is waiting when there shouldn't be
2958 any. However, for some commands, a delay in an ACL can mean that it is
2959 some time before the response is written. In this time, more input might
2960 arrive, invalidly. So now there are extra checks after an ACL has run for
2961 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2962 pipelining has not been advertised.
2964 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2966 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2967 returned in the SMTP response when smtp_return_error_details was false.
2968 This has been fixed.
2970 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2971 instead of the C I/O that was originally supplied, because problems were
2972 reported on Solaris.
2974 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2975 Exim which did not show up earlier: it was assuming that a call to
2976 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2977 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2978 was a macro that became an assignment, so it seemed to work. This has
2979 changed to a proper function call with a void return, hence the compile
2980 error. Exim's code has been fixed.
2982 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2985 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2987 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2989 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2990 because it turns out to be too big or there is a timeout) there is no
2991 "Completed" line in the log. When some messages of this type were
2992 selected by exigrep, they were listed as "not completed". Others were
2993 picked up by some special patterns. I have improved the selection
2994 criteria to be more general.
2996 PH/38 The host_find_failed option in the manualroute router can now be set
2997 to "ignore", to completely ignore a host whose IP address cannot be
2998 found. If all hosts are ignored, the behaviour is controlled by the new
2999 host_all_ignored option.
3001 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3002 homing or because of multiple MX records with /mx) generated more than
3003 one IP address, and the following item turned out to be the local host,
3004 all the secondary addresses of the first item were incorrectly removed
3005 from the list, along with the local host and any following hosts (which
3006 is what is supposed to happen).
3008 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3009 whoever called Exim into the -H file. In the case of the daemon it was
3010 behaving confusingly. When first started, it used values for whoever
3011 started the daemon, but after a SIGHUP it used the Exim user (because it
3012 calls itself on a restart). I have changed the code so that it now always
3015 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3016 message are rejected with the same error (e.g. no authentication or bad
3017 sender address), and a DATA command is nevertheless sent (as can happen
3018 with PIPELINING or a stupid MUA), the error message that was given to the
3019 RCPT commands is included in the rejection of the DATA command. This is
3020 intended to be helpful for MUAs that show only the final error to their
3023 PH/42 Another patch from the Sieve maintainer.
3025 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3026 Eximstats - Fixed some broken HTML links and added missing column headers
3028 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3029 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3031 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3033 PH/43 Yet another patch from the Sieve maintainer.
3035 PH/44 I found a way to check for a TCP/IP connection going away before sending
3036 the response to the final '.' that terminates a message, but only in the
3037 case where the client has not sent further data following the '.'
3038 (unfortunately, this is allowed). However, in many cases there won't be
3039 any further data because there won't be any more messages to send. A call
3040 to select() can be used: if it shows that the input is "ready", there is
3041 either input waiting, or the socket has been closed. An attempt to read
3042 the next input character can distinguish the two cases. Previously, Exim
3043 would have sent an OK response which the client would never have see.
3044 This could lead to message repetition. This fix should cure that, at
3045 least in a lot of common cases.
3047 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3048 advertised in response to EHLO.
3054 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3055 fixed by 4.65/MH/01 (is this a record?) are fixed:
3057 (i) An empty string was always treated as zero by the numeric comparison
3058 operators. This behaviour has been restored.
3060 (ii) It is documented that the numeric comparison operators always treat
3061 their arguments as decimal numbers. This was broken in that numbers
3062 starting with 0 were being interpreted as octal.
3064 While fixing these problems I realized that there was another issue that
3065 hadn't been noticed. Values of message_size_limit (both the global option
3066 and the transport option) were treated as octal if they started with 0.
3067 The documentation was vague. These values are now always treated as
3068 decimal, and I will make that clear in the documentation.
3074 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3075 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3078 MH/01 Don't check that the operands of numeric comparison operators are
3079 integers when their expansion is in "skipping" mode (fixes bug
3080 introduced by 4.64-PH/07).
3082 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3083 child addresses, Exim now panics and dies. Previously, because the count
3084 is held in a short int, deliveries were likely to be lost. As such a
3085 large number of recipients for a single message is ridiculous
3086 (performance will be very, very poor), I have chosen to impose a limit
3087 rather than extend the field.
3093 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3094 leftover -K file (the existence of which was triggered by #402).
3095 While we were at it, introduced process PID as part of the -K
3096 filename. This should rule out race conditions when creating
3099 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3100 processing considerably. Previous code took too long for large mails,
3101 triggering a timeout which in turn triggers #401.
3103 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3104 in the DK code in transports.c. sendfile() is not really portable,
3105 hence the _LINUX specificness.
3107 TF/01 In the add_headers option to the mail command in an Exim filter,
3108 there was a bug that Exim would claim a syntax error in any
3109 header after the first one which had an odd number of characters
3112 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3113 callout verification, Exim cached a "reject" for the entire domain. This
3114 is correct for most verifications, but it is not correct for a recipient
3115 verification with use_sender or use_postmaster set, because in that case
3116 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3117 case of MAIL FROM:<> rejection from other early rejections (e.g.
3118 rejection of HELO). When verifying a recipient using a non-null MAIL
3119 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3120 Whatever the result of the callout, the value of the domain cache is
3121 left unchanged (for any other kind of callout, getting as far as trying
3122 RCPT means that the domain itself is ok).
3124 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3127 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3128 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3129 session, the daemon ignores ECONNECT errors and logs others; it now
3130 ignores EPIPE as well.
3132 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3133 (quoted-printable decoding).
3135 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3136 later the small subsequent patch to fix an introduced bug.
3138 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3140 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3142 PH/08 An error is now given if message_size_limit is specified negative.
3144 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3145 to be given (somewhat) arbitrary names.
3147 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3150 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3151 miscellaneous code fixes
3153 PH/10 Added the log_reject_target ACL modifier to specify where to log
3156 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3157 hostname. This is wrong, because it relates to the incoming message (and
3158 probably the interface on which it is arriving) and not to the outgoing
3159 callout (which could be using a different interface). This has been
3160 changed to use the value of the helo_data option from the smtp transport
3161 instead - this is what is used when a message is actually being sent. If
3162 there is no remote transport (possible with a router that sets up host
3163 addresses), $smtp_active_hostname is used.
3165 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3166 tweaks were necessary in order to get it to work (see also 21 below):
3167 (a) The code assumed that strncpy() returns a negative number on buffer
3168 overflow, which isn't the case. Replaced with Exim's string_format()
3170 (b) There were several signed/unsigned issues. I just did the minimum
3171 hacking in of casts. There is scope for a larger refactoring.
3172 (c) The code used strcasecmp() which is not a standard C function.
3173 Replaced with Exim's strcmpic() function.
3174 (d) The code set only $1; it now sets $auth1 as well.
3175 (e) A simple test gave the error "authentication client didn't specify
3176 service in request". It would seem that Dovecot has changed its
3177 interface. Fortunately there's a specification; I followed it and
3178 changed what the client sends and it appears to be working now.
3180 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3183 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3184 address A is aliased to B and C, where B exists and C does not. Without
3185 -v the output is "A verified" because verification stops after a
3186 successful redirection if more than one address is generated. However,
3187 with -v the child addresses are also verified. Exim was outputting "A
3188 failed to verify" and then showing the successful verification for C,
3189 with its parentage. It now outputs "B failed to verify", showing B's
3190 parentage before showing the successful verification of C.
3192 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3193 look up a TXT record in a specific list after matching in a combined
3196 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3197 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3198 they consult the DNS. I had assumed they would set it the way they
3199 wanted; and indeed my experiments on Linux seem to show that in some
3200 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3201 To be on the safe side, however, I have now made the interface to
3202 host_find_byname() similar to host_find_bydns(), with an argument
3203 containing the DNS resolver options. The host_find_byname() function now
3204 sets these options at its start, just as host_find_bydns() does. The smtp
3205 transport options dns_qualify_single and dns_search_parents are passed to
3206 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3207 of host_find_byname() use the default settings of RES_DEFNAMES
3208 (qualify_single) but not RES_DNSRCH (search_parents).
3210 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3211 spool_read_header() do less string testing, by means of a preliminary
3212 switch on the second character of optional "-foo" lines. (This is
3213 overdue, caused by the large number of possibilities that now exist.
3214 Originally there were few.) While I was there, I also converted the
3215 str(n)cmp tests so they don't re-test the leading "-" and the first
3216 character, in the hope this might squeeze out yet more improvement.
3218 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3219 flag allowing group syntax was set by the header_syntax check but not
3220 turned off, possible causing trouble later; (2) The flag was not being
3221 set at all for the header_verify test, causing "group"-style headers to
3222 be rejected. I have now set it in this case, and also caused header_
3223 verify to ignore an empty address taken from a group. While doing this, I
3224 came across some other cases where the code for allowing group syntax
3225 while scanning a header line wasn't quite right (mostly, not resetting
3226 the flag correctly in the right place). These bugs could have caused
3227 trouble for malformed header lines. I hope it is now all correct.
3229 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3230 with the "reply" argument non-NULL. The code, however (which originally
3231 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3232 but it didn't always do it. This confused somebody who was copying the
3233 code for some other use. I have removed all the tests.
3235 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3236 feature that was used to support insecure browsers during the U.S. crypto
3237 embargo. It requires special client support, and Exim is probably the
3238 only MTA that supported it -- and would never use it because real RSA is
3239 always available. This code has been removed, because it had the bad
3240 effect of slowing Exim down by computing (never used) parameters for the
3241 RSA_EXPORT functionality.
3243 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3244 authenticator to fail if there's a tab character in the incoming data
3245 (there should never be unless someone is messing about, as it's supposed
3246 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3247 if the connection is using TLS or if the remote IP is the same as the
3248 local IP, and the "valid-client-cert option" if a client certificate has
3251 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3252 authenticators. This can be used for authorization after authentication
3253 succeeds. (In the case of plaintext, it servers for both authentication
3256 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3257 if any retry times were supplied.
3259 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3260 connection, where there is no client IP address to check. In this
3261 situation, the verify now always succeeds.
3263 PH/25 Applied John Jetmore's -Mset patch.
3265 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3267 PH/27 In a string expansion for a processed (not raw) header when multiple
3268 headers of the same name were present, leading whitespace was being
3269 removed from all of them, but trailing whitespace was being removed only
3270 from the last one. Now trailing whitespace is removed from each header
3271 before concatenation. Completely empty headers in a concatenation (as
3272 before) are ignored.
3274 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3275 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3277 PH/29 [Removed. This was a change that I later backed out, and forgot to
3278 correct the ChangeLog entry (that I had efficiently created) before
3279 committing the later change.]
3281 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3282 address errors (4xx response to RCPT) over the same connection as other
3283 messages routed to the same hosts. Such deliveries are always "forced",
3284 so retry times are not inspected. This resulted in far too many retries
3285 for the affected addresses. The effect occurred only when there were more
3286 hosts than the hosts_max_try setting in the smtp transport when it had
3287 the 4xx errors. Those hosts that it had tried were not added to the list
3288 of hosts for which the message was waiting, so if all were tried, there
3289 was no problem. Two fixes have been applied:
3291 (i) If there are any address or message errors in an SMTP delivery, none
3292 of the hosts (tried or untried) are now added to the list of hosts
3293 for which the message is waiting, so the message should not be a
3294 candidate for sending over the same connection that was used for a
3295 successful delivery of some other message. This seems entirely
3296 reasonable: after all the message is NOT "waiting for some host".
3297 This is so "obvious" that I'm not sure why it wasn't done
3298 previously. Hope I haven't missed anything, but it can't do any
3299 harm, as the worst effect is to miss an optimization.
3301 (ii) If, despite (i), such a delivery is accidentally attempted, the
3302 routing retry time is respected, so at least it doesn't keep
3303 hammering the server.
3305 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3306 in ${readsocket because some servers need this prod.
3308 PH/32 Added some extra debug output when updating a wait-xxx database.
3310 PH/33 The hint "could be header name not terminated by colon", which has been
3311 given for certain expansion errors for a long time, was not being given
3312 for the ${if def:h_colon_omitted{... case.
3314 PH/34 The spec says: "With one important exception, whenever a domain list is
3315 being scanned, $domain contains the subject domain." There was at least
3316 one case where this was not true.
3318 PH/35 The error "getsockname() failed: connection reset by peer" was being
3319 written to the panic log as well as the main log, but it isn't really
3320 panic-worthy as it just means the connection died rather early on. I have
3321 removed the panic log writing for the ECONNRESET error when getsockname()
3324 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3325 runs only) independently of the message's sender address. This meant
3326 that, if the 4xx error was in fact related to the sender, a different
3327 message to the same recipient with a different sender could confuse
3328 things. In particular, this can happen when sending to a greylisting
3329 server, but other circumstances could also provoke similar problems.
3330 I have changed the default so that the retry time for these errors is now
3331 based a combination of the sender and recipient addresses. This change
3332 can be overridden by setting address_retry_include_sender=false in the
3335 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3336 remote server are returned as part of bounce messages. This was not
3337 happening for LMTP over a pipe (the lmtp transport), but now it is the
3338 same for both kinds of LMTP.
3340 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3341 in header lines that were in fact CNAMEs. This is no longer the case.
3343 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3344 and queue runs started by the daemon processed all messages. This has
3345 been fixed so that -R and -S can now usefully be given with -q<time>.
3347 PH/40 Import PCRE release 6.7 (fixes some bugs).
3349 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3351 PH/42 Give an error if -q is specified more than once.
3353 PH/43 Renamed the variables $interface_address and $interface_port as
3354 $received_ip_address and $received_port, to make it clear that these
3355 values apply to message reception, and not to the outgoing interface when
3356 a message is delivered. (The old names remain recognized, of course.)
3358 PH/44 There was no timeout on the connect() call when using a Unix domain
3359 socket in the ${readsocket expansion. There now is.
3361 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3362 be meaningful with "accept".
3364 SC/01 Eximstats V1.43
3365 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3367 SC/02 Eximstats V1.44
3368 Use a glob alias rather than an array ref in the generated
3369 parser. This improves both readability and performance.
3371 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3372 Collect SpamAssassin and rejection statistics.
3373 Don't display local sender or destination tables unless
3374 there is data to show.
3375 Added average volumes into the top table text output.
3377 SC/04 Eximstats V1.46
3378 Collect data on the number of addresses (recipients)
3379 as well as the number of messages.
3381 SC/05 Eximstats V1.47
3382 Added 'Message too big' to the list of mail rejection
3383 reasons (thanks to Marco Gaiarin).
3385 SC/06 Eximstats V1.48
3386 Mainlog lines which have GMT offsets and are too short to
3387 have a flag are now skipped.
3389 SC/07 Eximstats V1.49 (Alain Williams)
3390 Added the -emptyok flag.
3392 SC/08 Eximstats V1.50
3393 Fixes for obtaining the IP address from reject messages.
3395 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3396 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3397 whitespace changes from 4.64-PH/27
3399 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3402 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3403 are found, allow negative numbers in numeric criteria)
3405 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3407 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3408 to match changes made in 4.64-PH/43
3410 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3412 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3413 "domain" is actually an IP address, and force a failure. This locks out
3414 those revolvers/nameservers that support "A-for-A" lookups, in
3415 contravention of the specifications.
3417 PH/48 When a host name was looked up from an IP address, and the subsequent
3418 forward lookup of the name timed out, the host name was left in
3419 $sender_host_name, contrary to the specification.
3421 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3422 restricted to single-key lookups, Exim was not diagnosing an error if
3423 * or *@ was used with a query-style lookup.
3425 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3427 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3428 long ago, but noone interested enough thought of it. Let's just say that
3429 the "1.1" means that there are some new functions that weren't there at
3430 some point in the past.
3432 PH/51 Error processing for expansion failure of helo_data from an smtp
3433 transport during callout processing was broken.
3435 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3436 tested/used via the -bh/-bhc/-bs options.
3438 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3439 bug, fixed in subsequent PCRE releases).
3441 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3442 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3444 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3450 SC/01 Use a glob alias rather than an array ref in eximstats generated
3451 parser. This improves both readability and performance.
3453 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3454 Don't display local sender or destination tables in eximstats unless
3455 there is data to show.
3456 Added average volumes into the eximstats top table text output.
3458 SC/03 Collect data on the number of addresses (recipients) as well
3459 as the number of messages in eximstats.
3461 TF/01 Correct an error in the documentation for the redirect router. Exim
3462 does (usually) call initgroups() when daemonizing.
3464 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3465 with consistent privilege compared to when running as a daemon.
3467 TF/03 Note in the spec that $authenticated_id is not set for local
3468 submissions from trusted users.
3470 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3471 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3473 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3474 by adding some example configuration directives to the default
3475 configuration file. A little bit of work is required to uncomment the
3476 directives and define how usernames and passwords are checked, but
3477 there is now a framework to start from.
3479 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3480 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3481 without this. I don't know how relevant this is to other LDAP libraries.
3483 PH/02 Add the verb name to the "unknown ACL verb" error.
3485 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3487 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3489 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3490 directory not expanded when it should be if an expanded home directory
3491 was set for the address (which is overridden by the transport).
3493 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3496 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3497 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3498 because it is too late at that time, and has no effect.
3500 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3501 security issue with \' (bugzilla #107). I could not use the
3502 PQescapeStringConn() function, because it needs a PGconn value as one of
3505 PH/08 When testing addresses using -bt, indicate those final addresses that
3506 are duplicates that would not cause an additional delivery. At least one
3507 person was confused, thinking that -bt output corresponded to deliveries.
3508 (Suppressing duplicates isn't a good idea as you lose the information
3509 about possibly different redirections that led to the duplicates.)
3511 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3512 systems where poll() doesn't work, in particular OS X.
3514 PH/10 Added more information to debugging output for retry time not reached.
3516 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3517 operations in malware.c.
3519 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3522 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3523 syslog_duplication set false, log lines that would normally be written
3524 both the the main log and to the reject log were not written to syslog at
3527 PH/14 In the default configuration, change the use of "message" in ACL warn
3528 statements to "add_header".
3530 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3531 not followed by a command (e.g. "seen endif").
3533 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3534 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3537 PH/17 Added extra conditions to the default value of delay_warning_condition
3541 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3542 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3543 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3546 The Auto-Submitted: and various List- headers are standardised, whereas I
3547 don't think Precedence: ever was.
3549 PH/18 Refactored debugging code in route_finduser() to show more information,
3550 in particular, the error code if getpwnam() issues one.
3552 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3553 This is apparently needed in addition to the PH/07 change above to avoid
3554 any possible encoding problems.
3556 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3557 but not after initializing Perl.
3559 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3560 output them only if debugging. By default they are written stderr,
3561 apparently, which is not desirable.
3563 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3566 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3569 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3571 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3572 authenticated or an ident call has been made. Suppress the default
3573 values for $authenticated_id and $authenticated_sender (but permit -oMai
3574 and -oMas) when testing with -bh.
