1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
20 HS/01 Fix handling of very log lines in -H files. If a -<key> <value>
21 line caused the extension of big_buffer, the following lines where
24 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
26 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
28 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
29 buffer overrun for (non-chunking) other transports.
31 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
32 TLS1.3, means that a server rejecting a client certificate is not visible
33 to the client until the first read of encrypted data (typically the
34 response to EHLO). Add detection for that case and treat it as a failed
35 TLS connection attempt, so that the normal retry-in-clear can work (if
38 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
39 and/or domain. Found and fixed by Jason Betts.
41 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
42 configuration). If a CNAME target was not a wellformed name pattern, a
45 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
46 the OS reports them interleaved with other addresses.
48 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
49 used both for input and for a verify callout, both encrypted, SMTP
50 responses being sent by the server could be lost. This resulted in
51 dropped connections and sometimes bounces generated by a peer sending
54 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
55 string. Previously it was possible to cause undefined behaviour in a
56 library routine (usually a crash). Found by "zerons".
58 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
61 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
62 API was removed, so update to use the newer ones.
64 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
65 any timeout set, is taking a long time. Previously we would hang on to a
66 rotated logfile "forever" if the input was arriving with long gaps
67 (a previous attempt to fix addressed lack, for a long time, of initial
70 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
71 shared (NFS) environment. The length of the tempfile name is now
72 4 + 16 ("hdr.$message_exim_id") which might break on file
73 systems which restrict the file name length to lower values.
76 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
77 shared (NFS) environment.
79 HS/02 Bug 2392: exigrep does case sensitive *option* processing (as it
80 did for all versions <4.90). Notably -M, -m, --invert, -I may be
83 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
84 on some platforms for bit 31.
86 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
87 to changes apparently associated with TLS1.3 handling some of the APIs
88 previously used were either nonfunctional or inappropriate. Strings
89 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
90 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
91 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
92 This affects log line X= elements, the $tls_{in,out}_cipher variables,
93 and the use of specific cipher names in the encrypted= ACL condition.
95 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
97 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
98 verification result was not updated unless hosts_require_ocsp applied.
100 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
101 queue_list_requires_admin set to false, non-admin users were denied the
104 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
105 directory-of-certs mode. Previously they were advertised despite the
108 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
109 A single TCP connection by a client will now hold a TLS connection open
110 for multiple message deliveries, by default. Previoud the default was to
113 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
114 default. If built with the facility, DANE will be used. The facility
115 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
117 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
118 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
119 must be defined and you must still, unless you define DISABLE_TLS, manage
120 the the include-dir and library-file requirements that go with that
121 choice. Non-TLS builds are still supported.
123 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
126 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
127 default. If the platform supports and has the facility enabled, it will
128 be requested on all coneections.
130 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
131 controlled by the build-time option SUPPORT_PIPE_CONNECT.
133 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
135 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
136 success-DSN messages. Previously the From: header was always the default
137 one for these; the option was ignored.
139 JH/28 Fix the timeout on smtp response to apply to the whole response.
140 Previously it was reset for every read, so a teergrubing peer sending
141 single bytes within the time limit could extend the connection for a
142 long time. Credit to Qualsys Security Advisory Team for the discovery.
144 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
145 delivery address, which leaked information of the results of local
146 forwarding. Change to the original envelope recipient address, per
149 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
150 requested. Previously not bounce was generated and a log entry of
151 error ignored was made.
153 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
155 JH/32 Introduce a general tainting mechanism for values read from the input
156 channel, and values derived from them. Refuse to expand any tainted
157 values, to catch one form of exploit.
159 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
160 was unused and the unexpanded text used for the test. Found and
161 fixed by Ruben Jenster.
163 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
164 an attempt to use a TLS library read routine dereffed a nul pointer,
167 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
168 them in our smtp response.
170 JH/36 Have the general router option retry_use_local_part default to true when
171 any of the restrictive preconditions are set (to anything). Previously it
172 was only for check_local user. The change removes one item of manual
173 configuration which is required for proper retries when a remote router
174 handles a subset of addresses for a domain.
176 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
177 link count into consideration.
183 JH/01 Remove code calling the customisable local_scan function, unless a new
184 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
186 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
187 non-signal-safe functions being used.
189 JH/03 Bug 2269: When presented with a received message having a stupidly large
190 number of DKIM-Signature headers, disable DKIM verification to avoid
191 a resource-consumption attack. The limit is set at twenty.
193 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
194 report of oldest_pass in ${authres } in consequence, and separate out
195 some descriptions of reasons for verification fail.
197 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
198 files in the spool were present and unlocked. A queue-runner could spot
199 them, resulting in a duplicate delivery. Fix that by doing the unlock
200 after the unlink. Investigation by Tim Stewart. Take the opportunity to
201 add more error-checking on spoolfile handling while that code is being
204 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
205 No known attacks, no CVE, this is defensive hardening.
207 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
208 a queue-runner could start a delivery while other operations were ongoing.
209 Cutthrough delivery was a common victim, resulting in duplicate delivery.
210 Found and investigated by Tim Stewart. Fix by using the open message data
211 file handle rather than opening another, and not locally closing it (which
212 releases a lock) for that case, while creating the temporary .eml format
213 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
215 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
216 $sender_verify_failure/$recipient_verify_failure to "random".
218 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
221 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
222 Previously this would segfault.
224 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
227 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
228 like zero, since the resolver should be doing this for us, But we need one
229 as a CNAME but no MX presence gets the CNAME returned; we need to check
230 that doesn't point to an MX to declare it "no MX returned" rather than
231 "error, loop". A new main option is added so the older capability of
232 following some limited number of chain links is maintained.
234 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
236 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
237 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
238 now-deprecated earlier definitions used only specified the range up to TLS
239 1.2 (in the older-version library docs).
241 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
243 JH/15 Rework TLS client-side context management. Stop using a global, and
244 explicitly pass a context around. This enables future use of TLS for
245 connections to service-daemons (eg. malware scanning) while a client smtp
246 connection is using TLS; with cutthrough connections this is quite likely.
248 JH/16 Fix ARC verification to do AS checks in reverse order.
250 JH/17 Support a "tls" option on the ${readsocket } expansion item.
252 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
253 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
254 Previously the "utf8" would be re-prepended for every additional message.
256 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
257 Previously thery were accepted, resulting in issues when attempting to
258 forward messages to a non-supporting MTA.
260 PP/02 Let -n work with printing macros too, not just options.
262 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
263 one parent address was copied, and bogus data was used at delivery-logging
264 time. Either a crash (after delivery) or bogus log data could result.
265 Discovery and analysis by Tim Stewart.
267 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
268 Previously if the string ended mid-character, we did not insert the
269 promised '?' replacement.
271 PP/04 Documentation: current string operators work on bytes, not codepoints.
273 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
274 should pack well giving a smaller memory footprint so better caching and
275 therefore performance. Group the declarations where this can't be done so
276 that the byte-sized flag variables are not interspersed among pointer
277 variables, giving a better chance of good packing by the compiler.
279 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
280 non-null, to avoid issues with sites running BATV. Previously reports were
281 sent with an empty envelope sender so looked like bounces.
283 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
284 The ignore_error flag wasn't being returned from the filter subprocess so
285 was not set for later routers. Investigation and fix by Matthias Kurz.
287 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
288 and a msg:complete for the whole, when a message is manually removed using
289 -Mrm. Developement by Matthias Kurz, hacked on by JH.
291 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
292 a "Gnu special" function, asprintf() in the DB utility binary builds; I
293 hope that is portable enough.
295 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
296 requiring a known-CA anchor certificate; make it now rely entirely on the
297 TLSA as an anchor. Checking the name on the leaf cert against the name
298 on the A-record for the host is still done for TA (but not for EE mode).
300 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
301 included in delivery lines for non-proxied connections, when compiled with
302 SUPPORT_SOCKS and running with proxy logging enabled.
304 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
305 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
306 move the existing event to fire before the normal logging of message
307 failure so that custom logging is bracketed by normal logging.
309 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
310 msg:fail:internal event. Developement by Matthias Kurz.
312 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
313 far too small for todays use of crypto signatures stored there. Go all
314 the way to the max DNS message size of 64kB, even though this might be
315 overmuch for IOT constrained device use.
317 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
318 copy a string over itself. The library routine is documented as not
319 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
321 JH/32 For main options check_spool_space and check_inode_space, where the
322 platform supports 64b integers, support more than the previous 2^31 kB
323 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
324 the previous G, M, k.
326 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
327 $authenticated_fail_id variable on authentication failure. Previously
330 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
331 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
332 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
333 GNUTLS_SEC_PARAM_MEDIUM.
335 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
336 side. Previously we would continue as if no SNI had been received.