3576 PH/24 Re-jigged the order of the tests in the default configuration so that the
3577 tests for valid domains and recipients precede the DNS black list and CSA
3578 tests, on the grounds that those ones are more expensive.
3580 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3581 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3582 This bug exists in every version of Exim that I still have, right back to
3585 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3586 However, an attempt to turn on case-sensitivity in a regex key by
3587 including (?-i) didn't work because the subject string was already
3588 lowercased, and the effects were non-intuitive. It turns out that a
3589 one-line patch can be used to allow (?-i) to work as expected.
3595 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3596 other effects) broke the use of negated acl sub-conditions.
3598 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3601 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3602 "Deny" causes Exim to reject the incoming connection with a 554 error.
3603 Unfortunately, if there is a major crisis, such as a disk failure,
3604 tcp-wrappers gives "deny", whereas what one would like would be some
3605 kind of temporary error. A kludge has been added to help with this.
3606 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3607 554 error is used if errno is still zero or contains ENOENT (which occurs
3608 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3611 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3613 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3614 errors. Otherwise a message that provokes a temporary error (when other
3615 messages do not) can cause a whole host to time out.
3617 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3618 addresses were routed directly to files or pipes from a redirect router.
3619 File deliveries just didn't batch; pipe deliveries might have suffered
3622 PH/06 A failure to get a lock for a hints database would erroneously always say
3623 "Failed to get write lock", even when it was really a read lock.
3625 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3626 of 0600. This has been changed to use the value of the lockfile_mode
3627 option (which defaults to 0600).
3629 PH/08 Applied small patch from the Sieve maintainer.
3631 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3632 folder from quota calculations, a direct delivery into this folder messed
3633 up the contents of the maildirsize file. This was because the regex was
3634 used only to exclude .Trash (or whatever) when the size of the mailbox
3635 was calculated. There was no check that a delivery was happening into an
3636 excluded directory. This bug has been fixed by ignoring all quota
3637 processing for deliveries into excluded directories.
3639 PH/10 Added the maildirfolder_create_regex option to appendfile.
3645 PH/01 The code for finding all the local interface addresses on a FreeBSD
3646 system running IPv6 was broken. This may well have applied to all BSD
3647 systems, as well as to others that have similar system calls. The broken
3648 code found IPv4 interfaces correctly, but gave incorrect values for the
3649 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3650 that it would not match correctly against @[] and not recognize the IPv6
3653 PH/02 The ipliteral router was not recognizing addresses of the form user@
3654 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3656 PH/03 Added disable_ipv6.
3658 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3659 decoded form, because it is most often used to construct To: headers
3660 lines in autoreplies, and the decoded form may well be syntactically
3661 invalid. However, $reply_address has leading white space removed, and all
3662 newlines turned into spaces so that the autoreply transport does not
3665 PH/05 If group was specified without a user on a router, and no group or user
3666 was specified on a transport, the group from the router was ignored.
3668 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3669 for visible compile-time settings that can be used to change these
3670 numbers, for those that want even more. Backwards compatibility with old
3671 spool files has been maintained. However, going back to a previous Exim
3672 release will lost any variables that are in spool files.
3674 PH/07 Two small changes when running in the test harness: increase delay when
3675 passing a TCP/IP connection to a new process, in case the original
3676 process has to generate a bounce, and remove special handling of
3677 127.0.0.2 (sic), which is no longer necessary.
3679 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3680 be the same on different OS.
3682 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3685 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3686 whether --show-vars was specified or not
3688 JJ/02 exipick: Added support for new ACL variable spool format introduced
3691 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3692 syntactically invalid From: or Reply-to: line, and a filter used this to
3693 generate an autoreply, and therefore failed to obtain an address for the
3694 autoreply, Exim could try to deliver to a non-existent relative file
3695 name, causing unrelated and misleading errors. What now happens is that
3696 it logs this as a hard delivery error, but does not attempt to create a
3699 PH/11 The exinext utility has a -C option for testing purposes, but although
3700 the given file was scanned by exinext itself; it wasn't being passed on
3701 when Exim was called.
3703 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3704 an end-of-file indication when reading a command response.
3706 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3707 compiled. In many other places in Exim, IPv6 addresses are always
3708 recognized, so I have changed this. It also means that IPv4 domain
3709 literals of the form [IPV4:n.n.n.n] are now always recognized.
3711 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3712 used if the router is not running as root, for example, when verifying at
3713 ACL time, or when using -bh. The debugging output from this situation was
3714 non-existent - all you got was a failure to exec. I have made two
3717 (a) Failures to set uid/gid, the current directory, or a process leader
3718 in a subprocess such as that created by queryprogram now generate
3719 suitable debugging output when -d is set.
3721 (b) The queryprogram router detects when it is not running as root,
3722 outputs suitable debugging information if -d is set, and then runs
3723 the subprocess without attempting to change uid/gid.
3725 PH/15 Minor change to Makefile for building test_host (undocumented testing
3728 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3729 additional section of a DNS packet that returns MX or SRV records.
3730 Instead, it always explicitly searches for A/AAAA records. This avoids
3731 major problems that occur when a DNS server includes only records of one
3732 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3733 fixed another bug: if SRV records were looked up and the corresponding
3734 address records were *not* found in the additional section, the port
3735 values from the SRV records were lost.
3737 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3738 using the correct key (the original address) when searching the retry
3739 rules in order to find which one to use for generating the retry hint.
3741 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3742 adding the default one. Similarly, if it contains a Reply-To: header, the
3743 errors_reply_to option, if set, is not used.
3745 PH/19 When calculating a retry time, Exim used to measure the "time since
3746 failure" by looking at the "first failed" field in the retry record. Now
3747 it does not use this if it is later than than the arrival time of the
3748 message. Instead it uses the arrival time. This makes for better
3749 behaviour in cases where some deliveries succeed, thus re-setting the
3750 "first failed" field. An example is a quota failure for a huge message
3751 when small messages continue to be delivered. Without this change, the
3752 "time since failure" will always be short, possible causing more frequent
3753 delivery attempts for the huge message than are intended.
3754 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3756 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3757 $1, $2, $3) because the numerical variables can be reset during some
3758 expansion items (e.g. "match"), thereby losing the authentication data.
3760 PH/21 Make -bV show the size of off_t variables so that the test suite can
3761 decide whether to run tests for quotas > 2G.
3763 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3764 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3765 filecount value is greater than 2G or if a quota value is greater than 2G
3766 on a system where the size of off_t is not greater than 4, a panic error
3769 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3770 never match. The debug and -bh output now contains an explicit error
3771 message indicating a malformed IPv4 address or mask.
3773 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3774 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3775 PH/23 above applies.
3777 PH/25 Do not write to syslog when running in the test harness. The only
3778 occasion when this arises is a failure to open the main or panic logs
3779 (for which there is an explicit test).
3781 PH/26 Added the /no_tell option to "control=freeze".
3783 PH/27 If a host name lookup failed very early in a connection, for example, if
3784 the IP address matched host_lookup and the reverse lookup yielded a name
3785 that did not have a forward lookup, an error message of the form "no IP
3786 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3787 could be logged. Now it outputs the IP address instead of "NULL".
3789 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3790 allows the sender and the authenticated sender to be set when
3791 submitting a message from within Exim. Since child_open_exim() is
3792 documented for local_scan(), the new function should be too.
3794 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3795 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3796 results in an empty string is now treated as unset.
3798 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3800 PH/31 Added sender_verify_fail logging option.
3802 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3803 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3804 tidied the source and removed it altogether.
3806 PH/33 When a queue run was abandoned because the load average was too high, a
3807 log line was always written; now it is written only if the queue_run log
3808 selector is set. In addition, the log line for abandonment now contains
3809 information about the queue run such as the pid. This is always present
3810 in "start" and "stop" lines but was omitted from the "abandon" line.
3812 PH/34 Omit spaces between a header name and the colon in the error message that
3813 is given when verify = headers_syntax fails (if there are lots of them,
3814 the message gets confusing).
3816 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3817 names, as there are now some PTR records that contain slashes. This check
3818 is only to protect against broken name servers that fall over on strange
3819 characters, so the fact that it applies to all lookups doesn't matter.
3821 PH/36 Now that the new test suite is complete, we can remove some of the
3822 special code in Exim that was needed for the old test suite. For example,
3823 sorting DNS records because real resolvers return them in an arbitrary
3824 order. The new test suite's fake resolver always returns records in the
3827 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3828 bounces) except when queue_only is set, to avoid logging races between
3829 the different processes.
3831 PH/38 Panic-die if .include specifies a non-absolute path.
3833 PH/39 A tweak to the "H" retry rule from its user.
3835 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3836 a label. They prevented compilation on older perls.
3838 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3839 a warning to be raised on newish perls.
3841 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3842 on queue. Changes to match documented behaviour of showing count of
3843 messages matching specified criteria.
3845 PH/40 Changed the default ident timeout from 30s to 5s.
3847 PH/41 Added support for the use of login_cap features, on those BSD systems
3848 that have them, for controlling the resources used by pipe deliveries.
3850 PH/42 The content-scanning code uses fopen() to create files in which to put
3851 message data. Previously it was not paying any attention to the mode of
3852 the files. Exim runs with umask(0) because the rest of the code creates
3853 files with open(), and sets the required mode explicitly. Thus, these
3854 files were ending up world-writeable. This was not a big issue, because,
3855 being within the spool directory, they were not world-accessible. I have
3856 created a function called modefopen, which takes an additional mode
3857 argument. It sets umask(777), creates the file, chmods it to the required
3858 mode, then resets the umask. All the relevant calls to fopen() in the
3859 content scanning code have been changed to use this function.
3861 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3862 to 24 hours. This avoids potential overflow problems when processing G
3863 and H retry rules. I suspect nobody ever tinkers with this value.
3865 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3867 PH/45 When the plaintext authenticator is running as a client, the server's
3868 challenges are checked to ensure they are valid base64 strings. By
3869 default, the authentication attempt is cancelled if an invalid string is
3870 received. Setting client_ignore_invalid_base64 true ignores these errors.
3871 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3872 they are received. Thus, the responses can be made to depend on the
3873 challenges. If an invalid string is ignored, an empty string is placed in
3876 PH/46 Messages that are created by the autoreply transport now contains a
3877 References: header, in accordance with RFCs 2822 and 3834.
3879 PH/47 Added authenticated_sender_force to the smtp transport.
3881 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3883 PH/49 Installed latest patch from the Sieve maintainer.
3885 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3886 was also set, the appendfile transport was unnecessarily scanning a
3887 directory of message files (e.g. for maildir delivery) to find the count
3888 of files (along with the size), even though it did not need this
3889 information. It now does the scan only if it needs to find either the
3890 size of the count of files.
3892 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3894 PH/52 Two bugs concerned with error handling when the smtp transport is
3897 (i) Exim was not creating retry information for temporary errors given
3898 for individual recipients after the DATA command when the smtp transport
3899 was used in LMTP mode. This meant that they could be retried too
3900 frequently, and not timed out correctly.
3902 (ii) Exim was setting the flag that allows error details to be returned
3903 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3904 recipients that were returned after the DATA command.
3906 PH/53 This is related to PH/52, but is more general: for any failing address,
3907 when detailed error information was permitted to be returned to the
3908 sender, but the error was temporary, then after the final timeout, only
3909 "retry timeout exceeded" was returned. Now it returns the full error as
3910 well as "retry timeout exceeded".
3912 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3913 do this, and (what is worse) MTAs that accept it.
3915 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3916 will now be deprecated.
3918 PH/56 New os.c-cygwin from the Cygwin maintainer.
3920 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3921 formats (previously only available in exim formats via -bpr, -bpru,
3922 and -bpra. Now also available in native and exiqgrep formats)
3924 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3925 with very large, slow to parse queues
3927 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3929 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3931 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3932 responses to authentication challenges, though it was showing the
3933 challenges; (ii) I've removed the CR characters from the debug output for
3936 PH/58 Allow for the insertion of a newline as well as a space when a string
3937 is turned into more than one encoded-word during RFC 2047 encoding. The
3938 Sieve code now uses this.
3940 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3941 data_4xx, lost_connection, tls_required.
3943 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3944 message was being sent as an SMTP response.
3946 PH/61 Add -l and -k options to exicyclog.
3948 PH/62 When verifying, if an address was redirected to one new address, so that
3949 verification continued, and the new address failed or deferred after
3950 having set something in $address_data, the value of $address_data was not
3951 passed back to the ACL. This was different to the case when no
3952 redirection occurred. The value is now passed back in both cases.
3954 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3955 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3956 use login_cap.h, so on its own it isn't the distinguishing feature. The
3957 new name refers directly to the setclassresources() function.
3959 PH/65 Added configuration files for NetBSD3.
3961 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3963 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3964 is preferred over IPv4.
3966 PH/68 The bounce_return_message and bounce_return_body options were not being
3967 honoured for bounces generated during the reception of non-SMTP messages.
3968 In particular, this applied to messages rejected by the ACL. This bug has
3969 been fixed. However, if bounce_return_message is true and bounce_return_
3970 body is false, the headers that are returned for a non-SMTP message
3971 include only those that have been read before the error was detected.
3972 (In the case of an ACL rejection, they have all been read.)
3974 PH/69 The HTML version of the specification is now built in a directory called
3975 spec_html instead of spec.html, because the latter looks like a path with
3976 a MIME-type, and this confuses some software.
3978 PH/70 Catch two compiler warnings in sieve.c.
3980 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3981 function verify_get_ident() calls ip_connect() to connect a socket, but
3982 if the "connect()" function timed out, ip_connect() used to close the
3983 socket. However, verify_get_ident() also closes the socket later, and in
3984 between Exim writes to the log, which may get opened at this point. When
3985 the socket was closed in ip_connect(), the log could get the same file
3986 descriptor number as the socket. This naturally causes chaos. The fix is
3987 not to close the socket in ip_connect(); the socket should be closed by
3988 the function that creates it. There was only one place in the code where
3989 this was missing, in the iplookup router, which I don't think anybody now
3990 uses, but I've fixed it anyway.
3992 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3993 well as to direct DNS lookups. Otherwise the handling of names in host
3994 lists is inconsistent and therefore confusing.
4000 PH/01 Two changes to the default runtime configuration:
4002 (1) Move the checks for relay_from_hosts and authenticated clients from
4003 after to before the (commented out) DNS black list checks.
4005 (2) Add control=submission to the relay_from_hosts and authenticated
4006 clients checks, on the grounds that messages accepted by these
4007 statements are most likely to be submissions.
4009 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4011 (1) Generate an error if the third argument for the ${prvs expansion is
4014 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4017 (3) Reset the variables that are obtained from the first argument of
4018 ${prvscheck and used in the second argument before leaving the code,
4019 because their memory is reclaimed, so using them afterwards may do
4022 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4023 one (it's much easier than Tom thought :-).
4025 (5) Because of (4), we can now allow for the use of $prvscheck_result
4026 inside the third argument.
4028 PH/03 For some reason, the default setting of PATH when running a command from
4029 a pipe transport was just "/usr/bin". I have changed it to
4032 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4033 anything to be listed in the output from -bV.
4035 PH/05 When a filter generated an autoreply, the entire To: header line was
4036 quoted in the delivery log line, like this:
4038 => >A.N.Other <ano@some.domain> <original@ddress> ...
4040 This has been changed so that it extracts the operative address. There
4041 may be more than one such address. If so, they are comma-separated, like
4044 => >ano@some.domain,ona@other.domain <original@ddress> ...
4046 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4047 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4048 not being looked up in the rDNS to get a host name, Exim was showing the
4049 IP address twice in Received: lines, even though the IP addresses were
4050 identical. For example:
4052 Received: from [1.2.3.4] (helo=[1.2.3.4])
4054 However, if the real host name was known, it was omitting the HELO data
4055 if it matched the actual IP address. This has been tidied up so that it
4056 doesn't show the same IP address twice.
4058 PH/07 When both +timestamp and +memory debugging was on, the value given by
4059 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4060 called by the debug printing, thereby overwriting the timestamp buffer.
4061 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4063 PH/08 When the original message was included in an autoreply transport, it
4064 always said "this is a copy of the message, including all the headers",
4065 even if body_only or headers_only was set. It now gives an appropriate
4068 PH/09 Applied a patch from the Sieve maintainer which:
4070 o fixes some comments
4071 o adds the (disabled) notify extension core
4072 o adds some debug output for the result of if/elsif tests
4073 o points to the current vacation draft in the documentation
4074 and documents the missing references header update
4078 o fixes a bug in processing the envelope test (when testing
4079 multiple envelope elements, the last element determined the
4082 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4083 Electronic Mail") by including:
4085 Auto-submitted: auto-generated
4087 in the messages that it generates (bounce messages and others, such as
4088 warnings). In the case of bounce messages for non-SMTP messages, there was
4089 also a typo: it was using "Auto_submitted" (underscore instead of
4090 hyphen). Since every message generated by Exim is necessarily in response
4091 to another message, thes have all been changed to:
4093 Auto-Submitted: auto-replied
4095 in accordance with these statements in the RFC:
4097 The auto-replied keyword:
4099 - SHOULD be used on messages sent in direct response to another
4100 message by an automatic process,
4102 - MUST NOT be used on manually-generated messages,
4104 - MAY be used on Delivery Status Notifications (DSNs) and Message
4105 Disposition Notifications (MDNs),
4107 - MUST NOT be used on messages generated by automatic or periodic
4108 processes, except for messages which are automatic responses to
4111 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4112 to the default Received: header definition.
4114 PH/12 Added log selector acl_warn_skipped (default on).
4116 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4117 variables because (a) they are in the wrong storage pool and (b) even if
4118 they were copied, it wouldn't work properly because of the caching.
4120 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4121 checking when decoding. Apparently there are clients that generate
4122 overlong encoded strings. Why am I not surprised?
4124 PH/15 If the first argument of "${if match_address" was not empty, but did not
4125 contain an "@" character, Exim crashed. Now it writes a panic log message
4126 and treats the condition as false.
4128 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4130 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4131 extension "envelope-auth". The code is finished and in agreement with
4132 other implementations, but there is no documentation so far and in fact,
4133 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4134 not changing the active code.
4136 Print executed "if" and "elsif" statements when debugging is used. This
4137 helps a great deal to understand what a filter does.
4139 Document more things not specified clearly in RFC3028. I had all this
4140 sorted out, when out of a sudden new issues came to my mind. Oops."
4142 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4145 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4146 canonical form (as documented). However, after a host name lookup from
4147 the IP address, check_host() was doing a simple string comparison with
4148 addresses acquired from the DNS when checking that the found name did
4149 have the original IP as one of its addresses. Since any found IPv6
4150 addresses are likely to be in abbreviated form, the comparison could
4151 fail. Luckily, there already exists a function for doing the comparison
4152 by converting both addresses to binary, so now that is used instead of
4153 the text comparison.
4155 PH/20 There was another similar case to PH/19, when a complete host name was
4156 given in a host list; looking up its IP address could give an abbreviated
4157 form, whereas the current host's name might or might not be abbreviated.
4158 The same fix has been applied.
4164 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4165 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4168 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4169 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4171 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4173 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4174 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4175 Also, for "Auto-Submitted", treat anything other than "no" as
4176 non-personal, in accordance with RFC 3834. (Previously it treated
4177 anything starting "auto-" as non-personal.)
4179 TF/01 The control=submission/name=... option had a problem with syntax
4180 errors if the name included a slash character. The /name= option
4181 now slurps the rest of the string, so it can include any characters
4182 but it must come last in the list of options (after /sender_retain
4185 PH/05 Some modifications to the interface to the fake nameserver for the new
4193 TK/01 Added the "success_on_redirect" address verification option. See
4194 NewStuff for rationale and an example.
4196 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4198 PH/02 Patch to exigrep to allow it to work on syslog lines.
4200 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4201 fread() to skip over the body file's header line, because in Cygwin the
4202 header line is locked and is inaccessible.
4204 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4205 co-exist for some time) to make it clear that it is the Exim ID that is
4206 referenced, not the Message-ID: header line.
4208 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4209 string_format() function, because snprintf() does not exist on all
4212 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4213 ${stat: expansion item.
4215 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4216 protocol synchronization error", to keep the pedants happy.
4218 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4219 well as for IRIX systems, when gcc is being used. See the host.c source
4222 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4224 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4227 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4228 transport and to the smtp transport in LMTP mode.
4230 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4232 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4233 run a filter in a subprocess. This could lead to confusion in subsequent
4234 lookups in the parent process. There should also be a search_tidyup() at
4235 the end of the subprocess.
4237 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4238 only if the host matched helo_try_verify_hosts, which caused the
4239 verification to occur when the EHLO/HELO command was issued. The ACL just
4240 tested the remembered result. Now, if a previous verification attempt has
4241 not happened, "verify = helo" does it there and then.
4243 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4245 TK/03 Fix log output including CR from clamd.
4247 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4248 exist provoked a memory error which could cause a segfault.
4250 PH/15 Installed PCRE 6.2
4252 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4254 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4255 of the problem. Specifically, suggested +O2 rather than +O1 for the
4258 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4260 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4261 record, the port was not correctly logged when the outgoing_port log
4262 selector was set (it logged the transort's default port).
4264 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4265 fallback_hosts, and "hosts" in the smtp transport.
4267 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4268 host errors such as "Connection refused".
4270 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4271 authentication with radiusclient 0.4.9:
4273 - Error returned from rc_read_config was caught wrongly
4274 - Username/password not passed on to radius server due to wrong length.
4276 The presumption is that some radiusclient API changes for 4.51/PH/17
4277 were not taken care of correctly. The code is still untested by me (my
4278 Linux distribution still has 0.3.2 of radiusclient), but it was
4279 contributed by a Radius user.
4281 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4282 expanding the "port" option of the smtp transport.
4284 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4285 while reading a MIME header. Thanks to Tom Hughes for a patch.
4287 PH/24 Include config.h inside local_scan.h so that configuration settings are
4290 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4291 that in an ACL for RCPT (for example), you can examine exactly what was
4294 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4295 commands, but it was not correctly comparing the address with the actual
4296 client host address. Thus, it would show the EHLO address in Received:
4297 header lines when this was not necessary.
4299 PH/27 Added the % operator to ${eval:}.
4301 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4302 it should be ignoring failures (because with -C, for example, it has lost
4303 privilege). It wasn't ignoring creation failures other than "already
4306 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4309 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4310 because an input line was too long, either on its own, or by virtue of
4311 too many continuations, the temporary file was not being removed, and the
4312 return code was incorrect.
4314 PH/31 Missing "BOOL" in function definition in filtertest.c.
4316 PH/32 Applied Sieve patches from the maintainer.
4318 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4320 PH/33 Added "verify = not_blind".
4322 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4323 Local/Makefile (with some defaults set). These are used in built scripts
4324 such as exicyclog, but they have never been used in the exim_install
4325 script (though there are many overriding facilities there). I have
4326 arranged that the exim_install script now takes note of these two
4329 PH/35 Installed configuration files for Dragonfly.
4331 PH/36 When a locally submitted message by a trusted user did not contain a
4332 From: header, and the sender address was obtained from -f or from an SMTP
4333 MAIL command, and the trusted user did not use -F to supply a sender
4334 name, $originator_name was incorrectly used when constructing a From:
4335 header. Furthermore, $originator_name was used for submission mode
4336 messages from external hosts without From: headers in a similar way,
4337 which is clearly wrong.
4339 PH/37 Added control=suppress_local_fixups.
4341 PH/38 When log_selector = +received_sender was set, and the addition of the
4342 sender made the log line's construction buffer exactly full, or one byte
4343 less than full, an overflow happened when the terminating "\n" was
4346 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4347 when the result of a list match is failure because a DNS lookup failed.
4349 PH/40 RM_COMMAND is now used in the building process.
4351 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4352 the "build-* directories that it finds.
4354 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4355 address in a domain literal was a prefix of an interface address.
4357 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4358 when verifying a sender address, unless rewrite_headers is false.
4360 PH/44 Wrote a long comment about why errors_to addresses are verified as
4361 recipients, not senders.
4363 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4364 the ratelimit ACL was added.
4366 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4368 PH/46 Added extra information about PostgreSQL errors to the error string.
4370 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4371 avoiding the need to install special zones in a real server. This is
4372 backwards compatible; if it can't find the fake resolver, it drops back.
4373 Thus, both old and new test suites can be run.
4375 TF/02 Added util/ratelimit.pl
4377 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4378 clock is set back in time.
4380 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4381 Candler <B.Candler@pobox.com>.
4383 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4384 OK for addresses that are the result of redirections.
4386 PH/48 A number of further additions for the benefit of the new test suite,
4387 including a fake gethostbyname() that interfaces to the fake DNS resolver
4390 TF/06 The fix for widen_domains has also been applied to qualify_single and
4391 search_parents which are the other dnslookup options that can cause
4394 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4397 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4399 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4400 service" if it encountered an MX record with an empty target hostname.
4401 The message is now "an MX or SRV record indicated no SMTP service".
4403 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4404 if the DNS of the sending site is misconfigured. This is quite a
4405 common situation. This change restores the behaviour of treating a
4406 helo verification defer as a failure.
4408 PH/51 If self=fail was set on a router, the bounce message did not include the
4409 actual error message.
4415 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4417 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4418 command itself ended in error, the underlying message about the transport
4419 filter timeout was being overwritten with the pipe command error. Now the
4420 underlying error message should be appended to the second error message.
4422 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4424 PH/02 Reduce the amount of output that "make" produces by default. Full output
4425 can still be requested.
4427 PH/03 The warning log line about a condition test deferring for a "warn" verb
4428 was being output only once per connection, rather than after each
4429 occurrence (because it was using the same function as for successful
4430 "warn" verbs). This seems wrong, so I have changed it.
4432 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4433 it should not have, which might have caused a crash in the right
4434 circumstances, but probably never did.
4436 PH/04 Installed a modified version of Tony Finch's patch to make submission
4437 mode fix the return path as well as the Sender: header line, and to
4438 add a /name= option so that you can make the user's friendly name appear
4441 TF/03 Added the control = fakedefer ACL modifier.
4443 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4444 Mark Lowes for thorough testing.
4446 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4448 TK/03 Merged latest SRS patch from Miles Wilton.
4450 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4451 with the definition in sysexits.h (which is #included earlier).
4452 Fortunately, Exim does not actually use EX_OK. The code used to try to
4453 preserve the sysexits.h value, by assuming that macro definitions were
4454 scanned for macro replacements. I have been disabused of this notion,
4455 so now the code just undefines EX_OK before #including unistd.h.
4457 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4458 in the smtp transport. When a block could not be written in a single
4459 write() function, the timeout was being re-applied to each part-write.
4460 This seems wrong - if the receiver was accepting one byte at a time it
4461 would take for ever. The timeout is now adjusted when this happens. It
4462 doesn't have to be particularly precise.
4464 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4465 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4467 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4468 without a domain if the check to <postmaster@domain> fails.
4470 SC/01 Eximstats: added -xls and the ability to specify output files
4471 (patch written by Frank Heydlauf).
4473 SC/02 Eximstats: use FileHandles for outputting results.
4475 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4477 SC/04 Eximstats: fixed display of large numbers with -nvr option
4479 SC/05 Eximstats: fixed merging of reports with empty tables.
4481 SC/06 Eximstats: added the -include_original_destination flag
4483 SC/07 Eximstats: removed tabs and trailing whitespace.
4485 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4487 TK/06 MBOX spool code: Add real "From " MBOX separator line
4488 so the .eml file is really in mbox format (even though
4489 most programs do not really care). Patch from Alex Miller.
4491 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4492 The latter is generated from $received_to and is only set if the
4493 message has one envelope recipient. SA can use these headers,
4494 obviously out-of-the-box. Patch from Alex Miller.
4496 PH/08 The ${def test on a variable was returning false if the variable's
4497 value was "0", contrary to what the specification has always said!
4498 The result should be true unless the variable is empty.
4500 PH/09 The syntax error of a character other than { following "${if
4501 def:variable_name" (after optional whitespace) was not being diagnosed.
4502 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4503 accidental colon was present, for example, could give incorrect results.
4505 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4506 result is used (not including appendfile, where other changes are about
4509 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4510 This involved changing a lot of size variables from int to off_t. It
4511 should work with maildirs and everything.
4513 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4514 spamd dying while we are connected to it.
4516 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4519 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4520 function for BDB 4.3.
4522 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4524 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4525 the sources for PCRE have been re-arranged and more files are now
4528 PH/15 The code I had for printing potentially long long variables in PH/11
4529 above was not the best (it lost precision). The length of off_t variables
4530 is now inspected at build time, and an appropriate printing format (%ld
4531 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4532 to be "long long int" or "long int". This is needed for the internal
4533 formatting function string_vformat().
4535 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4536 the configuration file to be ":syslog", then the script "guesses" where
4537 the logs files are, rather than using the compiled in default. In our
4538 case the guess is not the same as the compiled default, so the script
4539 suddenly stopped working when I started to use syslog. The patch checks
4540 to see if log_file_path is "". If so, it attempts to read it from exim
4541 with no configuration file to get the compiled in version, before it
4542 falls back to the previous guessing code."
4544 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4545 implementing BATV in an Exim configuration. See NewStuff for the gory
4548 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4549 Makefile that are specific to HP-UX.
4551 PH/18 If the "use_postmaster" option was set for a recipient callout together
4552 with the "random" option, the postmaster address was used as the MAIL
4553 FROM address for the random test, but not for the subsequent recipient
4554 test. It is now used for both.
4556 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4557 patch removes a few documentation additions to RFC 3028, because the
4558 latest draft now contains them. It adds the new en;ascii-case comparator
4559 and a new error check for 8bit text in MIME parts. Comparator and
4560 require names are now matched exactly. I enabled the subaddress
4561 extension, but it is not well tested yet (read: it works for me)."
4563 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4564 rework some of the code of TK/09 above to avoid the hardwired use of
4565 "%lld" and "long long". Replaced the call to snprintf() with a call to
4568 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4569 records point to non-existent hosts", "retry timeout exceeded", and
4570 "retry time not reached for any host after a long failure period".
4572 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4573 experimental DomainKeys support:
4575 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4576 (2) On an error such as an illegally used "control", the wrong name for
4577 the control was given.
4579 These problems did NOT occur unless DomainKeys support was compiled.
4581 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4583 PH/24 Added ${if match_ip condition.
4585 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4586 databases so that it will be absolutely obvious if a crash occurs in the
4587 DB library. This is a regular occurrence (often caused by mis-matched
4590 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4591 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4592 picked up on a user's system that detects such things. There doesn't seem
4593 to be a gcc warning option for this - only an attribute that has to be
4594 put on the function's prototype. It seems that in Fedora Core 4 they have
4595 set this on a number of new functions. No doubt there will be more in due
4598 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4599 specify a transport. However, if an address that was verified by such a
4600 router was the subject of a callout, Exim crashed because it tried to
4601 read the rcpt_include_affixes from the non-existent transport. Now it
4602 just assumes that the setting of that option is false. This bug was
4603 introduced by 4.51/PH/31.
4605 PH/28 Changed -d+all to exclude +memory, because that information is very
4606 rarely of interest, but it makes the output a lot bigger. People tend to
4607 do -d+all out of habit.
4609 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4610 code in os-type was giving problems when libc.so lives in lib64, like on
4613 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4614 aren't the modern standard, and it seems that some systems' include files
4615 don't always have them. Exim was already checking for some of the newer
4616 ones like T_AAAA, and defining it itself. I've added checks for all the
4617 record types that Exim uses.
4619 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4620 not automatically generating a new one, as it is supposed to. This
4621 prevented TLS from working. If the file did exist, but contained invalid
4622 data, a new version was generated, as expected. It was only the case of a
4623 non-existent file that was broken.
4625 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4626 with a change in libdomainkeys > 0.64.
4628 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4629 from DNS. If the selector record carries the flag, it now has
4630 precedence over the domain-wide flag.
4632 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4634 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4635 the use of a port name that isn't defined in /etc/services) occurred, the
4636 message was deferred as in a normal delivery, and thus remained on the
4637 spool, instead of being failed because of the mua_wrapper setting. This
4638 is now fixed, and I tidied up some of the mua_wrapper messages at the
4641 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4642 the messages in a hash of arrays rather than using individual hashes.
4643 This is a bit cleaner and results in dramatic memory savings, albeit
4644 at a slight CPU cost.
4646 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4647 as requested by Marc Sherman.
4649 SC/10 Eximstats: added histograms for user specified patterns as requested
4652 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4654 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4655 fopen() in the content-scanning modules that did not already have it.
4661 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4662 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4664 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4666 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4668 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4669 to test Sieve filters that use "vacation".
4671 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4672 that changes the way the GnuTLS parameters are stored in the cache file.
4673 The new format can be generated externally. For backward compatibility,
4674 if the data in the cache doesn't make sense, Exim assumes it has read an
4675 old-format file, and it generates new data and writes a new file. This
4676 means that you can't go back to an older release without removing the
4679 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4680 work if there are any delivery delays because "one_time" forces the
4681 parent to be marked "delivered", so its unseen clone is never tried
4682 again. For this reason, Exim now forbids the simultaneous setting of
4685 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4686 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4687 entry if you want to know the details. The fix, however, neglected to
4688 consider the case when local delivery batching is involved. The test for
4689 "previously delivered" was not happening when checking to see if an
4690 address could be batched with a previous (undelivered) one; under
4691 certain circumstances this could lead to multiple deliveries to the same
4694 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4695 in its include files, and this causes problems building Exim.
4697 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4698 header_syntax) but Exim was just ignoring anything given after a slash.
4699 In particular, this caused confusion with an attempt to use "verify =
4700 reverse_host_lookup/defer_ok". An error is now given when options are
4701 supplied for verify items that do not have them. (Maybe reverse_host_
4702 lookup should have a defer_ok option, but that's a different point.)
4704 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4705 defined by RFC 821) to 2048, because there were problems with some AUTH
4706 commands, and RFC 1869 says the size should be increased for extended
4707 SMTP commands that take arguments.
4709 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4712 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4713 "unknown" error; now it says that the functionality isn't in the binary.
4715 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4716 an address' error message when a string expansion fails (syntax or
4717 whatever). Otherwise the password may appear in the log. Following change
4718 PH/42 below, there is no longer a chance of it appearing in a bounce
4721 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4723 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4724 of its addresses was ever tried. (Bugzilla bug #2.)
4726 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4727 the result incorrectly in the debug output. (It correctly added a newline
4728 to what was transported.)
4730 TF/01 Added $received_time.
4732 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4733 commented out examples of how to interface to a virus scanner and to
4734 SpamAssassin. Also added commented examples of av_scanner and
4735 spamd_address settings.
4737 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4738 and controls are allowed in which ACLs. There were a couple of minor
4739 errors. Some of the entries in the conditions table (which is a table of
4740 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4741 negation of where the condition IS allowed.
4743 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4745 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4746 header file does not have a version number, so I've had to invent a new
4747 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4748 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4749 radiusclient), but it was contributed by a Radius user.
4751 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4752 files or directories, for OpenSSL.
4754 PH/19 When an Exim process that is running as root has to create an Exim log
4755 file, it does so in a subprocess that runs as exim:exim so as to get the
4756 ownership right at creation (otherwise, other Exim processes might see
4757 the file with the wrong ownership). There was no test for failure of this
4758 fork() call, which would lead to the process getting stuck as it waited
4759 for a non-existent subprocess. Forks do occasionally fail when resources
4760 run out. I reviewed all the other calls to fork(); they all seem to check
4763 PH/20 When checking for unexpected SMTP input at connect time (before writing
4764 the banner), Exim was not dealing correctly with a non-positive return
4765 from the read() function. If the client had disconnected by this time,
4766 the result was a log entry for a synchronization error with an empty
4767 string after "input=" when read() returned zero. If read() returned -1
4768 (an event I could not check), uninitialized data bytes were printed.
4769 There were reports of junk text (parts of files, etc) appearing after
4772 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4774 PH/22 Added support for macro redefinition, and (re)definition in between
4775 driver and ACL definitions.
4777 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4778 forgetting to use the resulting value; it was using the unexpanded value.
4780 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4781 hadn't been configured. The fix is from Juergen Kreileder, who
4782 understands it better than I do:
4784 "Here's what I see happening with three configured cyrus_sasl
4785 authenticators configured (plain, login, cram-md5):
4787 On startup auth_cyrus_sasl_init() gets called for each of these.
4788 This means three calls to sasl_listmech() without a specified mech_list.
4789 => SASL tests which mechs of all available mechs actually work
4790 => three warnings about OTP not working
4791 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4793 With the patch, sasl_listmech() also gets called three times. But now
4794 SASL's mech_list option is set to the server_mech specified in the the
4795 authenticator. Or in other words, the answer from sasl_listmech()
4796 gets limited to just the mech you're testing for (which is different
4798 => the return list contains just 'plain' or 'login', 'cram-md5' or
4799 nothing depending on the value of ob->server_mech.
4801 I've just tested the patch: Authentication still works fine,
4802 unavailable mechs specified in the exim configuration are still
4803 caught, and the auth.log warnings about OTP are gone."
4805 PH/25 When debugging is enabled, the contents of the command line are added
4806 to the debugging output, even when log_selector=+arguments is not
4809 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4810 answer is "GNU", and only if the return is "GNU/something" is the answer
4813 PH/27 $acl_verify_message is now set immediately after the failure of a
4814 verification in an ACL, and so is available in subsequent modifiers. In
4815 particular, the message can be preserved by coding like this:
4817 warn !verify = sender
4818 set acl_m0 = $acl_verify_message
4820 Previously, $acl_verify_message was set only while expanding "message"
4821 and "log_message" when a very denied access.