338 JH/36 Harden the handling of string-lists. When a list consisted of a sole
339 "<" character, which should be a list-separator specification, we walked
340 off past the nul-terimation.
342 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
343 causes) even when the retry time is not yet met. Previously they were
344 not, meaning that when (say) an account was over-quota and temp-rejecting,
345 and multiple senders' messages were queued, only one sender would get
346 notified on each configured delay_warning cycle.
348 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
350 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
351 cipher-suites, an error can be left on the stack even for a succeeding
352 accept; this results in impossible error messages when a later operation
355 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
356 return error codes indicating retry. Under TLS1.3 this becomes required.
358 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
359 it only wrote the new authenticators, resulting in a lack of tracking of
360 peer changes of ESMTP extensions until the next cache flush.
362 JH/41 Fix the loop reading a message header line to check for integer overflow,
363 and more-often against header_maxsize. Previously a crafted message could
364 induce a crash of the recive process; now the message is cleanly rejected.
366 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
367 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
373 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
374 When redis_servers is set to a list of > 1 element, and the Redis servers
375 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
376 case of MOVED into a DEFER case instead, thus moving the query onto the
377 next server in the list. For a cluster of N elements, all N servers must
378 be defined in redis_servers.
380 GF/02 Catch and remove uninitialized value warning in exiqsumm
381 Check for existence of @ARGV before looking at $ARGV[0]
383 JH/01 Replace the store_release() internal interface with store_newblock(),
384 which internalises the check required to safely use the old one, plus
385 the allocate and data copy operations duplicated in both (!) of the
386 extant use locations.
388 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
389 modifier. This matches the restriction on the commandline.
391 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
392 Previously only the last row was returned.
394 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
395 we assumed that tags in the header were well-formed, and parsed the
396 element content after inspecting only the first char of the tag.
397 Assumptions at that stage could crash the receive process on malformed
400 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
401 While running the DKIM ACL we operate on the Permanent memory pool so that
402 variables created with "set" persist to the DATA ACL. Also (at any time)
403 DNS lookups that fail create cache records using the Permanent pool. But
404 expansions release any allocations made on the current pool - so a dnsdb
405 lookup expansion done in the DKIM ACL releases the memory used for the
406 DNS negative-cache, and bad things result. Solution is to switch to the
407 Main pool for expansions.
408 While we're in that code, add checks on the DNS cache during store_reset,
409 active in the testsuite.
410 Problem spotted, and debugging aided, by Wolfgang Breyha.
412 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
413 When none of the hosts presented to a transport match an already-open
414 connection, close it and proceed with the list. Previously we would
415 queue the message. Spotted by Lena with Yahoo, probably involving
418 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
419 Previously a spurious "250 OK id=" response was appended to the proper
422 JH/08 The "support for" informational output now, which built with Content
423 Scanning support, has a line for the malware scanner interfaces compiled
424 in. Interface can be individually included or not at build time.
426 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
427 by the template makefile "src/EDITME". The "STREAM" support for an older
428 ClamAV interface method is removed.
430 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
431 rows affected is given instead).
433 JH/11 The runtime Berkeley DB library version is now additionally output by
434 "exim -d -bV". Previously only the compile-time version was shown.
436 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
437 SMTP connection. Previously, when one had more receipients than the
438 first, an abortive onward connection was made. Move to full support for
439 multiple onward connections in sequence, handling cutthrough connection
440 for all multi-message initiating connections.
442 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
443 routers. Previously, a multi-recipient message would fail to match the
444 onward-connection opened for the first recipient, and cause its closure.
446 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
447 a timeout on read on a GnuTLS initiating connection, resulting in the
448 initiating connection being dropped. This mattered most when the callout
449 was marked defer_ok. Fix to keep the two timeout-detection methods
452 JH/15 Relax results from ACL control request to enable cutthrough, in
453 unsupported situations, from error to silently (except under debug)
454 ignoring. This covers use with PRDR, frozen messages, queue-only and
457 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
459 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
460 metadata, resulting in a crash in free().
462 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
463 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
464 Broken also in d185889f4, with init system revamp.
466 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
467 Previously we abruptly closed the connection after reading a malware-
468 found indication; now we go on to read the "scan ok" response line,
471 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
472 ACL. Previously, a crash would result.
474 JH/19 Speed up macro lookups during configuration file read, by skipping non-
475 macro text after a replacement (previously it was only once per line) and
476 by skipping builtin macros when searching for an uppercase lead character.
478 JH/20 DANE support moved from Experimental to mainline. The Makefile control
479 for the build is renamed.
481 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
482 was allocated for every new TLS startup, meaning one per message. Fix
483 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
485 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
486 reported the original. Fix to report (as far as possible) the ACL
487 result replacing the original.
489 JH/23 Fix memory leak during multi-message connections using STARTTLS under
490 OpenSSL. Certificate information is loaded for every new TLS startup,
491 and the resources needed to be freed.
493 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
495 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
498 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
499 DATA response info to the (existing) per-recipient response info for
500 the "C=" log element. It can have useful tracking info from the
501 destination system. Patch from Simon Arlott.
503 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
504 length value. Previously this would segfault.
506 HS/02 Support Avast multiline protoocol, this allows passing flags to
507 newer versions of the scanner.
509 JH/28 Ensure that variables possibly set during message acceptance are marked
510 dead before release of memory in the daemon loop. This stops complaints
511 about them when the debug_store option is enabled. Discovered specifically
512 for sender_rate_period, but applies to a whole set of variables.
513 Do the same for the queue-runner and queue-list loops, for variables set
514 from spool message files. Do the same for the SMTP per-message loop, for
515 certain variables indirectly set in ACL operations.
517 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
518 as a multi-recipient message from a mailinglist manager). The coding had
519 an arbitrary cutoff number of characters while checking for more input;
520 enforced by writing a NUL into the buffer. This corrupted long / fast
521 input. The problem was exposed more widely when more pipelineing of SMTP
522 responses was introduced, and one Exim system was feeding another.
523 The symptom is log complaints of SMTP syntax error (NUL chars) on the
524 receiving system, and refused recipients seen by the sending system
525 (propating to people being dropped from mailing lists).
526 Discovered and pinpointed by David Carter.
528 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
529 replaced by the ${authresults } expansion.
531 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
533 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
534 allows proper process termination in container environments.
536 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
537 Previously the "final dot" had a newline after it; ensure it is CR,LF.
539 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
540 and "err_perm", deprecated since 4.83 when the RFC-defined words
541 "temperror" and "permerror" were introduced.
543 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
544 transport-filters or DKIM-signing. The restriction was lost in the
545 consolidation of verify-callout and delivery SMTP handling.
546 Extend the restriction to also cover ARC-signing.
548 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
549 in defer=pass mode supply a 450 to the initiator. Previously the message
552 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
553 tls_require_ciphers is used as before.
555 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
556 "pass_unscanned". Only tmpfails from the scanner are written to
557 the paniclog, as they may require admin intervention (permission
558 denied, license issues). Other scanner errors (like decompression
559 bombs) do not cause a paniclog entry.
561 JH/36 Fix reinitialisation of DKIM logging variable between messages.
562 Previously it was possible to log spurious information in receive log
565 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
566 triggered odd behaviour from Outlook Express clients.
568 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
571 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
572 since the IETF WG has not yet settled on that versus the original
573 "bare" representation.
575 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
576 Previously the millisecond value corrupted the output.
577 Fix also for syslog_pid=no and log_selector +pid, for which the pid
578 corrupted the output.
584 JH/01 Rework error string handling in TLS interface so that the caller in
585 more cases is responsible for logging. This permits library-sourced
586 string to be attached to addresses during delivery, and collapses
587 pairs of long lines into single ones.
589 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
590 during configuration. Wildcards are allowed and expanded.
592 JH/02 Rework error string handling in DKIM to pass more info back to callers.
593 This permits better logging.
595 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
596 do not close it down and have the child transport start it up again on
597 the passed-on TCP connection. Instead, proxy the child (and any
598 subsequent ones) for TLS via a unix-domain socket channel. Logging is
599 affected: the continued delivery log lines do not have any DNSSEC, TLS
600 Certificate or OCSP information. TLS cipher information is still logged.
602 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
603 identical IP addresses on different listening ports. Will also affect
606 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
607 add noisy ifdef guards to special-case this sillyness.
608 Patch from Bernd Kuhls.
610 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
611 than 255 are no longer allowed.
613 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
614 Disable the session-cache too, which might reduce our load. Since we
615 currrectly use a new context for every connection, both as server and
616 client, there is no benefit for these.