4823 PH/28 Modified OS/os.c-Linux with
4825 -#ifndef OS_LOAD_AVERAGE
4826 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4828 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4829 nomenclature these days.)
4831 PH/29 Installed patch from the Sieve maintainer that adds the options
4832 sieve_useraddress and sieve_subaddress to the redirect router.
4834 PH/30 In these circumstances:
4835 . Two addresses routed to the same list of hosts;
4836 . First host does not offer TLS;
4837 . First host accepts first address;
4838 . First host gives temporary error to second address;
4839 . Second host offers TLS and a TLS session is established;
4840 . Second host accepts second address.
4841 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4842 and peerdn, if requested) that were in fact used only for the second
4845 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4846 attention to any local part prefix or suffix that was matched by the
4847 router that accepted the address. It now behaves in the same way as it
4848 does for delivery: the affixes are removed from the local part unless
4849 rcpt_include_affixes is set on the transport.
4851 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4852 timeout during the DATA phase of an incoming message.
4854 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4855 applied a patch sanctioned by the Sieve maintainer.
4857 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4858 the uid or gid is negative. A case of a negative gid caused this to be
4859 noticed. The fix allows for either to be negative.
4861 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4862 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4865 PH/36 The change PH/12 above was broken. Fixed it.
4867 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4868 the grounds that routing the same address twice would always produce the
4869 same answer. This might have been true once, but it is certainly no
4870 longer true now. Routing a child address may depend on the previous
4871 routing that produced that child. Some complicated redirection strategies
4872 went wrong when messages had multiple recipients, and made Exim's
4873 behaviour dependent on the order in which the addresses were given.
4875 I have moved the duplicate checking until after the routing is complete.
4876 Exim scans the addresses that are assigned to local and remote
4877 transports, and removes any duplicates. This means that more work will be
4878 done, as duplicates will always all be routed, but duplicates are
4879 presumably rare, so I don't expect this is of any significance.
4881 For deliveries to pipes, files, and autoreplies, the duplicate checking
4882 still happens during the routing process, since they are not going to be
4885 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4886 It corrects a timeout issue with spamd. This is Ian's comment: "The
4887 background is that sometimes spamd either never reads data from a
4888 connection it has accepted, or it never writes response data. The exiscan
4889 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4890 blindly assumes that writes won't block so it may never time out."
4892 PH/39 Allow G after quota size as well as K and M.
4894 PH/40 The value set for $authenticated_id in an authenticator may not contain
4895 binary zeroes or newlines because the value is written to log lines and
4896 to spool files. There was no check on this. Now the value is run through
4897 the string_printing() function so that such characters are converted to
4898 printable escape sequences.
4900 PH/41 $message_linecount is a new variable that contains the total number of
4901 lines in the message. Compare $body_linecount, which is the count for the
4904 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4905 bounce and delay warning messages, except in certain special cases, which
4908 (a) An SMTP error message from a remote host;
4909 (b) A message specified in a :fail: redirection;
4910 (c) A message specified in a "fail" command in a system filter;
4911 (d) A message specified in a FAIL return from the queryprogram router;
4912 (e) A message specified by the cannot_route_message router option.
4914 In these cases only, Exim does include the error details in bounce and
4915 warning messages. There are also a few cases where bland messages such
4916 as "unrouteable address" or "local delivery error" are given.
4918 PH/43 $value is now also set for the "else" part of a ${run expansion.
4920 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4921 being worked on, but at least Exim now implements the latest version to
4924 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4925 process to complete was treated as a delivery failure, a timeout while
4926 writing the message to the pipe was logged, but erroneously treated as a
4927 successful delivery. Such timeouts include transport filter timeouts. For
4928 consistency with the overall process timeout, these timeouts are now
4929 treated as errors, giving rise to delivery failures by default. However,
4930 there is now a new Boolean option for the pipe transport called
4931 timeout_defer, which, if set TRUE, converts the failures into defers for
4932 both kinds of timeout. A transport filter timeout is now identified in
4935 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4936 systems where "make" and "gmake" are different, calling "gmake" at top
4937 level broke things. I've arranged for the value of $(MAKE) to be passed
4938 from the Makefile to this script so that it can call the same version of
4942 A note about Exim versions 4.44 and 4.50
4943 ----------------------------------------
4945 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4946 changes of various kinds. As a consequence, a big documentation update was
4947 needed. This delayed the release for rather longer than seemed good, especially
4948 in the light of a couple of (minor) security issues. Therefore, the changes
4949 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4950 release. So 4.44 and 4.50 are in effect two different branches that both start
4953 I have left the 4.50 change log unchanged; it contains all the changes since
4954 4.43. The change log for 4.44 is below; many of its items are identical to
4955 those for 4.50. This seems to be the most sensible way to preserve the
4956 historical information.
4962 1. Minor wording change to the doc/README.SIEVE file.
4964 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4965 computation of the current number of files was incorrect.
4967 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4968 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4971 4. Give more explanation in the error message when the command for a transport
4972 filter fails to execute.
4974 5. There are several places where Exim runs a non-Exim command in a
4975 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4976 was being done only for the command run by the queryprogram router. It is
4977 now done for all such subprocesses. The other cases are: ${run, transport
4978 filters, and the commands run by the lmtp and pipe transports.
4980 6. Added CONFIGURE_GROUP build-time option.
4982 7. Some older OS have a limit of 256 on the maximum number of file
4983 descriptors. Exim was using setrlimit() to set 1000 as a large value
4984 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4985 systems. I've change it so that if it can't get 1000, it tries for 256.
4987 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4988 was an oversight, and furthermore, ever since the addition of extra
4989 controls (e.g. 4.43/32), the checks on when to allow different forms of
4990 "control" were broken. There should now be diagnostics for all cases when a
4991 control that does not make sense is encountered.
4993 9. Added the /retain_sender option to "control=submission".
4995 10. $recipients is now available in the predata ACL (oversight).
4997 11. Tidy the search cache before the fork to do a delivery from a message
4998 received from the command line. Otherwise the child will trigger a lookup
4999 failure and thereby defer the delivery if it tries to use (for example) a
5000 cached ldap connection that the parent has called unbind on.
5002 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5003 of $address_data from the recipient verification was clobbered by the
5004 sender verification.
5006 13. The value of address_data from a sender verification is now available in
5007 $sender_address_data in subsequent conditions in the ACL statement.
5009 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5011 15. Added a new option "connect=<time>" to callout options, to set a different
5014 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5015 was its contents. (It was OK if the option was not defined at all.)
5017 17. A "Completed" log line is now written for messages that are removed from
5018 the spool by the -Mrm option.
5020 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5021 information about exactly what failed.
5023 19. Added -dd to debug only the daemon process.
5025 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5026 handles timeouts, both on the server side and network timeouts. Renamed the
5027 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5029 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5030 It is now set to "smtps".
5032 22. $host_address is now set to the target address during the checking of
5033 ignore_target_hosts.
5035 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5036 being passed; this would have caused $sender_host_name to have been used if
5037 matching the list had actually called for a host name (not very likely,
5038 since this list is usually IP addresses). A host name is now passed as
5041 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5042 code that specifies a non-restarting handler (typically sigaction() in
5043 modern systems) in an attempt to fix a rare and obscure crash bug.
5045 25. Narrowed the window for a race in the daemon that could cause it to ignore
5046 SIGCHLD signals. This is not a major problem, because they are used only to
5047 wake it up if nothing else does.
5049 26. A malformed maildirsize file could cause Exim to calculate negative values
5050 for the mailbox size or file count. Odd effects could occur as a result.
5051 The maildirsize information is now recalculated if the size or filecount
5054 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5055 support for a long time. Removed HAVE_SYS_VFS_H.
5057 28. Installed the latest version of exipick from John Jetmore.
5059 29. In an address list, if the pattern was not a regular expression, an empty
5060 subject address (from a bounce message) matched only if the pattern was an
5061 empty string. Non-empty patterns were not even tested. This was the wrong
5062 because it is perfectly reasonable to use an empty address as part of a
5063 database query. An empty address is now tested by patterns that are
5064 lookups. However, all the other forms of pattern expect the subject to
5065 contain a local part and a domain, and therefore, for them, an empty
5066 address still always fails if the pattern is not itself empty.
5068 30. Exim went into a mad DNS loop when attempting to do a callout where the
5069 host was specified on an smtp transport, and looking it up yielded more
5070 than one IP address.
5072 31. Re-factored the code for checking spool and log partition space into a
5073 function that finds that data and another that does the check. The former
5074 is then used to implement four new variables: $spool_space, $log_space,
5075 $spool_inodes, and $log_inodes.
5077 32. The RFC2047 encoding function was originally intended for short strings
5078 such as real names; it was not keeping to the 75-character limit for
5079 encoded words that the RFC imposes. It now respects the limit, and
5080 generates multiple encoded words if necessary. To be on the safe side, I
5081 have increased the buffer size for the ${rfc2047: expansion operator from
5084 33. It is now permitted to omit both strings after an "if" condition; if the
5085 condition is true, the result is "true". As before, when the second string
5086 is omitted, a false condition yields an empty string. This makes it less
5087 cumbersome to write custom ACL and router conditions.
5089 34. Failure to deliver a bounce message always caused it to be frozen, even if
5090 there was an errors_to setting on the router. The errors_to setting is now
5093 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5094 canonical form (fully expanded) before being placed in
5095 $sender_host_address.
5097 36. The table in the code that translates DNS record types into text (T_A to
5098 "A" for instance) was missing entries for NS and CNAME. It is just possible
5099 that this could have caused confusion if both these types were looked up
5100 for the same domain, because the text type is used as part of Exim's
5101 per-process caching. But the chance of anyone hitting this buglet seems
5104 37. The dnsdb lookup has been extended in a number of ways.
5106 (1) There is a new type, "zns", which walks up the domain tree until it
5107 finds some nameserver records. It should be used with care.
5109 (2) There is a new type, "mxh", which is like "mx" except that it returns
5110 just the host names, not the priorities.
5112 (3) It is now possible to give a list of domains (or IP addresses) to be
5113 looked up. The behaviour when one of the lookups defers can be
5114 controlled by a keyword.
5116 (4) It is now possible to specify the separator character for use when
5117 multiple records are returned.
5119 38. The dnslists ACL condition has been extended: it is now possible to supply
5120 a list of IP addresses and/or domains to be looked up in a particular DNS
5123 39. Added log_selector=+queue_time_overall.
5125 40. When running the queue in the test harness, wait just a tad after forking a
5126 delivery process, to get repeatability of debugging output.
5128 41. Include certificate and key file names in error message when GnuTLS fails
5129 to set them up, because the GnuTLS error message doesn't include the name
5130 of the failing file when there is a problem reading it.
5132 42. Allow both -bf and -bF in the same test run.
5134 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5136 44. The "Exiscan patch" is now merged into the mainline Exim source.
5138 45. Sometimes the final signoff response after QUIT could fail to get
5139 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5140 < 0 before doing a fflush(). This bug looks as though it goes back to the
5141 introduction of TLS in release 3.20, but "sometimes" must have been rare
5142 because the tests only now provoked it.
5144 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5145 (this can affect the format of dates).
5147 47. exim_tidydb, when checking for the continued existence of a message for
5148 which it has found a message-specific retry record, was not finding
5149 messages that were in split spool directories. Consequently, it was
5150 deleting retry records that should have stayed in existence.
5152 48. Steve fixed some bugs in eximstats.
5154 49. The SPA authentication driver was not abandoning authentication and moving
5155 on to the next authenticator when an expansion was forced to fail,
5156 contradicting the general specification for all authenticators. Instead it
5157 was generating a temporary error. It now behaves as specified.
5159 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5160 (the order specifies the preference for clients). The order is now AES256,
5161 AES128, 3DES, ARCFOUR128.
5163 51. Small patch to Sieve code - explicitly set From: when generating an
5166 52. Exim crashed if a remote delivery caused a very long error message to be
5167 recorded - for instance if somebody sent an entire SpamAssassin report back
5168 as a large number of 550 error lines. This bug was coincidentally fixed by
5169 increasing the size of one of Exim's internal buffers (big_buffer) that
5170 happened as part of the Exiscan merge. However, to be on the safe side, I
5171 have made the code more robust (and fixed the comments that describe what
5174 53. Now that there can be additional text after "Completed" in log lines (if
5175 the queue_time_overall log selector is set), a one-byte patch to exigrep
5176 was needed to allow it to recognize "Completed" as not the last thing in
5179 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5180 patch that reportedly fixes this has been added. I am not expert enough to
5181 create a test for it. This is what the patch creator wrote:
5183 "I found a little strange behaviour of ldap code when working with
5184 Windows 2003 AD Domain, where users was placed in more than one
5185 Organization Units. When I tried to give exim partial DN, the exit code
5186 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5187 But simultaneously result of request was absolutely normal ldap result,
5188 so I produce this patch..."
5190 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5191 REFERENCE, so I have modified the code to exclude the patch when that macro
5194 55. Some experimental protocols are using DNS PTR records for new purposes. The
5195 keys for these records are domain names, not reversed IP addresses. The
5196 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5197 leaves it alone. Component reversal etc. now happens only for IP addresses.
5200 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5202 57. Double the size of the debug message buffer (to 2048) so that more of very
5203 long debug lines gets shown.
5205 58. The exicyclog utility now does better if the number of log files to keep
5206 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5208 59. Two changes related to the smtp_active_hostname option:
5210 (1) $smtp_active_hostname is now available as a variable.
5211 (2) The default for smtp_banner uses $smtp_active_hostname instead
5212 of $primary_hostname.
5214 60. The host_aton() function is supposed to be passed a string that is known
5215 to be a valid IP address. However, in the case of IPv6 addresses, it was
5216 not checking this. This is a hostage to fortune. Exim now panics and dies
5217 if the condition is not met. A case was found where this could be provoked
5218 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5219 components; fortuitously, this particular loophole had already been fixed
5220 by change 4.50/55 above.
5222 If there are any other similar loopholes, the new check in host_aton()
5223 itself should stop them being exploited. The report I received stated that
5224 data on the command line could provoke the exploit when Exim was running as
5225 exim, but did not say which command line option was involved. All I could
5226 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5227 running as the user.
5230 61. There was a buffer overflow vulnerability in the SPA authentication code
5231 (which came originally from the Samba project). I have added a test to the
5232 spa_base64_to_bits() function which I hope fixes it.
5235 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5236 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5238 63. The daemon start-up calls getloadavg() while still root for those OS that
5239 need the first call to be done as root, but it missed one case: when
5240 deliver_queue_load_max is set with deliver_drop_privilege. This is
5241 necessary for the benefit of the queue runner, because there is no re-exec
5242 when deliver_drop_privilege is set.
5244 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5245 This has been fixed.
5247 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5248 was in use, was not putting the data itself into the right store pool;
5249 consequently, it could be overwritten for a subsequent message in the same
5250 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5253 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5255 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5256 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5257 the function were treating the return as a boolean value, which happened to
5258 work because 0=false and not-0=true, but is not correct code.
5260 68. The host_aton() function was not handling scoped IPv6 addresses (those
5261 with, for example, "%eth0" on the end) correctly.
5263 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5264 negated items (that is, ~something) in unsigned ints. Some compilers
5265 apparently mutter when there is no cast.
5267 70. If an address verification called from an ACL failed, and did not produce a
5268 user-specific message (i.e. there was only a "system" message), nothing was
5269 put in $acl_verify_message. In this situation, it now puts the system
5272 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5273 session; change 4.31/43 added the unwanted input to the log line - except
5274 that it did not do this in the start of session case. It now does.
5276 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5277 This is wrong and can cause the other end to generate a synchronization
5278 error if it is another Exim or anything else that does the synchronization
5279 check. A QUIT command is no longer sent after a timeout.
5281 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5282 during host lookups.
5284 74. The defer_ok option of callout verification was not working if it was used
5285 when verifying addresses in header lines, that is, for this case:
5287 verify = header_sender/callout=defer_ok
5289 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5290 those file descriptors could be used for SMTP connections. If anything
5291 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5292 could be sent to the SMTP client, causing chaos. The daemon now opens
5293 stdin, stdout, and stderr to /dev/null when it puts itself into the
5296 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5297 log by default. The user can override this with suitable Perl magic.
5299 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5300 the log message when discard triggers, was not working for the DATA ACL or
5301 for the non-SMTP ACL.
5303 78. Error message wording change in sieve.c.
5305 79. If smtp_accept_max_per_host was set, the number of connections could be
5306 restricted to fewer than expected, because the daemon was trying to set up
5307 a new connection before checking whether the processes handling previous
5308 connections had finished. The check for completed processes is now done
5309 earlier. On busy systems, this bug wouldn't be noticed because something
5310 else would have woken the daemon, and it would have reaped the completed
5313 80. If a message was submitted locally by a user whose login name contained one
5314 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5315 It caused a spool format error. I have fixed the spool reading code. A
5316 related problem was that the "from" clause in the Received: line became
5317 illegal because of the space(s). It is now covered by ${quote_local_part.
5319 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5322 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5324 83. Patch from Sieve maintainer to fix unterminated string problem in
5325 "vacation" handling.
5327 84. Some minor changes to the Linux configuration files to help with other
5328 OS variants using glibc.
5330 85. One more patch for Sieve to update vacation handling to latest spec.
5333 ----------------------------------------------------
5334 See the note above about the 4.44 and 4.50 releases.
5335 ----------------------------------------------------
5341 1. Change 4.43/35 introduced a bug that caused file counts to be
5342 incorrectly computed when quota_filecount was set in an appendfile
5345 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5346 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5349 3. Give more explanation in the error message when the command for a transport
5350 filter fails to execute.
5352 4. There are several places where Exim runs a non-Exim command in a
5353 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5354 was being done only for the command run by the queryprogram router. It is
5355 now done for all such subprocesses. The other cases are: ${run, transport
5356 filters, and the commands run by the lmtp and pipe transports.
5358 5. Some older OS have a limit of 256 on the maximum number of file
5359 descriptors. Exim was using setrlimit() to set 1000 as a large value
5360 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5361 systems. I've change it so that if it can't get 1000, it tries for 256.
5363 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5364 was an oversight, and furthermore, ever since the addition of extra
5365 controls (e.g. 4.43/32), the checks on when to allow different forms of
5366 "control" were broken. There should now be diagnostics for all cases when a
5367 control that does not make sense is encountered.
5369 7. $recipients is now available in the predata ACL (oversight).
5371 8. Tidy the search cache before the fork to do a delivery from a message
5372 received from the command line. Otherwise the child will trigger a lookup
5373 failure and thereby defer the delivery if it tries to use (for example) a
5374 cached ldap connection that the parent has called unbind on.
5376 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5377 of $address_data from the recipient verification was clobbered by the
5378 sender verification.
5380 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5381 was its contents. (It was OK if the option was not defined at all.)
5383 11. A "Completed" log line is now written for messages that are removed from
5384 the spool by the -Mrm option.
5386 12. $host_address is now set to the target address during the checking of
5387 ignore_target_hosts.
5389 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5390 being passed; this would have caused $sender_host_name to have been used if
5391 matching the list had actually called for a host name (not very likely,
5392 since this list is usually IP addresses). A host name is now passed as
5395 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5396 code that specifies a non-restarting handler (typically sigaction() in
5397 modern systems) in an attempt to fix a rare and obscure crash bug.
5399 15. Narrowed the window for a race in the daemon that could cause it to ignore
5400 SIGCHLD signals. This is not a major problem, because they are used only to
5401 wake it up if nothing else does.
5403 16. A malformed maildirsize file could cause Exim to calculate negative values
5404 for the mailbox size or file count. Odd effects could occur as a result.
5405 The maildirsize information is now recalculated if the size or filecount
5408 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5409 support for a long time. Removed HAVE_SYS_VFS_H.
5411 18. Updated exipick to current release from John Jetmore.
5413 19. Allow an empty sender to be matched against a lookup in an address list.
5414 Previously the only cases considered were a regular expression, or an
5417 20. Exim went into a mad DNS lookup loop when doing a callout where the
5418 host was specified on the transport, if the DNS lookup yielded more than
5421 21. The RFC2047 encoding function was originally intended for short strings
5422 such as real names; it was not keeping to the 75-character limit for
5423 encoded words that the RFC imposes. It now respects the limit, and
5424 generates multiple encoded words if necessary. To be on the safe side, I
5425 have increased the buffer size for the ${rfc2047: expansion operator from
5428 22. Failure to deliver a bounce message always caused it to be frozen, even if
5429 there was an errors_to setting on the router. The errors_to setting is now
5432 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5433 canonical form (fully expanded) before being placed in
5434 $sender_host_address.
5436 24. Updated eximstats to version 1.33
5438 25. Include certificate and key file names in error message when GnuTLS fails
5439 to set them up, because the GnuTLS error message doesn't include the name
5440 of the failing file when there is a problem reading it.
5442 26. Expand error message when OpenSSL has problems setting up cert/key files.
5445 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5446 (this can affect the format of dates).
5448 28. exim_tidydb, when checking for the continued existence of a message for
5449 which it has found a message-specific retry record, was not finding
5450 messages that were in split spool directories. Consequently, it was
5451 deleting retry records that should have stayed in existence.
5453 29. eximstats updated to version 1.35
5454 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5455 1.35 - bugfix such that pie charts by volume are generated correctly
5457 30. The SPA authentication driver was not abandoning authentication and moving
5458 on to the next authenticator when an expansion was forced to fail,
5459 contradicting the general specification for all authenticators. Instead it
5460 was generating a temporary error. It now behaves as specified.
5462 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5463 (the order specifies the preference for clients). The order is now AES256,
5464 AES128, 3DES, ARCFOUR128.
5466 31. Small patch to Sieve code - explicitly set From: when generating an
5469 32. Exim crashed if a remote delivery caused a very long error message to be
5470 recorded - for instance if somebody sent an entire SpamAssassin report back
5471 as a large number of 550 error lines. This bug was coincidentally fixed by
5472 increasing the size of one of Exim's internal buffers (big_buffer) that
5473 happened as part of the Exiscan merge. However, to be on the safe side, I
5474 have made the code more robust (and fixed the comments that describe what
5477 33. Some experimental protocols are using DNS PTR records for new purposes. The
5478 keys for these records are domain names, not reversed IP addresses. The
5479 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5480 leaves it alone. Component reversal etc. now happens only for IP addresses.
5483 34. The host_aton() function is supposed to be passed a string that is known
5484 to be a valid IP address. However, in the case of IPv6 addresses, it was
5485 not checking this. This is a hostage to fortune. Exim now panics and dies
5486 if the condition is not met. A case was found where this could be provoked
5487 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5488 components; fortuitously, this particular loophole had already been fixed
5489 by change 4.50/55 or 4.44/33 above.
5491 If there are any other similar loopholes, the new check in host_aton()
5492 itself should stop them being exploited. The report I received stated that
5493 data on the command line could provoke the exploit when Exim was running as
5494 exim, but did not say which command line option was involved. All I could
5495 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5496 running as the user.
5499 35. There was a buffer overflow vulnerability in the SPA authentication code
5500 (which came originally from the Samba project). I have added a test to the
5501 spa_base64_to_bits() function which I hope fixes it.
5504 36. The daemon start-up calls getloadavg() while still root for those OS that
5505 need the first call to be done as root, but it missed one case: when
5506 deliver_queue_load_max is set with deliver_drop_privilege. This is
5507 necessary for the benefit of the queue runner, because there is no re-exec
5508 when deliver_drop_privilege is set.
5510 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5511 was in use, was not putting the data itself into the right store pool;
5512 consequently, it could be overwritten for a subsequent message in the same
5513 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5516 38. Sometimes the final signoff response after QUIT could fail to get
5517 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5518 < 0 before doing a fflush(). This bug looks as though it goes back to the
5519 introduction of TLS in release 3.20, but "sometimes" must have been rare
5520 because the tests only now provoked it.
5526 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5527 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5528 Now it returns an integer. A number of calls were still expecting a T/F
5529 return. Fortuitously, in all cases, the tests worked in OK situations,
5530 which is the norm. However, things would have gone wrong on any write
5531 failures on the smtp file descriptor. This function is used when sending
5532 messages over SMTP and also when doing verify callouts.
5534 2. When Exim is called to do synchronous delivery of a locally submitted
5535 message (the -odf or -odi options), it no longer closes stderr before doing
5538 3. Implemented the mua_wrapper option.
5540 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5542 5. Implemented the functions header_remove(), header_testname(),
5543 header_add_at_position(), and receive_remove_recipient(), and exported them
5546 6. If an ACL "warn" statement specified the addition of headers, Exim already
5547 inserted X-ACL-Warn: at the start if there was no header name. However, it
5548 was not making this test for the second and subsequent header lines if
5549 there were newlines in the string. This meant that an invalid header could
5550 be inserted if Exim was badly configured.
5552 7. Allow an ACL "warn" statement to add header lines at the start or after all
5553 the Received: headers, as well as at the end.
5555 8. Added the rcpt_4xx retry error code.
5557 9. Added postmaster_mailfrom=xxx to callout verification option.
5559 10. Added mailfrom=xxxx to the callout verification option, for verify=
5562 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5563 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5565 12. Inserted some casts to stop certain compilers warning when using pointer
5566 differences as field lengths or precisions in printf-type calls (mostly
5567 affecting debugging statements).
5569 13. Added optional readline() support for -be (dynamically loaded).
5571 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5572 same clock tick as a message's arrival, so that its received time was the
5573 same as the "first fail" time on the retry record, and that message
5574 remained on the queue past the ultimate address timeout, every queue runner
5575 would try a delivery (because it was past the ultimate address timeout) but
5576 after another failure, the ultimate address timeout, which should have then
5577 bounced the address, did not kick in. This was a "< instead of <=" error;
5578 in most cases the first failure would have been in the next clock tick
5579 after the received time, and all would be well.
5581 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5582 being recognized when the domain list was tested by the match_domain
5583 condition in an expansion string.
5585 16. Added the ${str2b64: operator.
5587 17. Exim was always calling setrlimit() to set a large limit for the number of
5588 processes, without checking whether the existing limit was already
5589 adequate. (It did check for the limit on file descriptors.) Furthermore,
5590 errors from getrlimit() and setrlimit() were being ignored. Now they are
5591 logged to the main and panic logs, but Exim does carry on, to try to do its
5592 job under whatever limits there are.
5594 18. Imported PCRE 5.0.
5596 19. Trivial typo in log message " temporarily refused connection" (the leading
5599 20. If the log selector return_path_on_delivery was set and an address was
5600 redirected to /dev/null, the delivery process crashed because it assumed
5601 that a return path would always be set for a "successful" delivery. In this
5602 case, the whole delivery is bypassed as an optimization, and therefore no
5605 21. Internal re-arrangement: the function for sending a challenge and reading
5606 a response while authentication was assuming a zero-terminated challenge
5607 string. It's now changed to take a pointer and a length, to allow for
5608 binary data in such strings.
5610 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5612 23. Exim was not respecting finduser_retries when seeking the login of the
5613 uid under which it was called; it was always trying 10 times. (The default
5614 setting of finduser_retries is zero.) Also, it was sleeping after the final
5615 failure, which is pointless.
5617 24. Implemented tls_on_connect_ports.
5619 25. Implemented acl_smtp_predata.
5621 26. If the domain in control=submission is set empty, Exim assumes that the
5622 authenticated id is a complete email address when it generates From: or
5623 Sender: header lines.
5625 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5626 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5627 chown and chgrp in /bin and hostname in /usr/bin.
5629 28. Exim was keeping the "process log" file open after each use, just as it
5630 does for the main log. This opens the possibility of it remaining open for
5631 long periods when the USR1 signal hits a daemon. Occasional processlog
5632 errors were reported, that could have been caused by this. Anyway, it seems
5633 much more sensible not to leave this file open at all, so that is what now
5636 29. The long-running daemon process does not normally write to the log once it
5637 has entered its main loop, and it closes the log before doing so. This is
5638 so that log files can straightforwardly be renamed and moved. However,
5639 there are a couple of unusual error situations where the daemon does write
5640 log entries, and I had neglected to close the log afterwards.
5642 30. The text of an SMTP error response that was received during a remote
5643 delivery was being truncated at 512 bytes. This is too short for some of
5644 the long messages that one sometimes sees. I've increased the limit to
5647 31. It is now possible to make retry rules that apply only when a message has a
5648 specific sender, in particular, an empty sender.
5650 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5651 it possible to be selective about when SMTP synchronization is enforced.
5653 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5655 32. Implemented hosts_connection_nolog.
5657 33. Added an ACL for QUIT.
5659 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5662 35. Added mailbox_size and mailbox_filecount to appendfile.
5664 36. Added control = no_multiline_responses to ACLs.
5666 37. There was a bug in the logic of the code that waits for the clock to tick
5667 in the case where the clock went backwards by a substantial amount such
5668 that the microsecond fraction of "now" was more than the microsecond
5669 fraction of "then" (but the whole seconds number was less).
5671 38. Added support for the libradius Radius client library this is found on
5672 FreeBSD (previously only the radiusclient library was supported).
5678 1. When certain lookups returned multiple values in the form name=value, the
5679 quoting of the values was not always being done properly. Specifically:
5680 (a) If the value started with a double quote, but contained no whitespace,
5682 (b) If the value contained whitespace other than a space character (i.e.
5683 tabs or newlines or carriage returns) it was not quoted.
5684 This fix has been applied to the mysql and pgsql lookups by writing a
5685 separate quoting function and calling it from the lookup code. The fix
5686 should probably also be applied to nisplus, ibase and oracle lookups, but
5687 since I cannot test any of those, I have not disturbed their existing code.
5689 2. A hit in the callout cache for a specific address caused a log line with no
5690 reason for rejecting RCPT. Now it says "Previous (cached) callout
5691 verification failure".
5693 3. There was an off-by-one bug in the queryprogram router. An over-long
5694 return line was truncated at 256 instead of 255 characters, thereby
5695 overflowing its buffer with the terminating zero. As well as fixing this, I
5696 have increased the buffer size to 1024 (and made a note to document this).
5698 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5699 when Exim is waiting for an SMTP response from a remote server, Exim
5700 restarts its select() call on the socket, thereby resetting its timeout.
5701 This is not a problem when such interrupts are rare. Somebody set up a cron
5702 job to run exiwhat every 2 minutes, which is less than the normal select()
5703 timeout (5 or 10 minutes). This meant that the select() timeout never
5704 kicked in because it was always reset. I have fixed this by comparing the
5705 time when an interrupt arrives with the time at the start of the first call
5706 to select(). If more time than the timeout has elapsed, the interrupt is
5707 treated as a timeout.
5709 5. Some internal re-factoring in preparation for the addition of Sieve
5710 extensions (by MH). In particular, the "personal" test is moved to a
5711 separate function, and given an option for scanning Cc: and Bcc: (which is
5712 not set for Exim filters).
5714 6. When Exim created an email address using the login of the caller as the
5715 local part (e.g. when creating a From: or Sender: header line), it was not
5716 quoting the local part when it contained special characters such as @.
5718 7. Installed new OpenBSD configuration files.
5720 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5721 try to make them clearer.
5723 9. Callout options, other than the timeout value, were being ignored when
5724 verifying sender addresses in header lines. For example, when using
5726 verify = header_sender/callout=no_cache
5728 the cache was (incorrectly) being used.
5730 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5731 only the Cygwin environment.
5733 11. When return_path_on_delivery was set as a log selector, if different remote
5734 addresses in the same message used different return paths and parallel
5735 remote delivery occurred, the wrong values would sometimes be logged.
5736 (Whenever a remote delivery process finished, the return path value from
5737 the most recently started remote delivery process was logged.)
5739 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5740 lines when AUTH and/or TLS are in use. This is the "received protocol"
5741 field. Exim used to use "asmtp" for authenticated SMTP, without any
5742 indication (in the protocol name) for TLS use. Now it follows the RFC and
5743 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5744 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5745 names appear in log lines as well as in Received: header lines.
5747 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5748 extensions, and comparison tests, and to fix some bugs.
5750 14. Changes to the "personal" filter test:
5752 (1) The test was buggy in that it was just doing the equivalent of
5753 "contains" tests on header lines. For example, if a user's address was
5754 anne@some.where, the "personal" test would incorrectly be true for
5756 To: susanne@some.where
5758 This test is now done by extracting each address from the header in turn,
5759 and checking the entire address. Other tests that are part of "personal"
5760 are now done using regular expressions (for example, to check local parts
5761 of addresses in From: header lines).
5763 (2) The list of non-personal local parts in From: addresses has been
5764 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5765 taken from the Sieve specification recommendations.
5767 (3) If the message contains any header line starting with "List-" it is
5768 treated as non-personal.
5770 (4) The test for "circular" in the Subject: header line has been removed
5771 because it now seems ill-conceived.
5773 15. Minor typos in src/EDITME comments corrected.
5775 16. Installed latest exipick from John Jetmore.
5777 17. If headers_add on a router specified a text string that was too long for
5778 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5779 of string_sprintf() is now avoided.
5781 18. $message_body_size was not set (it was always zero) when running the DATA
5782 ACL and the local_scan() function.
5784 19. For the "mail" command in an Exim filter, no default was being set for
5785 the once_repeat time, causing a random time value to be used if "once" was
5786 specified. (If the value happened to be <= 0, no repeat happened.) The
5787 default is now 0s, meaning "never repeat". The "vacation" command was OK
5788 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5789 (I found it when inspecting the code).
5791 20. There is now an overall timeout for performing a callout verification. It
5792 defaults to 4 times the callout timeout, which applies to individual SMTP
5793 commands during the callout. The overall timeout applies when there is more
5794 than one host that can be tried. The timeout is checked before trying the
5795 next host. This prevents very long delays if there are a large number of
5796 hosts and all are timing out (e.g. when the network connections are timing
5797 out). The value of the overall timeout can be changed by specifying an
5798 additional sub-option for "callout", called "maxwait". For example:
5800 verify = sender/callout=5s,maxwait=20s
5802 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5803 to the end before writing, but this should make it even safer).
5805 22. Exim was forgetting that it had advertised PIPELINING for the second and
5806 subsequent messages on an SMTP connection. It was also not resetting its
5807 memory on STARTTLS and an internal HELO.
5809 23. When Exim logs an SMTP synchronization error within a session, it now
5810 records whether PIPELINING has been advertised or not.
5812 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5813 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5814 rather than long int.
5816 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5818 26. Added the never_mail option to autoreply.
5824 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5825 crash if the getsockname() call failed; this can happen if a connection is
5826 closed very soon after it is established. The problem was simply in the
5827 order in which certain operations were done, causing Exim to try to write
5828 to the SMTP stream before it had set up the file descriptor. The bug has
5829 been fixed by making things happen in the correct order.
5835 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5836 before the connection was closed, thus losing the rejection response.
5838 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5839 some early Solaris releases, but causes trouble in current releases where
5840 socklen_t is defined.
5842 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5845 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5848 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5849 that releasing the top of it at the end releases what was used for sub-
5850 expansions (unless the block got too big). However, discard this block if
5851 the first thing is a variable or header, so that we can use its block when
5852 it is dynamic (useful for very large $message_headers, for example).
5854 6. Lookups now cache *every* query, not just the most recent. A new, separate
5855 store pool is used for this. It can be recovered when all lookup caches are
5856 flushed. Lookups now release memory at the end of their result strings.
5857 This has involved some general refactoring of the lookup sources.
5859 7. Some code has been added to the store_xxx() functions to reduce the amount
5860 of flapping under certain conditions.
5862 8. log_incoming_interface used to affect only the <= reception log lines. Now
5863 it causes the local interface and port to be added to several more SMTP log
5864 lines, for example "SMTP connection from", and rejection lines.
5866 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5868 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5870 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5871 could be overwritten at the end of the current message (or the start of a
5872 new message if it was set in a HELO ACL). The value is now preserved for
5873 the duration of the SMTP connection.
5875 12. If a transport had a headers_rewrite setting, and a matching header line
5876 contained an unqualified address, that address was qualified, even if it
5877 did not match any rewriting rules. The underlying bug was that the values
5878 of the flags that permit the existence of unqualified sender and recipient
5879 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5880 non-local messages, and by -bnq for local messages) were not being
5881 preserved with the message after it was received.
5883 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5884 "next input=" as part of the text comprising the host identity instead of
5885 the correct text. The code was using the same buffer for two different
5886 strings. However, depending on which order the printing function evaluated
5887 its arguments, the bug did not always show up. Under Linux, for example, my
5888 test suite worked just fine.
5890 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5891 doesn't work with some older versions of Perl. It has been changed to "my",
5892 which in any case is probably the better facility to use.
5894 15. A really picky compiler found some instances of statements for creating
5895 error messages that either had too many or two few arguments for the format
5898 16. The size of the buffer for calls to the DNS resolver has been increased
5899 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5900 for addresses that have a lot of PTR records. This alleviates a problem; it
5901 does not fully solve it.
5903 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5904 buffer now truncates the list and logs the incident, which is the same
5905 action as happens when Exim is looking up a host name and its aliases.
5906 Previously in this situation something unpredictable would happen;
5907 sometimes it was "internal error: store_reset failed".
5909 18. If a server dropped the connection unexpectedly when an Exim client was
5910 using GnuTLS and trying to read a response, the client delivery process
5911 crashed while trying to generate an error log message.
5913 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5914 string, for example:
5916 warn message = H1: something\nH2: something
5918 the text was added as a single header line from Exim's point of view
5919 though it ended up OK in the delivered message. However, searching for the
5920 second and subsequent header lines using $h_h2: did not work. This has been
5921 fixed. Similarly, if a system filter added multiple headers in this way,
5922 the routers could not see them.