617 GnuTLS appears to not support tickets server-side by default (we don't
618 call gnutls_session_ticket_enable_server()) but client side is enabled
619 by default on recent versions (3.1.3 +) unless the PFS priority string
622 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
623 <https://reproducible-builds.org/specs/source-date-epoch/>.
625 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
626 the check for any unsuccessful recipients did not notice the limit, and
627 erroneously found still-pending ones.
629 JH/08 Pipeline CHUNKING command and data together, on kernels that support
630 MSG_MORE. Only in-clear (not on TLS connections).
632 JH/09 Avoid using a temporary file during transport using dkim. Unless a
633 transport-filter is involved we can buffer the headers in memory for
634 creating the signature, and read the spool data file once for the
635 signature and again for transmission.
637 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
638 4.77 as the kernel support then wasn't solid, having issues in 64bit
639 mode. Now, it's been long enough. Add support for FreeBSD also.
641 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
642 case where the routing stage had gathered several addresses to send to
643 a host before calling the transport for the first, we previously failed
644 to close down TLS in the old transport process before passing the TCP
645 connection to the new process. The new one sent a STARTTLS command
646 which naturally failed, giving a failed delivery and bloating the retry
647 database. Investigation and fix prototype from Wolfgang Breyha.
649 JH/12 Fix check on SMTP command input synchronisation. Previously there were
650 false-negatives in the check that the sender had not preempted a response
651 or prompt from Exim (running as a server), due to that code's lack of
652 awareness of the SMTP input buffering.
654 PP/04 Add commandline_checks_require_admin option.
655 Exim drops privileges sanely, various checks such as -be aren't a
656 security problem, as long as you trust local users with access to their
657 own account. When invoked by services which pass untrusted data to
658 Exim, this might be an issue. Set this option in main configuration
659 AND make fixes to the calling application, such as using `--` to stop
662 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
663 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
664 responses to those, into a single TLS record each way (this usually means
665 a single packet). As a side issue, smtp_enforce_sync now works on TLS
668 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
669 affects you only if you're dancing at the edge of the param size limits.
670 If you are, and this message makes sense to you, then: raise the
671 configured limit or use OpenSSL 1.1. Nothing we can do for older
674 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
675 cmdline element to get the documented default one. Previously it was
678 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
679 are unsuable for non-message-specific reasons - eg. connection timeout,
680 banner-time rejection.
682 JH/16 Fix logging of delivery remote port, when specified by router, under
685 PP/06 Repair manualroute's ability to take options in any order, even if one
686 is the name of a transport.
689 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
691 JH/17 Change the list-building routines interface to use the expanding-string
692 triplet model, for better allocation and copying behaviour.
694 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
695 Previously it was constructed the first time a possibly-matching string
696 was met in the configuration file input during startup; now it is done
699 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
700 compatible one, to avoid the (poorly documented) possibility of a config
701 file in the working directory redirecting the DB files, possibly correpting
702 some existing file. CVE-2017-10140 assigned for BDB.
704 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
705 cache-hot. Previously, although the result was properly cached, the
706 initial verify call returned a defer.
708 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
709 the main verify for receipient in uncached-mode.
711 JH/22 Retire historical build files to an "unsupported" subdir. These are
712 defined as "ones for which we have no current evidence of testing".
714 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
715 if present. Previously it was ignored.
717 JH/24 Start using specified-initialisers in C structure init coding. This is
718 a C99 feature (it's 2017, so now considered safe).
720 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
721 if was a fixed-sized field and bitmask ops via macros; it is now more
724 PP/07 GitHub PR 56: Apply MariaDB build fix.
725 Patch provided by Jaroslav Škarvada.
727 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
728 during Coverity cleanups [4.87 JH/47]
729 Diagnosis and fix provided by Michael Fischer v. Mollard.
731 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
732 the right size to place the terminating semicolon on its own folded
733 line, the header hash was calculated to an incorrect value thanks to
734 the (relaxed) space the fold became.
736 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
737 and confused the parent.
739 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
740 which could crash as a result. This could lead to undeliverable messages.
742 JH/28 Logging: "next input sent too soon" now shows where input was truncated
745 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
746 matters on fast-turnover and PID-randomising systems, which were getting
747 out-of-order delivery.
749 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
750 a possibly-overlapping copy. The symptom was that "Remote host closed
751 connection in response to HELO" was logged instead of the actual 4xx
754 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
755 Previously only that bufferd was discarded, resulting in SYMTP command
758 JH/32 DKIM: when a message has multiple signatures matching an identity given
759 in dkim_verify_signers, run the dkim acl once for each. Previously only
760 one run was done. Bug 2189.
762 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
763 "panic the current process" to "deliberately defer". The panic log is
764 still written with the problem list name; the mail and reject logs now
765 get a temp-reject line for the message that was being handled, saying
766 something like "domains check lookup or other defer". The SMTP 451
767 message is still "Temporary local problem".
769 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
770 A crafted sequence of BDAT commands could result in in-use memory beeing
771 freed. CVE-2017-16943.
773 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
774 from SMTP input. Previously it was always done; now only done for DATA
775 and not BDAT commands. CVE-2017-16944.
777 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
778 to the message (such as an overlong header line). Previously this was
779 not done and we did not exit BDAT mode. Followon from the previous item
780 though a different problem.
786 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
787 than -2003 did; needs libidn2 in addition to libidn.
789 JH/02 The path option on a pipe transport is now expanded before use.
791 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
792 Patch provided by "Björn", documentation fix added too.
794 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
795 missing a wire-to-host endian conversion.
797 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
798 close after a BDAT command line could be taken as a following command,
799 giving a synch failure. Fix by only checking for synch immediately
800 before acknowledging the chunk.
802 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
803 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
804 macro. Patches provided by Josh Soref.
806 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
807 Previously we did not; the RFC seems ambiguous and VRFY is not listed
808 by IANA as a service extension. However, John Klensin suggests that we
811 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
812 the dkim code may be unix-mode line endings rather than smtp wire-format
813 CRLF, so prepend a CR to any bare LF.
815 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
816 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
818 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
819 the first were themselves being wrongly included in the feed into dkim
820 processing; with most chunk sizes in use this resulted in an incorrect
821 body hash calculated value.
823 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
824 DKIM signature block, for verification. Although advised against by
825 standards it is specifically not ruled illegal.
827 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
829 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
830 missing a body hash (the bh= tag).
832 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
833 It seems that HAProxy sends the Proxy Protocol information in clear and
834 only then does a TLS startup, so do the same.
836 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
837 TCP connections (such as for Spamd) unless the daemon successfully set
838 Fast Open mode on its listening sockets. This fixes breakage seen on
839 too-old kernels or those not configured for Fast Open, at the cost of
840 requiring both directions being enabled for TFO, and TFO never being used
841 by non-daemon-related Exim processes.
843 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
844 endings, at least on the first header line. Try to canonify any that get
845 past that check, despite the cost.
847 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
848 now limited to an arbitrary five deep, while parsing addresses with the
849 strip_excess_angle_brackets option enabled.
851 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
852 instead leave the unprompted TLS handshake in socket buffer for the
853 TLS library to consume.
855 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
857 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
859 JH/16 Drop variables when they go out of scope. Memory management drops a whole
860 region in one operation, for speed, and this leaves assigned pointers
861 dangling. Add checks run only under the testsuite which checks all
862 variables at a store-reset and panics on a dangling pointer; add code
863 explicitly nulling out all the variables discovered. Fixes one known
864 bug: a transport crash, where a dangling pointer for $sending_ip_address
865 originally assigned in a verify callout, is re-used.
867 PP/06 Drop '.' from @INC in various Perl scripts.
869 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
871 PP/08 Reduce a number of compilation warnings under clang; building with
872 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
873 should be warning-free.
875 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
877 HS/01 Fix portability problems introduced by PP/08 for platforms where
878 realloc(NULL) is not equivalent to malloc() [SunOS et al].
880 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
881 chunk. This allows us to accept broken chunked messages. We need a more
882 general solution here.
884 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
885 already-broken messages in the queue.
887 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
889 JH/19 Fix reference counting bug in routing-generated-address tracking.
895 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
896 supports it and a size is available (ie. the sending peer gave us one).
898 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
899 years of being deprecated). The replacements are the ACLs
900 acl_smtp_mime and acl_not_smtp_mime.
902 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
903 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
904 or even an in-clear connection were permitted. Now, if the host lookup
905 was dnssec and dane was requested then the host is only used if the
906 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
907 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
908 if one fails this test.
909 This means that a poorly-configured remote DNS will make it incommunicado;
910 but it protects against a DNS-interception attack on it.