5924 20. Expanded the error message when iplsearch is called with an invalid key to
5925 suggest using net-iplsearch in a host list.
5927 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5928 ACLs are no longer actually imposed (and a message to that effect is
5931 22. If a "gecos" field in a passwd entry contained escaped characters, in
5932 particular, if it contained a \" sequence, Exim got it wrong when building
5933 a From: or a Sender: header from that name. A second bug also caused
5934 incorrect handling when an unquoted " was present following a character
5935 that needed quoting.
5937 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5938 was not being matched caselessly.
5940 24. Arranged for all hyphens in the exim.8 source to be escaped with
5943 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5944 properly. Recipient callout cache records were still being keyed to include
5945 the sender, even when use_sender was set false. This led to far more
5946 callouts that were necessary. The sender is no longer included in the key
5947 when use_sender is false.
5949 26. Added "control = submission" modifier to ACLs.
5951 27. Added the ${base62d: operator to decode base 62 numbers.
5953 28. dnsdb lookups can now access SRV records.
5955 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5956 the configuration file.
5958 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5959 (-v) mode. This makes the output for a verbose queue run more intelligible.
5961 31. Added a use_postmaster feature to recipient callouts.
5963 32. Added the $body_zerocount variable, containing the number of binary zero
5964 bytes in the message body.
5966 33. The time of last modification of the "new" subdirectory is now used as the
5967 "mailbox time last read" when there is a quota error for a maildir
5970 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5972 35. Added +ignore_unknown as a special item in host lists.
5974 36. Code for decoding IPv6 addresses in host lists is now included, even if
5975 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5976 address was recognized as an IP address, but was then not correctly decoded
5977 into binary, causing unexpected and incorrect effects when compared with
5984 1. Very minor rewording of debugging text in manualroute to say "list of
5985 hosts" instead of "hostlist".
5987 2. If verify=header_syntax was set, and a header line with an unqualified
5988 address (no domain) and a large number of spaces between the end of the
5989 name and the colon was received, the reception process suffered a buffer
5990 overflow, and (when I tested it) crashed. This was caused by some obsolete
5991 code that should have been removed. The fix is to remove it!
5993 3. When running in the test harness, delay a bit after writing a bounce
5994 message to get a bit more predictability in the log output.
5996 4. Added a call to search_tidyup() just before forking a reception process. In
5997 theory, someone could use a lookup in the expansion of smtp_accept_max_
5998 per_host which, without the tidyup, could leave open a database connection.
6000 5. Added the variables $recipient_data and $sender_data which get set from a
6001 lookup success in an ACL "recipients" or "senders" condition, or a router
6002 "senders" option, similar to $domain_data and $local_part_data.
6004 6. Moved the writing of debug_print from before to after the "senders" test
6007 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6008 problems for message scanning, either using a data ACL, or using
6009 local_scan() because the Received: header was not generated till after they
6010 were called (in order to set the time as the time of reception completion).
6011 I have revised the way this works. The header is now generated after the
6012 body is received, but before the ACL or local_scan() are called. After they
6013 are run, the timestamp in the header is updated.
6019 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6020 before starting a queue runner without re-exec. This happened only when
6021 deliver_drop_privilege was set or when the Exim user was set to root. The
6022 effect of the bug was that timeouts during subsequent deliveries caused
6023 crashes instead of being properly handled. The handler is now left at its
6024 default (and expected) setting.
6026 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6027 message, again when deliver_drop_privilege is set or Exim is run as root.
6028 The bug described in (1) was not present in this case, but the tidying up
6029 of the other signals was missing. I have made the two cases consistent.
6031 3. The ignore_target_hosts setting on a manualroute router was being ignored
6032 for hosts that were looked up using the /MX notation.
6034 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6037 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6038 operated on the sender address. After changing the $sender_address to <>
6039 for the sender address verify, Exim was re-instated it as the original
6040 (before rewriting) address, but remembering that it had rewritten it, so it
6041 wasn't rewriting it again. This bug also had the effect of breaking the
6042 sender address verification caching when the sender address was rewritten.
6044 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6045 This has been changed so that if the ip literal address matches
6046 ignore_target_hosts, the router declines.
6048 7. Added expansion conditions match_domain, match_address, and match_local_
6049 part (NOT match_host).
6051 8. The placeholder for the Received: header didn't have a length field set.
6053 9. Added code to Exim itself and to exim_lock to test for a specific race
6054 condition that could lead to file corruption when using MBX delivery. The
6055 issue is with the lockfile that is created in /tmp. If this file is removed
6056 after a process has opened it but before that process has acquired a lock,
6057 there is the potential for a second process to recreate the file and also
6058 acquire a lock. This could lead to two Exim processes writing to the file
6059 at the same time. The added code performs the same test as UW imapd; it
6060 checks after acquiring the lock that its file descriptor still refers to
6061 the same named file.
6063 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6064 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6067 11. Added the smtp_active_hostname option. If used, this will typically be made
6068 to depend on the incoming interface address. Because $interface_address is
6069 not set up until the daemon has forked a reception process, error responses
6070 that can happen earlier (such as "too many connections") no longer contain
6073 12. If an expansion in a condition on a "warn" statement fails because a lookup
6074 defers, the "warn" statement is abandoned, and the next ACL statement is
6075 processed. Previously this caused the whole ACL to be aborted.
6077 13. Added the iplsearch lookup type.
6079 14. Added ident_timeout as a log selector.
6081 15. Added tls_certificate_verified as a log selector.
6083 16. Added a global option tls_require_ciphers (compare the smtp transport
6084 option of the same name). This controls incoming TLS connections.
6086 17. I finally figured out how to make tls_require_ciphers do a similar thing
6087 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6088 before starting the TLS session.
6090 18. Tabs are now shown as \t in -bP output.
6092 19. If the log selector return_path_on_delivery was set, Exim crashed when
6093 bouncing a message because it had too many Received: header lines.
6095 20. If two routers both had headers_remove settings, and the first one included
6096 a superfluous trailing colon, the final name in the first list and the
6097 first name in the second list were incorrectly joined into one item (with a
6098 colon in the middle).
6104 1. Added -C and -D options to the exinext utility, mainly to make it easier
6105 to include in the automated testing, but these could be helpful when
6106 multiple configurations are in use.
6108 2. The exinext utility was not formatting the output nicely when there was
6109 an alternate port involved in the retry record key, nor when there was a
6110 message id as well (for retries that were specific to a specific message
6111 and a specific host). It was also confused by IPv6 addresses, because of
6112 the additional colons they contain. I have fixed the IPv4 problem, and
6113 patched it up to do a reasonable job for IPv6.
6115 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6116 delivery, the log line now contains "pipelined" if PIPELINING was used.
6118 4. An SMTP transport process used to panic and die if the bind() call to set
6119 an explicit outgoing interface failed. This has been changed; it is now
6120 treated in the same way as a connect() failure.
6122 5. A reference to $sender_host_name in the part of a conditional expansion
6123 that was being skipped was still causing a DNS lookup. This no longer
6126 6. The def: expansion condition was not recognizing references to header lines
6127 that used bh_ and bheader_.
6129 7. Added the _cache feature to named lists.
6131 8. The code for checking quota_filecount in the appendfile transport was
6132 allowing one more file than it should have been.
6134 9. For compatibility with Sendmail, the command line option
6142 and sets the incoming protocol and host name (for trusted callers). The
6143 host name and its colon can be omitted when only the protocol is to be set.
6144 Note the Exim already has two private options, -pd and -ps, that refer to
6145 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6146 "s", but I don't think that's a major issue.
6148 10. A number of refactoring changes to the code, none of which should affect
6151 (a) The number of logging options was getting close to filling up the
6152 32-bit word that was used as a bit map. I have split them into two classes:
6153 those that are passed in the argument to log_write(), and those that are
6154 only ever tested independently outside of that function. These are now in
6155 separate 32-bit words, so there is plenty of room for expansion again.
6156 There is no change in the user interface or the logging behaviour.
6158 (b) When building, for example, log lines, the code previously used a
6159 macro that called string_cat() twice, in order to add two strings. This is
6160 not really sufficiently general. Furthermore, there was one instance where
6161 it was actually wrong because one of the argument was used twice, and in
6162 one call a function was used. (As it happened, calling the function twice
6163 did not affect the overall behaviour.) The macro has been replaced by a
6164 function that can join an arbitrary number of extra strings onto a growing
6167 (c) The code for expansion conditions now uses a table and a binary chop
6168 instead of a serial search (which was left over from when there were very
6169 few conditions). Also, it now recognizes conditions like "pam" even when
6170 the relevant support is not compiled in: a suitably worded error message is
6171 given if an attempt is made to use such a condition.
6173 11. Added ${time_interval:xxxxx}.
6175 12. A bug was causing one of the ddress fields not to be passed back correctly
6176 from remote delivery subprocesses. The field in question was not being
6177 subsequently used, so this caused to problems in practice.
6179 13. Added new log selectors queue_time and deliver_time.
6181 14. Might have fixed a bug in maildirsizefile handling that threw up
6182 "unexpected character" debug warnings, and recalculated the data
6183 unnecessarily. In any case, I expanded the warning message to give more
6186 15. Added the message "Restricted characters in address" to the statements in
6187 the default ACL that block characters like @ and % in local parts.
6189 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6190 Three changes have been made:
6192 (a) There was a serious bug; a negative response to MAIL caused the whole
6193 recipient domain to be cached as invalid, thereby blocking all messages
6194 to all local parts at the same domain, from all senders. This bug has
6195 been fixed. The domain is no longer cached after a negative response to
6196 MAIL if the sender used is not empty.
6198 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6201 (c) A new callout option, "use_sender" has been added for people who want
6202 the modified behaviour.
6208 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6211 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6212 indeed breaks things for older releases.
6214 3. Added additional logging to the case where there is a problem reading data
6215 from a filter that is running in a subprocess using a pipe, in order to
6216 try to track down a specific problem.
6218 4. Testing facility fudge: when running in the test harness and attempting
6219 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6220 getting "No route to host". Convert this to a timeout.
6222 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6225 6. Some OS don't have socklen_t but use size_t instead. This affects the
6226 fifth argument of getsockopt() amongst other things. This is now
6227 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6228 can be set for individual OS. I have set it for SunOS5, OSF1, and
6229 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6230 some earlier ones do not.
6232 7. Change 4.30/15 was not doing the test caselessly.
6234 8. The standard form for an IPv6 address literal was being rejected by address
6235 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6236 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6237 this, as well as the form without the "IPv6" on the front (but only when
6238 address literals are enabled, of course).
6240 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6242 10. Exim crashed if a message with an empty sender address specified by -f
6243 encountered a router with an errors_to setting. This could be provoked only
6244 by a command such as
6248 where an empty string was supplied; "<>" did not hit this bug.
6250 11. Installed PCRE release 4.5.
6252 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6253 remained set. It is now erased.
6255 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6256 times from message ids (which are base 36 rather than the normal 62).
6258 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6259 were being counted as actual protocol errors, and logged if the log
6260 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6261 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6262 and DATA following a set of rejected RCPTs do not count as protocol errors.
6263 In other words, Exim assumes they were pipelined, though this may not
6264 actually be the case. Of course, in all cases the client gets an
6265 appropriate error code.
6267 15. If a lookup fails in an ACL condition, a message about the failure may
6268 be available; it is used if testing the ACL cannot continue, because most
6269 such messages specify what the cause of the deferral is. However, some
6270 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6271 that caused an old message to be retained and used if a later statement
6272 caused a defer, replacing the real cause of the deferral.
6274 16. If an IP address had so many PTR records that the DNS lookup buffer
6275 was not large enough to hold them, Exim could crash while trying to process
6276 the truncated data. It now detects and logs this case.
6278 17. Further to 4.21/58, another change has been made: if (and only if) the
6279 first line of a message (the first header line) ends with CRLF, a bare LF
6280 in a subsequent header line has a space inserted after it, so as not to
6281 terminate the header.
6283 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6284 unnecessarily, used atoi() instead of strtol(), and didn't check the
6285 termination when getting file sizes from file names by regex.
6287 19. Completely re-implemented the support for maildirsize files, in the light
6288 of a number of problems with the previous contributed implementation
6289 (4.30/29). In particular:
6291 . If the quota is zero, the maildirsize file is maintained, but no quota is
6294 . If the maildir directory does not exist, it is created before any attempt
6295 to write a maildirsize file.
6297 . The quota value in the file is just a cache; if the quota is changed in
6298 the transport, the new value overrides.
6300 . A regular expression is available for excluding directories from the
6303 20. The autoreply transport checks the characters in options that define the
6304 message's headers; it allows continued headers, but it was checking with
6305 isspace() after an embedded newline instead of explicitly looking for a
6308 21. If all the "regular" hosts to which an address was routed had passed their
6309 expiry times, and had not reached their retry times, the address was
6310 bounced, even if fallback hosts were defined. Now Exim should go on to try
6313 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6314 equivalent code in the SMTP transport. Some hosts send humungous responses
6315 to HELO/EHLO, more than 1024 it seems.
6317 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6318 gives compiler warnings in some environments. I've now done it "properly",
6321 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6322 (because of problems with the built-in one) was declared to return uschar *
6323 instead of char *, causing compiler failure.
6325 25. Fixed a file descriptor leak when processing alias/forward files.
6327 26. Fixed a minor format string issue in dbfn.c.
6329 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6331 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6332 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6335 29. When a sender address is verified, it is cached, to save repeating the test
6336 when there is more than one recipient in a message. However, when the
6337 verification involves a callout, it is possible for different callout
6338 options to be set for different recipients. It is too complicated to keep
6339 track of this in the cache, so now Exim always runs a verification when a
6340 callout is required, relying on the callout cache for the optimization.
6341 The overhead is duplication of the address routing, but this should not be
6344 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6345 to be verified with callout=postmaster, and the main callout worked but the
6346 postmaster check failed, the verification correctly failed. However, if a
6347 subsequent RCPT command asked for sender verification *without* the
6348 postmaster check, incorrect caching caused this verification also to fail,
6351 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6352 it was not caching the DNS options (qualify_single, search_parents) that
6353 were used when the lookup failed. A subsequent lookup with different
6354 options therefore always gave the same answer, though there were cases
6355 where it should not have. (Example: a "domains = !$mx_any" option on a
6356 dnslookup router: the "domains" option is always processed without any
6357 widening, but the router might have qualify_single set.) Now Exim uses the
6358 cached value only when the same options are set.
6360 32. Added John Jetmore's "exipick" utility to the distribution.
6362 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6363 than a timeout (e.g. a certificate is required, and is not provided), an
6364 Exim server now closes the connection immediately. Previously it waited for
6365 the client to close - but if the client is SSL, it seems that they each
6366 wait for each other, leading to a delay before one of them times out.
6368 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6369 maintained 0.8.x compatibility because I don't think many are using it, and
6370 it is clearly obsolete.
6372 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6375 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6376 client certificate was expired. A simple patch fixes this, though I don't
6377 understand the full logic of why the verify callback is called multiple
6380 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6381 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6382 which causes problems with some clients (such as the Certicom SSL Plus
6383 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6384 disables the coutermeasure allowing Eudora to connect."
6386 38. Exim was not checking that a write() to a log file succeeded. This could
6387 lead to Bad Things if a log got too big, in particular if it hit a file
6388 size limit. Exim now panics and dies if it cannot write to a log file, just
6389 as it does if it cannot open a log file.
6391 39. Modified OS/Makefile-Linux so that it now contains
6393 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6395 The two -D definitions ensure that Exim is compiled with large file
6396 support, which makes it possible to handle log files that are bigger than
6399 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6400 instance) a domain was checked against a named list that involved a lookup,
6401 causing $domain_data to be set, then another domain was checked against the
6402 same list, then the first domain was re-checked, the value of $domain_data
6403 after the final check could be wrong. In particular, if the second check
6404 failed, it could be set empty. This bug probably also applied to
6407 41. The strip_trailing_dot option was not being applied to the address given
6408 with the -f command-line option.
6410 42. The code for reading a message's header from the spool was incrementing
6411 $received_count, but never initializing it. This meant that the value was
6412 incorrect (doubled) while delivering a message in the same process in which
6413 it was received. In the most common configuration of Exim, this never
6414 happens - a fresh exec is done - but it can happen when
6415 deliver_drop_privilege is set.
6417 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6418 it now includes up to 150 characters of the unexpected data in the log
6421 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6422 and building data strings. The size of both of these buffers was 10 000
6423 bytes - far larger than anybody would *ever* want, thought I. Needless to
6424 say, somebody hit the limit. I have increased the maximum line length to
6425 20 000 and the maximum data length of concatenated lines to 100 000. I have
6426 also fixed two bugs, because there was no checking on these buffers. Tsk,
6427 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6428 buffer is too small.
6430 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6431 lsearch lookups. Now it does.
6433 46. When parsing a route_list item in a manualroute router, a fixed-length
6434 buffer was used for the list of hosts. I made this 1024 bytes long,
6435 thinking that nobody would ever have a list of hosts that long. Wrong.
6436 Somebody had a whole pile of complicated expansion conditions, and the
6437 string was silently truncated, leading to an expansion error. It turns out
6438 that it is easier to change to an unlimited length (owing to other changes
6439 that have happened since this code was originally written) than to build
6440 structure for giving a limitation error. The length of the item that
6441 expands into the list of hosts is now unlimited.
6443 47. The lsearch lookup could not handle data where the length of text line was
6444 more than 4095 characters. Such lines were truncated, leading to shortened
6445 data being returned. It should now handle lines of any length.
6447 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6448 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6451 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6452 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6453 no longer happens in comment lines. A list of replacements is now placed
6454 at the head of all of the source files, except those whose only change is
6455 to replace PERL_COMMAND in the very first #! line.
6457 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6458 messages on the queue, with a bottom-up merge sort, using code contributed
6459 by Michael Haardt. This should make operations like -bp somewhat faster on
6460 large queues. It won't affect queue runners, except when queue_run_in_order
6463 51. Installed eximstats 1.31 in the distribution.
6465 52. Added support for SRV lookups to the dnslookup router.
6467 53. If an ACL referred to $message_body or $message_body_end, the value was not
6468 reset for any messages that followed in the same SMTP session.
6470 54. The store-handling optimization for building very long strings was not
6471 differentiating between the different store pools. I don't think this
6472 actually made any difference in practice, but I've tidied it.
6474 55. While running the routers to verify a sender address, $sender_address
6475 was still set to the sender address. This is wrong, because when routing to
6476 send a bounce to the sender, it would be empty. Therefore, I have changed
6477 it so that, while verifying a sender address, $sender_address is set to <>.
6478 (There is no change to what happens when verifying a recipient address.)
6480 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6481 target A or AAAA records (if not already returned) without resetting the
6482 qualify_single or search_parents options of the DNS resolver. These are
6483 inappropriate in this case because the targets of MX and SRV records must
6484 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6485 target that, when qualified, matched something in the local domain. These
6486 two options are now turned off when doing these lookups.
6488 57. It seems that at least some releases of Reiserfs (which does not have the
6489 concept of a fixed number of inodes) returns zero and not -1 for the
6490 number of available inodes. This interacted badly with check_spool_inodes,
6491 which assumed that -1 was the "no such thing" setting. What I have done is
6492 to check that the total number of inodes is greater than zero before doing
6493 the test of how many are available.
6495 58. When a "warn" ACL statement has a log_message modifier, the message is
6496 remembered, and not repeated. This is to avoid a lot of repetition when a
6497 message has many recipients that cause the same warning to be written.
6498 However, Exim was preserving the list of already written lines for an
6499 entire SMTP session, which doesn't seem right. The memory is now reset if a
6500 new message is started.
6502 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6503 debugging output unless log_rewrite was also set.
6505 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6506 of (int)(handle) when we know that handle contains (void *)(-1).
6508 61. The Exim daemon panic-logs an error return when it closes the incoming
6509 connection. However "connection reset by peer" seems to be common, and
6510 isn't really an error worthy of noting specially, so that particular error
6513 62. When Exim is trying to find all the local interfaces, it used to panic and
6514 die if the ioctl to get the interface flags failed. However, it seems that
6515 on at least one OS (Solaris 9) it is possible to have an interface that is
6516 included in the list of interfaces, but for which you get a failure error
6517 for this call. This happens when the interface is not "plumbed" into a
6518 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6519 failure of the "get flags" call assumes that the interface is down.
6521 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6522 makes life easier for people who are doing arithmetic on fields extracted
6523 from dates, where you often get leading zeros that should not be
6524 interpreted as octal.
6526 64. Added qualify_domain to the redirect router, to override the global
6529 65. If a pathologically long header line contained very many addresses (the
6530 report of this problem mentioned 10 000) and each of them was rewritten,
6531 Exim could use up a very large amount of memory. (It kept on making new
6532 copies of the header line as it rewrote, and never released the old ones.)
6533 At the expense of a bit more processing, the header rewriting function has
6534 been changed so that it no longer eats memory in this way.
6536 66. The generation of the Received: header has been moved from the time that a
6537 message starts to be received, to the time that it finishes. The timestamp
6538 in the Received: header should now be very close to that of the <= log
6539 line. There are two side-effects of this change:
6541 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6542 logged header lines no longer include the local Received: line, because
6543 it has not yet been created. The same applies to a copy of the message
6544 that is returned to a non-SMTP sender when a message is rejected.
6546 (b) When a filter file is tested using -bf, no additional Received: header
6547 is added to the test message. After some thought, I decided that this
6550 This change does not affect the value of $received_for. It is still set
6551 after address rewriting, but before local_scan() is called.
6553 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6555 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6556 gave an unhelpful panic error message, and a defer error. I have managed to
6557 change this behaviour so that it now rejects any supplied certificate,
6558 which seems right, as the list of acceptable certificates is empty.
6560 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6561 gave an unhelpful defer error. I have not managed to make this reject any
6562 supplied certificates, but the error message it gives is "no certificate
6563 supplied", which is not helpful.
6565 70. exigrep's output now also includes lines that are not associated with any
6566 message, but which match the given pattern. Implemented by a patch from
6567 Martin Sluka, which also tidied up the Perl a bit.
6569 71. Recipient callout verification, like sender verification, was using <> in
6570 the MAIL FROM command. This isn't really the right thing, since the actual
6571 sender may affect whether the remote host accepts the recipient or not. I
6572 have changed it to use the actual sender in the callout; this means that
6573 the cache record is now keyed on a recipient/sender pair, not just the
6574 recipient address. There doesn't seem to be a real danger of callout loops,
6575 since a callout by the remote host to check the sender would use <>.
6576 [SEE ABOVE: changed after hitting problems.]
6578 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6579 temporary errors. However, in the case of such a code being given after
6580 the end of a data transmission (i.e. after ".") Exim was failing to write
6581 a retry record for the message. (Yes, there was some broken host that was
6582 actually sending 8xx at this point.)
6584 73. An unknown lookup type in a host list could cause Exim to panic-die when
6585 the list was checked. (An example that provoked this was putting <; in the
6586 middle of a list instead of at the start.) If this happened during a DATA
6587 ACL check, a -D file could be left lying around. This kind of configuration
6588 error no longer causes Exim to die; instead it causes a defer error. The
6589 incident is still logged to the main and panic logs.
6591 74. Buglet left over from Exim 3 conversion. The message "too many messages
6592 in one connection" was written to the rejectlog but not the mainlog, except
6593 when address rewriting (yes!) was being logged.
6595 75. Added write_rejectlog option.
6597 76. When a system filter was run not as root (that is, when system_filter_user
6598 was set), the values of the $n variables were not being returned to the
6599 main process; thus, they were not subsequently available in the $sn
6602 77. Added +return_path_on_delivery log selector.
6604 78. A connection timeout was being treated differently from recipients deferred
6605 when testing hosts_max_try with a message that was older than the host's
6606 retry timeout. (The host should not be counted, thus allowing all hosts to
6607 be tried at least once before bouncing.) This may have been the cause of an
6608 occasionally reported bug whereby a message would remain on the queue
6609 longer than the retry timeout, but would be bounced if a delivery was
6610 forced. I say "may" because I never totally pinned down the problem;
6611 setting up timeout/retry tests is difficult. See also the next item.
6613 79. The ultimate address timeout was not being applied to errors that involved
6614 a combination of host plus message (for example, a timeout on a MAIL
6615 command). When an address resolved to a number of possible hosts, and they
6616 were not all tried for each delivery (e.g. because of hosts_max_try), a
6617 message could remain on the queue longer than the retry timeout.
6619 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6622 81. Fixed an obscure SMTP outgoing bug which required at least the following
6623 conditions: (a) there was another message waiting for the same server;
6624 (b) the server returned 5xx to all RCPT commands in the first message so
6625 that the message was not completed; (c) the server dropped the connection
6626 or gave a negative response to the RSET that Exim sends to abort the
6627 transaction. The observed case was a dropped connection after DATA that had
6628 been sent in pipelining mode. That is, the server had advertised PIPELINING
6629 but was not implementing it correctly. The effect of the bug was incorrect
6630 behaviour, such as trying another host, and this could lead to a crash.
6636 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6637 and daemon.c were passed as pointers to ints; they should have been
6638 pointers to socklen_t variables (which are typically unsigned ints).
6640 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6643 3. Fixed a really odd bug that affected only the testing scheme; patching a
6644 certain fixed string in the binary changed the value of another string that
6645 happened to be identical to the end of the original first string.
6647 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6648 name", it returns that address as the IP address. On some operating
6649 systems (e.g. Solaris), it also passes back the IP address string as the
6650 "host name". However, on others (e.g. Linux), it passes back an empty
6651 string. Exim wasn't checking for this, and was changing the host name to an
6652 empty string, assuming it had been canonicalized.
6654 5. Although rare, it is permitted to have more than one PTR record for a given
6655 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6656 all the names associated with an address, because they do in Solaris.
6657 However, it seems that they do not in Linux for data that comes from the
6658 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6659 I found this out when I moved to a new Linux workstation and tried to run
6660 the Exim test suite.
6662 To get round this problem I have changed the code so that it now does its
6663 own call to the DNS to look up PTR records when searching for a host name.
6664 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6665 addresses that are only in /etc/hosts are still found.
6667 This behaviour is, however, controlled by an option called host_lookup_
6668 order, which defaults to "bydns:byaddr". If people want to use the other
6669 order, or indeed, just use one or the other means of lookup, they can
6670 specify it in this variable.
6672 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6673 some operating systems, this comes back from gethostbyaddr() as an empty
6674 string, and this is what Exim used to test for. However, it seems that in
6675 other systems, "." is yielded. Exim now tests for this case too.
6677 7. The values of check_spool_space and check_log_space are now held internally
6678 as a number of kilobytes instead of an absolute number of bytes. If a
6679 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6680 kilobyte. This means that much larger values can be stored.
6682 8. Exim monitor: an attempt to get the action menu when not actually pointing
6683 at a message produces an empty menu entitled "No message selected". This
6684 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6685 no entries in it ("Shell widget menu has zero width and/or height"). So I
6686 have added a single, blank menu entry in this case.
6688 9. Added ${quote_local_part.
6690 10. MIME decoding is now applied to the contents of Subject: header lines when
6693 11. Now that a reference to $sender_host_address automatically causes a reverse
6694 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6695 host lookup before query-style lookups in lists that might use this
6696 variable. This has therefore been abolished, and the "net-" prefix is no
6697 longer necessary for query-style lookups.
6699 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6700 have been a typo for LFLAGS, so it has been changed.
6702 13. The install script calls Exim with "-C /dev/null" in order to find the
6703 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6704 to be output. However, since Exim outputs its version number before the
6705 error, it didn't break the script. It just looked ugly. I fixed this by
6706 always allowing "-C /dev/null" if the caller is root.
6708 14. Ignore overlarge ACL variable number when reading spool file - insurance
6709 against a later release with more variables having written the file.
6711 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6712 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6713 this, as well as the form without the "IPv6" on the front.
6715 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6716 OS/Makefile-Darwin file.
6718 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6719 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6721 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6722 variable. However, some OS use ints for inodes. I've added cast to long int
6723 to get rid of the compiler warning.
6725 19. I had forgotten to lock out "/../" in configuration file names when
6726 ALT_CONFIG_PREFIX was set.
6728 20. Routers used for verification do not need to specify transports. However,
6729 if such a router generated a host list, and callout was configured, Exim
6730 crashed, because it could not find a port number from the (non-existent)
6731 transport. It now assumes port 25 in this circumstance.
6733 21. Added the -t option to exigrep.
6735 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6736 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6737 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6738 them is set, LOOKUP_LSEARCH is forced.
6740 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6742 24. Added sender and host information to the "rejected by local_scan()" log
6743 line; previously there was no indication of these.
6745 25. Added .include_if_exists.
6747 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6748 Linux. It turns out that not all file systems support this. Apparently some
6749 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6750 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6751 not supported on the file descriptor, is now ignored when Exim is trying to
6752 sync a directory. This applies only to Linux.
6754 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6756 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6757 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6760 29. Added support for maildirsize files from supplied patch (modified a bit).
6762 30. The use of :fail: followed by an empty string could lead Exim to respond to
6763 sender verification failures with (e.g.):
6765 550 Verification failed for <xxx>
6766 550 Sender verify failed
6768 where the first response line was missing the '-' that indicates it is not
6769 the final line of the response.
6771 31. The loop for finding the name of the user that called Exim had a hardwired
6772 limit of 10; it now uses the value of finduser_retries, which is used for
6773 all other user lookups.
6775 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6778 33. Exim was neglecting to zero errno before one call of strtol() when
6779 expanding a string and expecting an integer value. On some systems this
6780 resulted in spurious "integer overflow" errors. Also, it was casting the
6781 result into an int without checking.
6783 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6784 did not work. The code looks as if it has *never* worked, though it appears
6785 to have been documented since at least release 1.62. I have made it work.
6787 35. The "timeout_DNS" error in retry rules, also documented since at least
6788 1.62, also never worked. As it isn't clear exactly what this means, and
6789 clearly it isn't a major issue, I have abolished the feature by treating it
6790 as "timeout", and writing a warning to the main and panic logs.
6792 36. The display of retry rules for -brt wasn't always showing the error code
6795 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6796 timeout_connect_A, timeout_connect_MX.
6798 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6799 to the empty sender.
6801 39. The daemon was not analysing the content of -oX till after it had closed
6802 stderr and disconnected from the controlling terminal. This meant that any
6803 syntax errors were only noted on the panic log, and the return code from
6804 the command was 0. By re-arranging the code a little, I've made the
6805 decoding happen first, so such errors now appear on stderr, and the return
6806 code is 1. However, the actual setting up of the sockets still happens in
6807 the disconnected process, so errors there are still only recorded on the
6810 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6811 connections (as happens on some IP stacks) was logged at start up time as
6812 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6813 it from "IPv6 and IPv4", which means that two separate sockets are being
6816 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6817 says whether AF_INET or AF_INET6 was passed as an argument.
6819 42. Exiwhat output was messed up when time zones were included in log
6822 43. Exiwhat now gives more information about the daemon's listening ports,
6823 and whether -tls-on-connect was used.
6825 44. The "port" option of the smtp transport is now expanded.
6827 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6828 silently ignored. Now an error message is written to the main and panic
6831 46. There's a new ACL modifier called "logwrite" which writes to a log file
6832 as soon as it is encountered.
6834 47. Added $local_user_uid and $local_user_gid at routing time.
6836 48. Exim crashed when trying to verify a sender address that was being
6839 49. Exim was recognizing only a space character after ".include". It now also
6840 recognizes a tab character.
6842 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6843 extracting the relevant information from the specification. The man page no
6844 longer contains scrambled data for the -d option, and I've added a section
6845 at the front about calling Exim under different names.
6847 51. Added "extra_headers" argument to the "mail" command in filter files.
6849 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6852 53. Installed eximstats 1.29.
6854 54. Added transport_filter_timeout as a generic transport option.
6856 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6857 Cc: header lines. This was required by RFC 822, but it not required by RFC
6860 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6861 incoming messages that did not have them. Now it does so only if the
6862 message originates locally, that is, if there is no associated remote host
6863 address. When Resent- header lines are present, this applies to the Resent-
6864 lines rather than the non-Resent- lines.
6866 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6867 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6869 58. Messages for configuration errors now include the name of the main
6870 configuration file - useful now that there may be more than one file in a
6871 list (.included file names were always shown).
6873 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6874 for those rare installations that do not start the daemon as root or run it
6875 setuid root. I've cut out the call to initgroups() if the daemon is not
6878 60. The Exim user and group can now be bound into the binary as text strings
6879 that are looked up at the start of Exim's processing.
6881 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6883 62. Added $mailstore_basename variable.
6885 63. Installed patch to sieve.c from Michael Haardt.
6887 64. When Exim failed to open the panic log after failing to open the main log,
6888 the original message it was trying to log was written to stderr and debug
6889 output, but if they were not available (the usual case in production), it
6890 was lost. Now it is written to syslog before the two lines that record the
6891 failures to open the logs.
6893 65. Users' Exim filters run in subprocesses under the user's uid. It is
6894 possible for a "deliver" command or an alias in a "personal" command to
6895 provoke an address rewrite. If logging of address rewriting is configured,
6896 this fails because the process is not running as root or exim. There may be
6897 a better way of dealing with this, but for the moment (because 4.30 needs
6898 to be released), I have disabled address rewrite logging when running a
6899 filter in a non-root, non-exim process.
6905 1. The buildconfig auxiliary program wasn't quoting the value set for
6906 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6907 not defined. This bug was masked in 4.22 by the effect that was fixed in
6910 2. Some messages that were rejected after a message id was allocated were
6911 shown as "incomplete" by exigrep. It no longer does this for messages that
6912 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6914 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6915 have allow_domain_literals set, the ID did not get logged in the <= line.
6916 Domain literals are now always recognized in Message-ID: header lines.
6918 4. The first argument for a ${extract expansion item is the key name or field
6919 number. Leading and trailing spaces in this item were not being ignored,
6920 causing some misleading effects.
6922 5. When deliver_drop_privilege was set, single queue runner processes started
6923 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6924 same command in the process it spins off) were not dropping privilege.
6926 6. When the daemon running as "exim" started a queue runner, it always
6927 re-executed Exim in the spun-off process. This is a waste of effort when
6928 deliver_drop_privilege is set. The new process now just calls the
6929 queue-runner function directly.
6935 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6938 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6939 ignored. Though the use of -oP was forcing the writing of a pid file, it
6940 was always written to the default place.
6942 3. If the message "no IP address found for host xxxx" is generated during
6943 incoming verification, it is now followed by identification of the incoming
6944 connection (so you can more easily find what provoked it).
6946 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6948 5. Added some features to "harden" Exim a bit more against certain attacks:
6950 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6951 be put in Local/Makefile. This is like the never_users runtime option,
6952 but it cannot be overridden. The default setting is "root".
6954 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6955 prefix string with which any file named in a -C command line option
6958 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6959 is retained for -C and -D only if the caller of Exim is root. Without
6960 it, the exim user may also use -C and -D and retain privilege.
6962 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6963 command line option is disabled.
6965 6. Macro names set by the -D option must start with an upper case letter, just
6966 like macro names defined in the configuration file.
6968 7. Added "dereference=" facility to LDAP.
6970 8. Two instances of the typo "uknown" in the source files are fixed.
6972 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6973 the Configure-Makefile script screwed up while processing it.
6975 10. Incorporated PCRE 4.4.
6977 11. The SMTP synchronization check was not operating right at the start of an
6978 SMTP session. For example, it could not catch a HELO sent before the client
6979 waited for the greeting. There is now a check for outstanding input at the
6980 point when the greeting is written. Because of the duplex, asynchronous
6981 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6982 way, but not yet received, when the check is performed.
6984 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6985 on TCP/IP sockets, because this apparently causes some broken clients to
6988 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6989 unchanged) from the Cygwin maintainer.
6991 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6992 was supported instead of testing for mbx. Effectively a typo.
6994 15. The spa authenticator server code was not checking that the input it
6995 received was valid base64.
6997 16. The debug output line for the "set" modifier in ACLs was not showing the
6998 name of the variable that was being set.
7000 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7002 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7003 lookup on its own. Something else was needed to trigger the lookup. For
7004 example, a match in host_lookup or the need for a host name in a host list.
7005 Now, if $sender_host_name is referenced and the host name has not yet been
7006 looked up, a lookup is performed. If the lookup fails, the variable remains
7007 empty, and $host_lookup_failed is set to "1".
7009 19. Added "eqi" as a case-independent comparison operator.
7011 20. The saslauthd authentication condition could segfault if neither service
7012 nor realm was specified.
7014 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7015 error message that was logged was misleading, and incoming SMTP
7016 connections were dropped. The message is now more accurate, and temporary
7017 errors are given to SMTP connections.
7019 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7020 (except RSET) with a 421 error, until QUIT is received. However, it was
7021 failing to send a response to QUIT.
7023 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7024 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7025 ACL is now run after the helo_try_verify_hosts code.
7027 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7028 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7029 case-independent, so other case variants are also recognized). Apparently
7030 some systems use these upper case variants.
7032 25. If more than two messages were waiting for the same host, and a transport
7033 filter was specified for the transport, Exim sent two messages over the
7034 same TCP/IP connection, and then failed with "socket operation on non-
7035 socket" when it tried to send the third.
7037 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7039 27. The extern definition of crypt16() in expand.c was not being excluded when
7040 the OS had its own crypt16() function.
7042 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7043 as a preferred synonym for return_size_limit, both as an option and as an
7046 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7048 30. Changed the default configuration ACL to relax the local part checking rule
7049 for addresses that are not in any local domains. For these addresses,
7050 slashes and pipe symbols are allowed within local parts, but the sequence
7051 /../ is explicitly forbidden.