912 JH/04 Bug 1810: make continued-use of an open smtp transport connection
913 non-noisy when a race steals the message being considered.
915 JH/05 If main configuration option tls_certificate is unset, generate a
916 self-signed certificate for inbound TLS connections.
918 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
919 in rewrites and routers.
921 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
922 and logged a warning sing 4.83; now they are a configuration file error.
924 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
925 (lacking @domain). Apply the same qualification processing as RCPT.
927 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
929 JH/10 Support ${sha256:} applied to a string (as well as the previous
932 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
933 a cutthrough deliver is pending, as we always want to make a connection.
934 This also avoids re-routing the message when later placing the cutthrough
935 connection after a verify cache hit.
936 Do not update it with the verify result either.
938 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
939 when routing results in more than one destination address.
941 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
942 signing (which inhibits the cutthrough capability). Previously only
943 the presence of an option was tested; now an expansion evaluating as
944 empty is permissible (obviously it should depend only on data available
945 when the cutthrough connection is made).
947 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
948 the relevant preceding SMTP command did not note the pipelining mode.
950 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
951 Previously they were not counted.
953 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
954 as one having no matching records. Previously we deferred the message
955 that needed the lookup.
957 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
958 distinguished as "(=".
960 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
961 for missing MX records. Previously it only worked for missing A records.
963 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
965 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
966 after the data-go-ahead and data-ack. Patch from Jason Betts.
968 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
969 even for a "none" policy. Patch from Tony Meyer.
971 JH/22 Fix continued use of a connection for further deliveries. If a port was
972 specified by a router, it must also match for the delivery to be
975 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
976 When one of the recipients of a message was unsuitable for the connection
977 (has no matching addresses), we lost track of needing to mark it
978 deferred. As a result mail would be lost.
980 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
982 JH/25 Decoding ACL controls is now done using a binary search; the source code
983 takes up less space and should be simpler to maintain. Merge the ACL
984 condition decode tables also, with similar effect.
986 JH/26 Fix problem with one_time used on a redirect router which returned the
987 parent address unchanged. A retry would see the parent address marked as
988 delivered, so not attempt the (identical) child. As a result mail would
991 JH/27 Fix a possible security hole, wherein a process operating with the Exim
992 UID can gain a root shell. Credit to http://www.halfdog.net/ for
993 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
996 JH/28 Enable {spool,log} filesystem space and inode checks as default.
997 Main config options check_{log,spool}_{inodes,space} are now
998 100 inodes, 10MB unless set otherwise in the configuration.
1000 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1001 Previously it only applied to the main-section connection policy
1004 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1006 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1007 by me. Added RFC7919 DH primes as an alternative.
1009 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1010 are not in the system include path.
1012 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1013 GnuTLS, when a session startup failed (eg because the client disconnected)
1014 Exim did stdio operations after fclose. This was exposed by a recent
1015 change which nulled out the file handle after the fclose.
1017 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1018 signed directly by the cert-signing cert, rather than an intermediate
1019 OCSP-signing cert. This is the model used by LetsEncrypt.
1021 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1023 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1024 an incoming connection.
1026 HS/02 Bug 1802: Do not half-close the connection after sending a request
1029 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1030 fallback to "prime256v1".
1032 JH/34 SECURITY: Use proper copy of DATA command in error message.
1033 Could leak key material. Remotely exploitable. CVE-2016-9963.
1039 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1040 and 3.4.4 - once the server is enabled to respond to an OCSP request
1041 it does even when not requested, resulting in a stapling non-aware
1042 client dropping the TLS connection.
1044 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1045 support variable-length bit vectors. No functional change.
1047 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1048 The I= interface field on outgoing lines is now after the H= remote
1049 host field, same as incoming lines. There is a separate
1050 outgoing_interface log selector which allows you to disable the
1053 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1054 If not running log_selector +smtp_connection the mainlog would be held
1055 open indefinitely after a "too many connections" event, including to a
1056 deleted file after a log rotate. Leave the per net connection logging
1057 leaving it open for efficiency as that will be quickly detected by the
1058 check on the next write.
1060 HS/01 Bug 1671: Fix post transport crash.
1061 Processing the wait-<transport> messages could crash the delivery
1062 process if the message IDs didn't exist for some reason. When
1063 using 'split_spool_directory=yes' the construction of the spool
1064 file name failed already, exposing the same netto behaviour.
1066 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1067 mime_regex ACL conditions.
1069 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1070 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1071 to HELO, local diagnostic string.
1073 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1074 log line to debug. Even when configured with a tls authenticator many
1075 client connections are expected to not authenticate in this way, so
1076 an authenticate fail is not an error.
1078 HS/02 Add the Exim version string to the process info. This way exiwhat
1079 gives some more detail about the running daemon.
1081 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1082 matter for fast-change records such as DNSBLs.
1084 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1085 retry record, even if constant. There may be multiple transports with
1086 different interface settings and the retry behaviour needs to be kept
1089 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1091 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1093 JH/10 Bug 840: fix log_defer_output option of pipe transport
1095 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1096 hosts_randomize. This matters a lot when combined with mua_wrapper.
1098 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1099 ${quote_pgsql:<string>} operator.
1101 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1103 JH/14 Change header file name for freeradius-client. Relevant if compiling
1104 with Radius support; from the Gentoo tree and checked under Fedora.
1106 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1108 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1109 option transport_filter, meaning no filtering.
1111 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1113 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1114 defaults to "*" (all hosts). The variable is now available when not built
1115 with TLS, default unset, mainly to enable keeping the testsuite sane.
1116 If a server certificate is not supplied (via tls_certificate) an error is
1117 logged, and clients will find TLS connections fail on startup. Presumably
1118 they will retry in-clear.
1119 Packagers of Exim are strongly encouraged to create a server certificate
1120 at installation time.
1122 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1123 with the $config_file variable.
1125 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1126 in transport context, after the attempt, and per-recipient. The latter type
1127 is per host attempted. The event data is the error message, and the errno
1128 information encodes the lookup type (A vs. MX) used for the (first) host,
1129 and the trailing two digits of the smtp 4xx response.
1131 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1132 to write to mainlog (or rejectlog, paniclog) in the window between file
1133 creation and permissions/ownership being changed. Particularly affects
1134 installations where exicyclog is run as root, rather than exim user;
1135 result is that the running daemon panics and dies.
1137 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1139 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1140 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1141 "pri" and "weight". Note that the previous implicit priority given by the
1142 list order is no longer honoured.
1144 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1145 for DKIM processing.
1147 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1148 by defining SUPPORT_SOCKS.
1150 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1151 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1152 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1153 variables are renamed to proxy_{local,external}_{address,port}.
1155 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1156 for a build by defining SUPPORT_I18N
1158 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1159 of the query string, and make ${quote_redis:} do that quoting.
1161 JH/29 Move Events support from Experimental to mainline, enabled by default
1162 and removable for a build by defining DISABLE_EVENT.
1164 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1166 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1167 cached by the daemon.
1169 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1170 by defining LOOKUP_REDIS. The libhiredis library is required.
1172 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1173 keys are given for lookup.
1175 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1176 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1177 only supported when built with TLS support. The PolarSSL SHA routines
1178 are still used when the TLS library is too old for convenient support.
1180 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1181 openssl_options), for security. OpenSSL forces this from version 1.1.0
1182 server-side so match that on older versions.
1184 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1185 allocation for $value could be released as the expansion processing
1186 concluded, but leaving the global pointer active for it.
1188 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1189 and to use the domains and local_parts ACL conditions.
1191 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1192 incorrectly not doubled on cutthrough transmission, hence seen as a
1193 body-termination at the receiving system - resulting in truncated mails.
1194 Commonly the sender saw a TCP-level error, and retransmitted the message
1195 via the normal store-and-forward channel. This could result in duplicates
1196 received - but deduplicating mailstores were liable to retain only the
1197 initial truncated version.
1199 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1201 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1203 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1204 we're in there, support oversigning also; bug 1309.
1206 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1208 HS/04 Add support for keep_environment and add_environment options.
1210 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1211 either intentional arithmetic overflow during PRNG, or testing config-
1214 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1215 delivery resulted in actual delivery. Cancel cutthrough before DATA
1218 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1219 rejects a recipient: pass the reject to the originator.
1221 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1222 Many were false-positives and ignorable, but it's worth fixing the
1225 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1226 for the new environment-manipulation done at startup. Move the routines
1227 from being local to tls.c to being global via the os.c file.
1229 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1230 an extract embedded as result-arg for a map, the first arg for extract
1231 is unavailable so we cannot tell if this is a numbered or keyed
1232 extraction. Accept either.
1238 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1241 JH/02 The smtp transport option "multi_domain" is now expanded.
1243 JH/03 The smtp transport now requests PRDR by default, if the server offers
1246 JH/04 Certificate name checking on server certificates, when exim is a client,
1247 is now done by default. The transport option tls_verify_cert_hostnames
1248 can be used to disable this per-host. The build option
1249 EXPERIMENTAL_CERTNAMES is withdrawn.
1251 JH/05 The value of the tls_verify_certificates smtp transport and main options
1252 default to the word "system" to access the system default CA bundle.
1253 For GnuTLS, only version 3.0.20 or later.
1255 JH/06 Verification of the server certificate for a TLS connection is now tried
1256 (but not required) by default. The verification status is now logged by
1257 default, for both outbound TLS and client-certificate supplying inbound
1260 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1263 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1264 Status Notification (bounce) messages are now MIME format per RFC 3464.
1265 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1266 under the control of the dsn_advertise_hosts option, and routers may
1267 have a dsn_lasthop option.
1269 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1270 default, modifiable by a malware= option. The list separator for
1271 the options can now be changed in the usual way. Bug 68.
1273 JH/10 The smtp_receive_timeout main option is now expanded before use.
1275 JH/11 The incoming_interface log option now also enables logging of the
1276 local interface on delivery outgoing connections.
1278 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1279 if the interface and destination host and port all match.
1281 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1284 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1285 Patch from Andrew Lewis.
1287 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1288 now supports optional time-restrictions, weighting, and priority
1289 modifiers per server. Patch originally by <rommer@active.by>.
1291 JH/16 The spamd_address main option now supports a mixed list of local
1292 and remote servers. Remote servers can be IPv6 addresses, and
1293 specify a port-range.
1295 JH/17 Bug 68: The spamd_address main option now supports an optional
1296 timeout value per server.
1298 JH/18 Bug 1581: Router and transport options headers_add/remove can
1299 now have the list separator specified.
1301 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1304 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1307 JH/21 Support for the A6 type of dns record is withdrawn.
1309 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1310 rather than the verbs used.
1312 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1313 from 255 to 1024 chars.
1315 JH/24 Verification callouts now attempt to use TLS by default.
1317 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1318 are generic router options now. The defaults didn't change.
1320 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1321 Original patch from Alexander Shikoff, worked over by JH.
1323 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1324 files if ZCAT_COMMAND is not executable.
1326 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1328 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1330 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1331 Normally benign, it bites when the pair was led to by a CNAME;
1332 modern usage is to not canonicalize the domain to a CNAME target
1333 (and we were inconsistent anyway for A-only vs AAAA+A).
1335 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1337 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1338 when evaluating $sender_host_dnssec.
1340 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1341 $sender_helo_dnssec variable.
1343 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1345 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1347 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1349 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1350 documented as working, but never had. Support all but $spam_report.
1352 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1353 added for tls authenticator.
1355 HS/03 Add perl_taintmode main config option
1361 TL/01 When running the test suite, the README says that variables such as
1362 no_msglog_check are global and can be placed anywhere in a specific
1363 test's script, however it was observed that placement needed to be near
1364 the beginning for it to behave that way. Changed the runtest perl
1365 script to read through the entire script once to detect and set these
1366 variables, reset to the beginning of the script, and then run through
1367 the script parsing/test process like normal.
1369 TL/02 The BSD's have an arc4random API. One of the functions to induce
1370 adding randomness was arc4random_stir(), but it has been removed in
1371 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1372 function when detected.
1374 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1375 cause callback expansion.
1377 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1378 syntax errors in an expansion can be treated as a string instead of
1379 logging or causing an error, due to the internal use of bool_lax
1380 instead of bool when processing it.
1382 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1383 server certificates when making smtp deliveries.
1385 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1387 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1389 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1391 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1392 Merged patch from Sebastian Wiedenroth.
1394 JH/05 Fix results-pipe from transport process. Several recipients, combined
1395 with certificate use, exposed issues where response data items split
1396 over buffer boundaries were not parsed properly. This eventually
1397 resulted in duplicates being sent. This issue only became common enough
1398 to notice due to the introduction of connection certificate information,
1399 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1401 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1402 size buffer was used, resulting in syntax errors when an expansion
1405 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1406 version 3.3.6 or later.
1408 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1409 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1410 both become "event_action", the variables become $event_name, $event_data
1411 and $event_defer_errno. There is a new variable $verify_mode, usable in
1412 routers, transports and related events. The tls:cert event is now also
1413 raised for inbound connections, if the main configuration event_action
1416 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1417 early OCSP support, but no stapling (appears to be less than 1.0.0).
1419 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1420 server certificate names available under the smtp transport option
1421 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1424 JH/10 Time-related extraction expansions from certificates now use the main
1425 option "timezone" setting for output formatting, and are consistent
1426 between OpenSSL and GnuTLS compilations. Bug 1541.
1428 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1429 encoded parameter in the incoming message. Bug 1558.
1431 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1432 include certificate info, eximon was claiming there were spoolfile
1435 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1437 JH/14 Log delivery-related information more consistently, using the sequence
1438 "H=<name> [<ip>]" wherever possible.
1440 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1441 are problematic for Debian distribution, omit them from the release
1444 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1446 JH/16 Fix string representation of time values on 64bit time_t architectures.
1449 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1450 output list separator was used.
1455 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1456 checkers that were complaining about end of non-void function with no
1459 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1460 This was a regression introduced in 4.83 by another bugfix.
1462 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1464 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1465 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1471 TF/01 Correctly close the server side of TLS when forking for delivery.
1473 When a message was received over SMTP with TLS, Exim failed to clear up
1474 the incoming connection properly after forking off the child process to
1475 deliver the message. In some situations the subsequent outgoing
1476 delivery connection happened to have the same fd number as the incoming
1477 connection previously had. Exim would try to use TLS and fail, logging
1478 a "Bad file descriptor" error.
1480 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1481 utilities have not been installed.
1483 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1484 temporary space as the ACL may create new global variables.
1486 TL/01 LDAP support uses per connection or global context settings, depending
1487 upon the detected version of the libraries at build time.
1489 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1490 to extract and use the src ip:port in logging and expansions as if it
1491 were a direct connection from the outside internet. PPv2 support was
1492 updated based on HAProxy spec change in May 2014.
1494 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1496 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1497 Properly escape header and check for NULL return.
1499 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1502 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1504 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1505 characters in header names, implemented as a verify condition.
1506 Contributed by Michael Fischer v. Mollard.
1508 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1509 results permerror and temperror. Previous values are deprecated but
1510 still accepted. In a future release, err_perm and err_temp will be
1511 completely removed, which will be a backward incompatibility if the
1512 ACL tests for either of these two old results. Patch contributed by
1513 user bes-internal on the mailing list.
1515 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1517 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1518 selectors, in both main and reject logs.
1520 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1523 JH/07 Add malware type "sock" for talking to simple daemon.
1525 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1527 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1528 routers/transports under cutthrough routing.
1530 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1531 numbers. Touch up "bool" conditional to keep the same definition.
1533 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1535 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1537 JH/12 Expand items in router/transport headers_add or headers_remove lists
1538 individually rather than the list as a whole. Bug 1452.
1540 Required for reasonable handling of multiple headers_ options when
1541 they may be empty; requires that headers_remove items with embedded
1542 colons must have them doubled (or the list-separator changed).
1544 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1545 view the policy declared in the DMARC record. Currently, $dmarc_status
1546 is a combined value of both the record presence and the result of the
1549 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1551 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1552 dnslookup router and the smtp transport (applying to the forward
1555 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1556 of ldap servers used for a specific lookup. Patch provided by Heiko
1559 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1560 New variable $lookup_dnssec_authenticated for observability.
1562 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1563 Patch submitted by Lars Timman.
1565 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1567 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1568 Requires trusted mode and valid format message id, aborts otherwise.
1569 Patch contributed by Heiko Schlichting.
1571 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1572 certextract with support for various fields. Bug 1358.
1574 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1575 is requested by default, modifiable by smtp transport option
1578 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1579 operate on certificate variables to give certificate fingerprints
1580 Also new ${sha256:cert_variable}.
1582 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1584 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1585 Christian Aistleitner.
1587 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1589 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1590 file. Patch from Wolfgang Breyha.
1592 JH/25 Expand the coverage of the delivery $host and $host_address to
1593 client authenticators run in verify callout. Bug 1476.
1595 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1596 align with daemon_smtp_ports. Bug 72.
1598 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1599 support and error reporting did not work properly.
1601 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1602 and is readable. Patch from Andrew Colin Kissa.
1604 TL/14 Enhance documentation of ${run expansion and how it parses the
1605 commandline after expansion, particularly in the case when an
1606 unquoted variable expansion results in an empty value.
1608 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1610 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1611 by renaming away from it. Addresses GH issue 15, reported by
1614 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1616 TL/15 SECURITY: prevent double expansion in math comparison functions
1617 (can expand unsanitized data). Not remotely exploitable.
1624 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1626 PP/02 Make -n do something, by making it not do something.
1627 When combined with -bP, the name of an option is not output.
1629 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1632 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1633 $sender_host_name and config options to manage this, and basic check
1636 PP/05 DSCP support for outbound connections and control modifier for inbound.
1638 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1639 (Only plugin which currently uses this is kerberos4, which nobody should
1640 be using, but we should make it available and other future plugins might
1641 conceivably use it, even though it would break NAT; stuff *should* be
1642 using channel bindings instead).
1644 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1645 name; added for Sendmail compatibility; requires admin caller.
1646 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1647 just ignore it); requires trusted caller.
1648 Also parse but ignore: -Ac -Am -X<logfile>
1651 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1653 TL/02 Add +smtp_confirmation as a default logging option.
1655 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1656 Patch by Magnus Holmgren from 2007-02-20.
1658 TL/04 Bugzilla 1281 - Spec typo.
1659 Bugzilla 1283 - Spec typo.
1660 Bugzilla 1290 - Spec grammar fixes.
1662 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1664 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1666 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1667 mailing list by by Dmitry Isaikin.
1669 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1671 JH/02 Support "G" suffix to numbers in ${if comparisons.
1673 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1675 NM/01 Bugzilla 1197 - Spec typo
1676 Bugzilla 1196 - Spec examples corrections
1678 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1680 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1681 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1684 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1685 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1687 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1688 "acl {{name}{arg}...}", and optional args on acl condition
1691 JH/05 Permit multiple router/transport headers_add/remove lines.
1693 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1695 JH/07 Avoid using a waiting database for a single-message-only transport.
1696 Performance patch from Paul Fisher. Bugzilla 1262.
1698 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1701 JH/09 Add $headers_added variable, with content from use of ACL modifier
1702 add_header (but not yet added to the message). Bugzilla 199.
1704 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1705 Pulled from Bugzilla 817 by Wolfgang Breyha.
1707 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1709 (nb: this is the same fix as in Exim 4.80.1)
1711 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1714 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1717 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1719 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1720 advertise SMTP AUTH mechanism to us, instead of a generic
1721 protocol violation error. Also, make Exim more robust to bad
1722 data from the Dovecot auth socket.
1724 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1726 When a queue runner is handling a message, Exim first routes the
1727 recipient addresses, during which it prunes them based on the retry
1728 hints database. After that it attempts to deliver the message to
1729 any remaining recipients. It then updates the hints database using
1732 So if a recipient address works intermittently, it can get repeatedly
1733 deferred at routing time. The retry hints record remains fresh so the
1734 address never reaches the final cutoff time.
1736 This is a fairly common occurrence when a user is bumping up against
1737 their storage quota. Exim had some logic in its local delivery code
1738 to deal with this. However it did not apply to per-recipient defers
1739 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1741 This change adds a proper retry rule check during routing so that the
1742 final cutoff time is checked against the message's age. We only do
1743 this check if there is an address retry record and there is not a
1744 domain retry record; this implies that previous attempts to handle
1745 the address had the retry_use_local_parts option turned on. We use
1746 this as an approximation for the destination being like a local
1747 delivery, as in LMTP.
1749 I suspect this new check makes the old local delivery cutoff check
1750 redundant, but I have not verified this so I left the code in place.
1752 TF/02 Correct gecos expansion when From: is a prefix of the username.
1754 Test 0254 submits a message to Exim with the header
1758 When I ran the test suite under the user fanf2, Exim expanded
1759 the header to contain my full name, whereas it should have added
1760 a Resent-Sender: header. It erroneously treats any prefix of the
1761 username as equal to the username.
1763 This change corrects that bug.
1765 GF/01 DCC debug and logging tidyup
1766 Error conditions log to paniclog rather than rejectlog.
1767 Debug lines prefixed by "DCC: " to remove any ambiguity.
1769 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1771 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1772 Bug spotted by Jeremy Harris; was flawed since initial commit.
1773 Would have resulted in OCSP responses post-SNI triggering an Exim
1774 NULL dereference and crash.
1776 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1778 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1779 Bug detection, analysis and fix by Samuel Thibault.
1780 Bugzilla 1331, Debian bug #698092.
1782 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1784 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1785 Server implementation by Todd Lyons, client by JH.
1786 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1787 config variable "prdr_enable" controls whether the server
1788 advertises the facility. If the client requests PRDR a new
1789 acl_data_smtp_prdr ACL is called once for each recipient, after
1790 the body content is received and before the acl_smtp_data ACL.
1791 The client is controlled by both of: a hosts_try_prdr option
1792 on the smtp transport, and the server advertisement.
1793 Default client logging of deliveries and rejections involving
1794 PRDR are flagged with the string "PRDR".
1796 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1797 fclose(). Diagnosis by Todd Lyons.
1799 PP/17 Update configure.default to handle IPv6 localhost better.
1800 Patch by Alain Williams (plus minor tweaks).
1803 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1804 This is now consistent with GnuTLS, and is now documented: the
1805 previous undocumented portable approach to treating the option as
1806 unset was to force an expansion failure. That still works, and
1807 an empty string is now equivalent.
1809 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1810 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1811 not performing validation itself.
1813 PP/20 Added force_command boolean option to pipe transport.
1814 Patch from Nick Koston, of cPanel Inc.
1816 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1819 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1821 PP/21 Fix eximon continuous updating with timestamped log-files.
1822 Broken in a format-string cleanup in 4.80, missed when I repaired the
1823 other false fix of the same issue.
1824 Report and fix from Heiko Schlichting.
1827 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1828 Report from Prashanth Katuri.
1830 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1831 It's SecureTransport, so affects any MacOS clients which use the
1832 system-integrated TLS libraries, including email clients.
1834 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1835 using a MIME ACL for non-SMTP local injection.
1836 Report and assistance in diagnosis by Warren Baker.
1838 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1840 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1842 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1843 last id that failed so it may be referenced in subsequent ACL's.
1845 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1848 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1849 ldap library initialization, allowing self-signed CA's to be
1850 used. Also properly sets require_cert option later in code by
1851 using NULL (global ldap config) instead of ldap handle (per
1852 session). Bug diagnosis and testing by alxgomz.
1854 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1855 the src/util/ subdirectory.
1857 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1858 renamed to Transport Post Delivery Action by Jeremy Harris, as
1861 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1862 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1863 redis_servers = needs to be configured which will be used by the redis
1864 lookup. Patch from Warren Baker, of The Packet Hub.
1866 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1868 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1869 hostname or reverse DNS when processing a host list. Used suggestions
1870 from multiple comments on this bug.
1872 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1874 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1875 Missed a few lines, added it to make the runtest require no keyboard
1878 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1879 contains upper case chars. Make router use caseful_local_part.
1881 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1882 support when GnuTLS has been built with p11-kit.
1888 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1890 This, or similar/improved, will also be change PP/11 of 4.82.
1896 PP/01 Handle short writes when writing local log-files.
1897 In practice, only affects FreeBSD (8 onwards).
1898 Bugzilla 1053, with thanks to Dmitry Isaikin.
1900 NM/01 Bugzilla 949 - Documentation tweak
1902 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1905 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1907 PP/02 Implemented gsasl authenticator.
1909 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1911 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1912 `pkg-config foo` for cflags/libs.
1914 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1915 with rest of GSASL and with heimdal_gssapi.
1917 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1918 `pkg-config foo` for cflags/libs for the TLS implementation.
1920 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1921 properties get this fed in as external SSF. A number of robustness
1922 and debugging improvements to the cyrus_sasl authenticator.
1924 PP/08 cyrus_sasl server now expands the server_realm option.
1926 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1927 Patch by Jeremy Harris.
1929 PP/10 Added dbmjz lookup type.
1931 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1933 PP/12 MAIL args handles TAB as well as SP, for better interop with
1934 non-compliant senders.
1935 Analysis and variant patch by Todd Lyons.
1937 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1938 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1939 Patch from Dirk Mueller <dmueller@suse.com>
1941 PP/13 tls_peerdn now print-escaped for spool files.
1942 Observed some $tls_peerdn in wild which contained \n, which resulted
1943 in spool file corruption.
1945 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1946 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1947 or write after TLS renegotiation, which otherwise led to messages
1950 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1951 as a tracking header (ie: a signed header comes before the signature).
1952 Patch from Wolfgang Breyha.
1954 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1955 comma-sep list; embedded commas doubled.
1957 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1959 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1961 Report and patch from Dmitry Banschikov.
1963 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1964 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1965 failures. We appear to now support TLS1.1+ with Exim.
1967 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1968 lets Exim select keys and certificates based upon TLS SNI from client.
1969 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1970 before an outbound SMTP session. New log_selector, +tls_sni.
1972 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1973 NULL dereference. Report and patch from Alun Jones.
1975 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1976 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1977 Not seeing resolver debug output on NetBSD, but suspect this is a
1978 resolver implementation change.
1980 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1981 Left warnings. Added "eximon gdb" invocation mode.
1983 PP/21 Defaulting "accept_8bitmime" to true, not false.
1985 PP/22 Added -bw for inetd wait mode support.
1987 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1988 locate the relevant includes and libraries. Made this the default.
1990 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1991 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1993 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1994 This may cause build issues on older platforms.
1996 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1997 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1998 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1999 Added SNI support via GnuTLS too.
2000 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2002 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2004 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2005 Applied second patch from Janne, implementing suggestion to default
2006 multiple-strings-in-record handling to match SPF spec.
2008 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2010 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2011 read-only, out of scope).
2012 Patch from Wolfgang Breyha, report from Stuart Northfield.
2014 PP/29 Fix three issues highlighted by clang analyser static analysis.
2015 Only crash-plausible issue would require the Cambridge-specific
2016 iplookup router and a misconfiguration.
2017 Report from Marcin Mirosław.
2019 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2021 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2022 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2023 As part of this, removing so much warning spew let me fix some minor
2024 real issues in debug logging.
2026 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2027 assignment on my part. Fixed.
2029 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2030 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2031 Janne Snabb (who went above and beyond: thank you).
2033 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2034 string otherwise requires a connection and a bunch more work and it's
2035 relatively easy to get wrong. Should also expose TLS library linkage
2038 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2039 64-bit ${eval} (JH/03).
2041 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2042 GNU libc to support some of the 64-bit stuff, should not lead to
2043 conflicts. Defined before os.h is pulled in, so if a given platform
2044 needs to override this, it can.
2046 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2047 protection layer was required, which is not implemented.
2048 Bugzilla 1254, patch from Wolfgang Breyha.
2050 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2051 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2052 tls_dhparam take prime identifiers. Also unbreak combination of
2053 OpenSSL+DH_params+TLSSNI.
2055 PP/39 Disable SSLv2 by default in OpenSSL support.
2061 PP/01 Solaris build fix for Oracle's LDAP libraries.
2062 Bugzilla 1109, patch from Stephen Usher.
2064 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2066 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2069 TF/02 Fix a couple more cases where we did not log the error message
2070 when unlink() failed. See also change 4.74-TF/03.
2072 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2073 lock up or crash if it happened to be inside a call to libc when it
2074 got a SIGUSR1 from exiwhat.
2076 The SIGUSR1 handler appends the current process status to the process
2077 log which is later printed by exiwhat. It used to use the general
2078 purpose logging code to do this, but several functions it calls are
2079 not safe for signals.
2081 The new output code in the SIGUSR1 handler is specific to the process
2082 log, and simple enough that it's easy to inspect for signal safety.
2083 Removing some special cases also simplifies the general logging code.
2084 Removing the spurious timestamps from the process log simplifies
2087 TF/04 Improved ratelimit ACL condition.
2089 The /noupdate option has been deprecated in favour of /readonly which
2090 has clearer semantics. The /leaky, /strict, and /readonly update modes
2091 are mutually exclusive. The update mode is no longer included in the
2092 database key; it just determines when the database is updated. (This
2093 means that when you upgrade Exim will forget old rate measurements.)
2095 Exim now checks that the per_* options are used with an update mode that
2096 makes sense for the current ACL. For example, when Exim is processing a
2097 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2098 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2099 must specify per_mail/readonly. If you omit the update mode it defaults to
2100 /leaky where that makes sense (as before) or /readonly where required.
2102 The /noupdate option is now undocumented but still supported for
2103 backwards compatibility. It is equivalent to /readonly except that in
2104 ACLs where /readonly is required you may specify /leaky/noupdate or
2105 /strict/noupdate which are treated the same as /readonly.
2107 A useful new feature is the /count= option. This is a generalization
2108 of the per_byte option, so that you can measure the throughput of other
2109 aggregate values. For example, the per_byte option is now equivalent
2110 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2112 The per_rcpt option has been generalized using the /count= mechanism
2113 (though it's more complicated than the per_byte equivalence). When it is
2114 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2115 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2116 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2117 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2118 Note that using per_rcpt with a non-readonly update mode in more than
2119 one ACL will cause the recipients to be double-counted. (The per_mail
2120 and per_byte options don't have this problem.)
2122 The handling of very low rates has changed slightly. If the computed rate
2123 is less than the event's count (usually one) then this event is the first
2124 after a long gap. In this case the rate is set to the same as this event's
2125 count, so that the first message of a spam run is counted properly.
2127 The major new feature is a mechanism for counting the rate of unique
2128 events. The new per_addr option counts the number of different
2129 recipients that someone has sent messages to in the last time period. It
2130 behaves like per_rcpt if all the recipient addresses are different, but
2131 duplicate recipient addresses do not increase the measured rate. Like
2132 the /count= option this is a general mechanism, so the per_addr option
2133 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2134 example, measure the rate that a client uses different sender addresses
2135 with the options per_mail/unique=$sender_address. There are further
2136 details in the main documentation.
2138 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2140 TF/06 Removed a few PCRE remnants.
2142 TF/07 Automatically extract Exim's version number from tags in the git
2143 repository when doing development or release builds.
2145 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2146 Bugzilla 879. Patch from Paul Fisher.
2148 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2149 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2152 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2154 PP/05 Variable $av_failed, true if the AV scanner deferred.
2155 Bugzilla 1078. Patch from John Horne.
2157 PP/06 Stop make process more reliably on build failure.
2158 Bugzilla 1087. Patch from Heiko Schlittermann.
2160 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2161 Bugzilla 1089. Patch from Heiko Schlittermann.
2163 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2164 Bugzilla 1131. Patch from Holger Weiß.
2166 PP/09 Handle IPv6 addresses with SPF.
2167 Bugzilla 860. Patch from Wolfgang Breyha.
2169 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2171 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2174 PP/11 match_* no longer expand right-hand-side by default.
2175 New compile-time build option, EXPAND_LISTMATCH_RHS.
2176 New expansion conditions, "inlist", "inlisti".
2178 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2180 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2182 PP/14 fix log_write() format string regression from TF/03.
2183 Bugzilla 1152. Patch from Dmitry Isaikin.
2189 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2191 PP/02 Harmonised TLS library version reporting; only show if debugging.
2192 Layout now matches that introduced for other libraries in 4.74 PP/03.
2194 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2196 PP/04 New "dns_use_edns0" global option.
2198 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2201 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2202 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2204 TK/01 Updated PolarSSL code to 0.14.2.
2205 Bugzilla 1097. Patch from Andreas Metzler.
2207 PP/07 Catch divide-by-zero in ${eval:...}.
2208 Fixes bugzilla 1102.
2210 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2213 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2214 format-string attack -- SECURITY: remote arbitrary code execution.
2216 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2217 time unintentionally subject to list matching rules, letting the header
2218 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2219 arbitrary string expansion). This allowed for information disclosure.
2221 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2222 INT_MIN/-1 -- value coerced to INT_MAX.
2228 NM/01 Workaround for PCRE version dependency in version reporting
2231 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2232 This fixes portability to compilers other than gcc, notably
2233 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2235 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2236 makefiles for portability to HP-UX and POSIX correctness.
2238 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2239 Also via indented variable definition in the Makefile.
2240 (Debugging by Oliver Heesakkers).
2242 PP/02 Restore caching of spamd results with expanded spamd_address.
2243 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2245 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2246 Improves build reliability. Fix from: Frank Elsner
2248 NM/02 Fix wide character breakage in the rfc2047 coding
2249 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2251 NM/03 Allow underscore in dnslist lookups
2252 Fixes bug 1026. Patch from Graeme Fowler
2254 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2255 Code patches from Adam Ciarcinski of NetBSD.
2257 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2260 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2261 is logged, to avoid truncation. Patch from John Horne.
2263 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2264 Patch from Jakob Hirsch.
2266 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2267 SQL string expansion failure details.
2268 Patch from Andrey Oktyabrski.
2270 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2271 Patch from Simon Arlott.
2273 PP/09 New lookups functionality failed to compile on old gcc which rejects
2274 extern declarations in function scope.
2275 Patch from Oliver Fleischmann
2277 PP/10 Use sig_atomic_t for flags set from signal handlers.
2278 Check getgroups() return and improve debugging.
2279 Fixed developed for diagnosis in bug 927 (which turned out to be
2282 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2283 Patch from Mark Zealey.
2285 PP/12 Bugzilla 1056: Improved spamd server selection.
2286 Patch from Mark Zealey.
2288 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2289 Based on patch from Heiko Schlittermann.
2291 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2292 Patch from Uwe Doering, sign-off by Michael Haardt.
2294 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2295 variable declaration deep within a block. Bug and patch from
2298 PP/15 lookups-Makefile IRIX compatibility coercion.
2300 PP/16 Make DISABLE_DKIM build knob functional.
2302 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2303 Patch by Simon Arlott
2305 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2306 variable argument macros. Our copy now differs from upstream.
2312 TF/01 Failure to get a lock on a hints database can have serious
2313 consequences so log it to the panic log.
2315 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2316 controlled using the smtp_confirmation log selector.
2318 TF/03 Include the error message when we fail to unlink a spool file.
2320 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2321 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2322 for maintaining out-of-tree patches for some time.
2324 PP/01 Bugzilla 139: Documentation and portability issues.
2325 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2326 Handle per-OS dynamic-module compilation flags.
2328 PP/02 Let /dev/null have normal permissions.
2329 The 4.73 fixes were a little too stringent and complained about the
2330 permissions on /dev/null. Exempt it from some checks.
2331 Reported by Andreas M. Kirchwitz.
2333 PP/03 Report version information for many libraries, including
2334 Exim version information for dynamically loaded libraries. Created
2335 version.h, now support a version extension string for distributors
2336 who patch heavily. Dynamic module ABI change.
2338 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2339 privilege escalation vulnerability whereby the Exim run-time user
2340 can cause root to append content of the attacker's choosing to
2343 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2346 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2347 If dropping privileges for untrusted macros, we disabled normal logging
2348 on the basis that it would fail; for the Exim run-time user, this is not
2349 the case, and it resulted in successful deliveries going unlogged.
2350 Fixed. Reported by Andreas Metzler.
2356 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2357 only prepend for the Resent-* case. Fixes regression introduced in
2358 Exim 4.70 by NM/22 for Bugzilla 607.
2360 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2361 increasing numbers of administrators be bitten by this.
2363 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2365 PP/03 Bugzilla 994: added openssl_options main configuration option.
2367 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2369 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2371 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2373 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2374 without a peer certificate, leading to a segfault because of an
2375 assumption that peers always have certificates. Be a little more
2376 paranoid. Problem reported by Martin Tscholak.
2378 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2379 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2380 NB: ClamAV planning to remove STREAM in "middle of 2010".
2381 CL also introduces -bmalware, various -d+acl logging additions and
2382 more caution in buffer sizes.
2384 PP/09 Implemented reverse_ip expansion operator.
2386 PP/10 Bugzilla 937: provide a "debug" ACL control.
2388 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2390 PP/12 Bugzilla 973: Implement --version.
2392 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2394 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2396 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2398 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2399 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2400 ignore trailing whitespace.
2402 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2404 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2407 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2408 Notification from Dr Andrew Aitchison.
2410 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2411 ExtendedDetectionInfo response format.
2412 Notification from John Horne.
2414 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2417 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2418 XSL and documented dependency on system catalogs, with examples of how
2421 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2424 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2425 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2426 configuration file which is writeable by the Exim user or group.
2428 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2429 of configuration files to cover files specified with the -C option if
2430 they are going to be used with root privileges, not just the default
2433 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2434 option (effectively making it always true).
2436 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2437 files to be used while preserving root privileges.
2439 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2440 that rogue child processes cannot use them.
2442 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2443 run-time user, instead of root.
2445 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2446 Exim run-time user without dropping privileges.
2448 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2449 result string, instead of calling string_vformat() twice with the same
2452 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2453 for other users. Others should always drop root privileges if they use
2454 -C on the command line, even for a whitelisted configure file.
2456 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2458 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2464 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2465 $data_path, and $header_path variables; fixed documentation bugs and
2468 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2469 exipick to access non-standard spools, including the "frozen" queue
2472 NM/01 Bugzilla 965: Support mysql stored procedures.
2473 Patch from Alain Williams
2475 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2477 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2478 Patch from Andreas Metzler
2480 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2481 Patch from Kirill Miazine
2483 NM/05 Bugzilla 671: Added umask to procmail example.
2485 JJ/03 installed exipick 20100323.0, fixing doc bug
2487 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2488 directory. Notification and patch from Dan Rosenberg.
2490 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2492 TK/02 Improve log output when DKIM signing operation fails.
2494 MH/01 Treat the transport option dkim_domain as a colon separated
2495 list, not as a single string, and sign the message with each element,
2496 omitting multiple occurences of the same signer.
2498 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2499 Bugzilla 985, 986. Patch by Simon Arlott
2501 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2502 Patch by Simon Arlott
2504 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2505 MBX locking. Notification from Dan Rosenberg.
2511 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2513 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2515 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2517 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2519 NM/04 Bugzilla 915: Flag broken perl installation during build.
2525 TK/01 Added patch by Johannes Berg that expands the main option
2526 "spamd_address" if it starts with a dollar sign.
2528 TK/02 Write list of recipients to X-Envelope-Sender header when building
2529 the mbox-format spool file for content scanning (suggested by Jakob
2532 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2533 (http://www.dcc-servers.net/) support via dccifd. Activated by
2534 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2536 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2537 by Mark Daniel Reidel <mr@df.eu>.
2539 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2540 When building exim an external PCRE library is now needed -
2541 PCRE is a system library on the majority of modern systems.
2542 See entry on PCRE_LIBS in EDITME file.
2544 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2545 conversation. Added nologin parameter to request.
2546 Patch contributed by Kirill Miazine.
2548 TF/01 Do not log submission mode rewrites if they do not change the address.
2550 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2552 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2553 log files in place. Contributed by Roberto Lima.
2555 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2557 TF/03 Bugzilla 615: When checking the local_parts router precondition
2558 after a local_part_suffix or local_part_prefix option, Exim now
2559 does not use the address's named list lookup cache, since this
2560 contains cached lookups for the whole local part.
2562 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2563 Robert Millan. Documentation is in experimental-spec.txt.
2565 TF/04 Bugzilla 668: Fix parallel build (make -j).
2567 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2569 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2570 Patch provided by Jan Srzednicki.
2572 TF/05 Leading white space used to be stripped from $spam_report which
2573 wrecked the formatting. Now it is preserved.
2575 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2576 that they are available at delivery time.
2578 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2580 TF/08 TLS error reporting now respects the incoming_interface and
2581 incoming_port log selectors.
2583 TF/09 Produce a more useful error message if an SMTP transport's hosts
2584 setting expands to an empty string.
2586 NM/06 Bugzilla 744: EXPN did not work under TLS.
2587 Patch provided by Phil Pennock.
2589 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2590 Patch provided by Richard Godbee.
2592 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2593 acl_smtp_notquit, added index entry.
2595 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2596 Patch provided by Eugene Bujak.
2598 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2599 accept(). Patch provided by Maxim Dounin.
2601 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2602 Patch provided by Phil Pennock.
2604 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2606 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2607 Patch provided by Brad "anomie" Jorsch.
2609 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2610 Patch provided by Dean Brooks.
2612 TK/05 Add native DKIM support (does not depend on external libraries).
2614 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2615 Patch provided by Graeme Fowler.
2617 NM/16 Bugzilla 851: Documentation example syntax fix.
2619 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2621 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2624 NM/19 Bugzilla 745: TLS version reporting.
2625 Patch provided by Phil Pennock.
2627 NM/20 Bugzilla 167: bool: condition support.
2628 Patch provided by Phil Pennock.
2630 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2631 clients. Patch provided by Phil Pennock.
2633 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2634 Patch provided by Brad "anomie" Jorsch.
2636 NM/23 Bugzilla 687: Fix misparses in eximstats.
2637 Patch provided by Heiko Schlittermann.
2639 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2640 Patch provided by Heiko Schlittermann.
2642 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2643 plus update to original patch.
2645 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2647 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2648 Patch provided by David Brownlee.
2650 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2652 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2654 NM/30 Bugzilla 888: TLS documentation bugfixes.
2656 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2658 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2659 Unlike the original bugzilla I have changed all shell scripts in src tree.
2661 NM/33 Bugzilla 898: Transport filter timeout fix.
2662 Patch by Todd Rinaldo.
2