7053 31. SPA server authentication was not clearing the challenge buffer before
7056 32. log_message in a "warn" ACL statement was writing to the reject log as
7057 well as to the main log, which contradicts the documentation and doesn't
7058 seem right (because no rejection is happening). So I have stopped it.
7060 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7061 However, I am unable to do any testing of this.
7063 34. Fixed an infelicity in the appendfile transport. When checking directories
7064 for a mailbox, to see if any needed to be created, it was accidentally
7065 using path names with one or more superfluous leading slashes; tracing
7066 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7068 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7069 contents are added to the log line that is written for every discarded
7070 recipient. (Previously a log_message setting was ignored.)
7072 36. The ${quote: operator now quotes the string if it is empty.
7074 37. The install script runs exim in order to find its version number. If for
7075 some reason other than non-existence or emptiness, which it checks, it
7076 could not run './exim', it was installing it with an empty version number,
7077 i.e. as "exim-". This error state is now caught, and the installation is
7080 38. An argument was missing from the function that creates an error message
7081 when Exim fails to connect to the socket for saslauthd authentication.
7082 This could cause Exim to crash, or give a corrupted message.
7084 39. Added isip, isip4, and isip6 to ${if conditions.
7086 40. The ACL variables $acl_xx are now saved with the message, and can be
7087 accessed later in routers, transports, and filters.
7089 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7090 strings in the file are not string-expanded.
7092 42. If a MAIL command specified a SIZE value that was too large to fit into an
7093 int variable, the check against message_size_limit failed. Such values are
7094 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7095 day this will have to be increased, but I don't think I want to be around
7096 when emails are that large.
7103 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7104 iconv() is not standard in FreeBSD.
7106 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7107 IPv6 enabled. The observed symptom was a segmentation fault on return from
7108 the function os_common_find_running_interfaces() in src/os.c.
7110 3. In the check_special_case() function in daemon.c I had used "errno" as an
7111 argument name, which causes warnings on some systems. This was basically a
7112 typo, since it was named "eno" in the comments!
7114 4. The code that waits for the clock to tick (at a resolution of some fraction
7115 of a second) so as to ensure message-id uniqueness was always waiting for
7116 at least one whole tick, when it could have waited for less. [This is
7117 almost certainly not relevant at current processor speeds, where it is
7118 unlikely to ever wait at all. But we try to future-proof.]
7120 5. The function that sleeps for a time interval that includes fractions of a
7121 second contained a race. It did not block SIGALRM between setting the
7122 timer, and suspending (a couple of lines later). If the interval was short
7123 and the sigsuspend() was delayed until after it had expired, the suspension
7124 never ended. On busy systems this could lead to processes getting stuck for
7127 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7128 process, before it forks any delivery processes. The open lookup caching
7129 mechanism meant that the open file or database connection was passed into
7130 the delivery process. The problem was that delivery processes always tidy
7131 up cached lookup data. This could cause a problem for the next delivery
7132 process started by the queue runner, because the external queue runner
7133 process does not know about the closure. So the next delivery process
7134 still has data in the lookup cache. In the case of a file lookup, there was
7135 no problem because closing a file descriptor in a subprocess doesn't affect
7136 the parent. However, if the lookup was caching a connection to a database,
7137 the connection was closed, and the second delivery process was likely to
7138 see errors such as "PGSQL: query failed: server closed the connection
7139 unexpectedly". The problem has been fixed by closing all cached lookups
7140 in a queue runner before running a delivery process.
7142 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7143 seem to have the "const" qualifier which it has on other OS. I've
7146 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7147 *of the same type* (client or server) with the same public name that an
7148 error should be diagnosed.
7150 9. When Exim looked up a host name for an IP address, but failed to find the
7151 original IP address when looking up the host name (a safety check), it
7152 output the message "<ip address> does not match any IP for NULL", which was
7153 confusing, to say the least. The bug was that the host name should have
7154 appeared instead of "NULL".
7156 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7157 user that is built into the binary, and the -C or -D options is used, root
7158 privilege is dropped before the configuration file is read. In addition,
7159 logging is switched to stderr instead of the normal log files. If the
7160 configuration then re-defines the Exim user, the unprivileged environment
7161 is probably not what is expected, so Exim logs a panic warning message (but
7164 However, if deliver_drop_privilege is set, the unprivileged state may well
7165 be exactly what is intended, so the warning has been cut out in that case,
7166 and Exim is allowed to try to write to its normal log files.
7172 1. smtp_return_error_details was not giving details for temporary sender
7173 or receiver verification errors.
7175 2. Diagnose a configuration error if two authenticators have the same public
7178 3. Exim used not to create the message log file for a message until the first
7179 delivery attempt. This could be confusing when incoming messages were held
7180 for policy or load reasons. The message log file is now created at the time
7181 the message is received, and an initial "Received" line is written to it.
7183 4. The automatically generated man page for command line options had a minor
7184 bug that caused no ill effects; however, a more serious problem was that
7185 the procedure for building the man page automatically didn't always
7186 operate. Consequently, release 4.20 contains an out-of-date version. This
7187 shouldn't happen again.
7189 5. When building Exim with embedded Perl support, the script that builds the
7190 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7191 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7193 6. The freeze_tell option was not being used for messages that were frozen on
7194 arrival, either by an ACL or by local_scan().
7196 7. Added the smtp_incomplete_transaction log selector.
7198 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7199 was accepting AUTH without a new EHLO.
7201 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7202 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7205 10. Logging of TCP/IP connections (when configured) now happens in the main
7206 daemon process instead of the child process, so that the TCP/IP connection
7207 count is more accurate (but it can never be perfect).
7209 11. The use of "drop" in a nested ACL was not being handled correctly in the
7210 outer ACL. Now, if condition failure induced by the nested "drop" causes
7211 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7212 or "require"), the connection is dropped.
7214 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7215 that yield "discard" can now be used with an "accept" or a "discard" verb,
7216 but an error is generated for any others (because I can't see a useful way
7217 to define what should happen).
7219 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7220 are now processed in the same way as lines in the Exim configuration file.
7221 In particular, continuation lines are supported.
7223 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7225 15. Added -ti meaning -t -i.
7227 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7228 domains, and warn by logging if others are found.
7230 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7231 returned from GIFCONF when Exim is trying to find the list of interfaces on
7232 a host. The code in os.c has been modified to copy each ifreq to an aligned
7233 structure in all cases.
7235 Also, in some cases, the returned ifreq's were being copied to a 'struct
7236 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7237 means the last couple of bytes of an IPv6 address could be chopped if the
7238 ifreq contained only a normal sockaddr (14 bytes storage).
7240 18. Named domain lists were not supported in the hosts_treat_as_local option.
7241 An entry such as +xxxx was not recognized, and was treated as a literal
7244 19. Ensure that header lines added by a DATA ACL are included in the reject log
7245 if the ACL subsequently rejects the message.
7247 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7248 MD5 (which is deprecated).
7250 21. When testing a filter file using -bf, Exim was writing a message when it
7251 took the sender from a "From " line in the message, but it was not doing so
7252 when it took $return_path from a Return-Path: header line. It now does.
7254 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7255 with a valid header line field name (a series of printing characters
7256 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7258 23. Changed "disc" in the source to "disk" to conform to the documentation and
7259 the book and for uniformity.
7261 24. Ignore Sendmail's -Ooption=value command line item.
7263 25. When execve() failed while trying to run a command in a pipe transport,
7264 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7265 could be confused with a return value of 69 from the command itself. This
7266 has been changed to 127, the value the shell returns if it is asked to run
7267 a non-existent command. The wording for the related log line suggests a
7268 non-existent command as the problem.
7270 26. If received_header_text expands to an empty string, do not add a Received:
7271 header line to the message. (Well, it adds a token one on the spool, but
7272 marks it "old" so that it doesn't get used or transmitted.)
7274 27. Installed eximstats 1.28 (addition of -nt option).
7276 28. There was no check for failure on the call to getsockname() in the daemon
7277 code. This can fail if there is a shortage of resources on the system, with
7278 ENOMEM, for example. A temporary error is now given on failure.
7280 29. Contrary to the C standard, it seems that in some environments, the
7281 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7282 program. Exim now does this explicitly; it affects the formatting of
7283 timestamps using strftime().
7285 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7286 complaints. I've now initialized all the variables, to avoid this.
7288 32. Header lines added by a system filter were not being "seen" during
7289 transport-time rewrites.
7291 33. The info_callback() function passed to OpenSSL is set up with type void
7292 (*)(SSL *, int, int), as described somewhere. However, when calling the
7293 function (actually a macro) that sets it up, the type void(*)() is
7294 expected. I've put in a cast to prevent warnings from picky compilers.
7296 34. If a DNS black list lookup found a CNAME record, but there were no A
7297 records associated with the domain it pointed at, Exim crashed.
7299 35. If a DNS black list lookup returned more than one A record, Exim ignored
7300 all but the first. It now scans all returned addresses if a particular IP
7301 value is being sought. In this situation, the contents of the
7302 $dnslist_value variable are a list of all the addresses, separated by a
7305 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7306 to accept a host name and all its aliases if the forward lookup for any of
7307 them yielded the IP address of the incoming connection. Now it accepts only
7308 those names whose forward lookup yields the correct IP address. Any other
7309 names are discarded. This closes a loophole whereby a rogue DNS
7310 administrator could create reverse DNS records to break through a
7311 wildcarded host restriction in an ACL.
7313 37. If a user filter or a system filter that ran in a subprocess used any of
7314 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7315 the wrong values were passed to the pipe command ($thisaddress had the
7316 value of $0, $0 had the value of $1, etc). This bug was introduced by
7317 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7319 38. Improved the line breaking for long SMTP error messages from ACLs.
7320 Previously, if there was no break point between 40 and 75 characters, Exim
7321 left the rest of the message alone. Two changes have been made: (a) I've
7322 reduced the minimum length to 35 characters; (b) if it can't find a break
7323 point between 35 and 75 characters, it looks ahead and uses the first one
7324 that it finds. This may give the occasional overlong line, but at least the
7325 remaining text gets split now.
7327 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7328 file descriptors might be low, and that setting 1000 would always raise it.
7329 It turns out that in some environments, the limit is already over 1000 and
7330 that lowering it causes trouble. So now Exim takes care not to decrease it.
7332 40. When delivering a message, the value of $return_path is set to $sender_
7333 address at the start of routing (routers may change the value). By an
7334 oversight, this default was not being set up when an address was tested by
7335 -bt or -bv, which affected the outcome if any router or filter referred to
7338 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7339 log only when "message" or "log_message" are set. However, if one of the
7340 conditions was an address verification, or a call to a nested ACL, the
7341 messages generated by the underlying test were being passed through. This
7342 no longer happens. The underlying message is available in $acl_verify_
7343 message for both "message" and "log_message" expansions, so it can be
7344 passed through if needed.
7346 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7347 new expansion $bh_ to give the encoded byte string without charset
7348 translation. Translation happens only if iconv() is available; HAVE_ICONV
7349 indicates this at build time. HEADERS_CHARSET gives the charset to
7350 translate to; headers_charset can change it in the configuration, and
7351 "headers charset" can change it in an individual filter file.
7353 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7354 that creates RFC 2047 encoded "words" labels them as that charset instead
7355 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7356 expansion operator; (ii) when Exim creates a From: line for a local
7357 message; (iii) when a header line is rewritten to include a "phrase" part.
7359 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7360 buggy, causing it to skip the first lines of messages whose message ID
7361 ended in 'D'. This would not have bitten before Exim release 4.14, because
7362 message IDs were unlikely to end in 'D' before then. The effect was to have
7363 incorrect size information for certain domains.
7365 45. #include "config.h" was missing at the start of the crypt16.c module. This
7366 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7369 46. If there was a timeout during a "random" callout check, Exim treated it as
7370 a failure of the random address, and carried on sending RSET and the real
7371 address. If the delay was just some slowness somewhere, the response to the
7372 original RCPT would be taken as a response to RSET and so on, causing
7373 mayhem of various kinds.
7375 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7376 when I implemented it. It didn't allow for the fact that some option values
7377 may legitimately be negative (e.g. size_addition), and it didn't even do
7378 the right test for positive values.
7380 48. Domain names in DNS records are case-independent. Exim always looks them up
7381 in lower case. Some resolvers return domain names in exactly the case they
7382 appear in the zone file, that is, they may contain uppercase letters. Not
7383 all resolvers do this - some return always lower case. Exim was treating a
7384 change of case by a resolver as a change of domain, similar to a widening
7385 of a domain abbreviation. This triggered its re-routing code and so it was
7386 trying to route what was effectively the same domain again. This normally
7387 caused routing to fail (because the router wouldn't handle the domain
7388 twice). Now Exim checks for this case specially, and just changes the
7389 casing of the domain that it ultimately uses when it transmits the message
7392 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7395 50. If a filter generated a file delivery with a non-absolute name (possible if
7396 no home directory exists for the router), the forbid_file option was not
7399 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7400 the existing equality matching.
7402 52. Exim was using ints instead of ino_t variables in some places where it was
7403 dealing with inode numbers.
7405 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7406 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7407 called TMPDIR, and if it finds it is different, it changes its value.
7409 54. The smtp_printf() function is now made available to local_scan() so
7410 additional output lines can be written before returning. There is also an
7411 smtp_fflush() function to enable the detection of a dropped connection.
7412 The variables smtp_input and smtp_batched_input are exported to
7415 55. Changed the default runtime configuration: the message "Unknown user"
7416 has been removed from the ACL, and instead placed on the localuser router,
7417 using the cannot_route_message feature. This means that any verification
7418 failures that generate their own messages won't get overridden. Similarly,
7419 the "Unrouteable address" message that was in the ACL for unverifiable
7420 relay addresses has also been removed.
7422 56. Added hosts_avoid_esmtp to the smtp transport.
7424 57. The exicyclog script was not checking for the esoteric option
7425 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7426 will work only if exicyclog is run under the appropriate euid.
7428 58. Following a discussion on the list, the rules by which Exim recognises line
7429 endings on incoming messages have been changed. The -dropcr and drop_cr
7430 options are now no-ops, retained only for backwards compatibility. The
7431 following line terminators are recognized: LF CRLF CR. However, special
7432 processing applies to CR:
7434 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7435 nor a local message in the state where . is a terminator.
7437 (ii) If a bare CR is encountered in a header line, an extra space is added
7438 after the line terminator so as not to end the header. The reasoning
7439 behind this is that bare CRs in header lines are most likely either
7440 to be mistakes, or people trying to play silly games.
7442 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7443 was being incorrectly given as 18 bytes larger than it should have been.
7444 This is a VOB (very old bug).
7446 60. This may never have affected anything current, but just in case it has:
7447 When the local host is found other than at the start of a list of hosts,
7448 the local host, those with the same MX, and any that follow, are discarded.
7449 When the list in question was part of a longer list of hosts, the following
7450 hosts (not currently being processed) were also being discarded. This no
7451 longer happens. I'm not sure if this situation could ever has previously
7454 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7457 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7458 header. This is something that is recommended in a new Internet Draft, and
7459 is something that is documented as being done by Sendmail. There are two
7460 possible values. For messages generated by the autoreply transport, Exim
7463 Auto-Submitted: auto-replied
7465 whereas for all other generated messages (e.g. bounces) it adds
7467 Auto-Submitted: auto-generated
7469 63. The "personal" condition in filters now includes a test for the
7470 Auto-Submitted: header. If it contains the string "auto-" the message it
7471 not considered personal.
7473 64. Added rcpt_include_affixes as a generic transport option.
7475 65. Added queue_only_override (default true).
7477 66. Added the syslog_duplication option.
7479 67. If what should have been the first header line of a message consisted of
7480 a space followed by a colon, Exim was mis-interpreting it as a header line.
7481 It isn't of course - it is syntactically invalid and should therefore be
7482 treated as the start of the message body. The misbehaviour could have
7483 caused a number of strange effects, including loss of data in subsequent
7484 header lines, and spool format errors.
7486 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7487 client host had authenticated. This control can now be exercised by an ACL
7488 for more flexibility.
7490 69. By default, callouts do not happen when testing with -bh. There is now a
7491 variant, -bhc, which does actually run the callout code, including
7492 consulting and updating the callout cache.
7494 70. Added support for saslauthd authentication, courtesy of Alexander
7497 71. If statvfs() failed on the spool or log directories while checking their
7498 size for availability, Exim confusingly gave the error "space shortage".
7499 Furthermore, in debugging mode it crashed with a floating point exception.
7500 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7501 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7502 serious problem, Exim now writes to the main and panic logs when this
7503 happens, with details of the failure. It then refuses to accept the
7504 incoming message, giving the message "spool directory problem" or "log
7505 directory problem" with a 421 code for SMTP messages.
7507 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7508 0, 1, and 2 exist, because some OS complain for execs without them (see
7509 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7510 descriptors. However, the code omitted to check that the open succeeded,
7511 causing mysterious errors if for some reason the permissions on /dev/null
7512 got screwed. Now Exim writes a message to the main and panic logs, and
7513 bombs out if it can't open /dev/null.
7515 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7516 interact so that it is all more flexible. It is supposed to remain
7517 backwards compatible. Also added extra_local_interfaces.
7519 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7520 to bomb out with an assertion failure - to the client this appears as a
7521 connection drop. This problem occurs in the part of the code that was taken
7522 from the Samba project. Fortunately, the assertion is in a very simple
7523 function, so I have fixed this by reproducing the function inline in the
7524 one place where it is called, and arranging for authentication to fail
7525 instead of killing the process with assert().
7527 75. The SPA client code was not working when the server requested OEM rather
7528 than Unicode encoding.
7530 76. Added code to make require_files with a specific uid setting more usable in
7531 the case where statting the file as root fails - usually a non-root-mounted
7532 NFS file system. When this happens and the failure is EACCES, Exim now
7533 forks a subprocess and does the per-uid checking as the relevant uid.
7535 77. Added process_log_path.
7537 78. If log_file_path was not explicitly set, a setting of check_log_space or
7538 check_log_inodes was ignored.
7540 79. If a space check for the spool or log partitions fails, the incident is now
7541 logged. Of course, in the latter case the data may get lost...
7543 80. Added the %p formatting code to string_format() so that it can be used to
7544 print addresses in debug_print(). Adjusted all the address printing in the
7545 debugging in store.c to use %p rather than %d.
7547 81. There was a concern that a line of code in smtp_in.c could overflow a
7548 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7549 initially expressed, the concern was not well-founded, because trailing
7550 spaces are removed early. However, if the trailing spaces were followed by
7551 a NULL, they did not get removed, so the overflow was possible. Two fixes
7554 (a) I re-wrote the offending code in a cleaner fashion.
7555 (b) If an incoming SMTP command contains a NULL character, it is rejected
7558 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7559 runs initgroups(), so that if the Exim user is in any additional groups,
7560 they will be used during message reception.
7566 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob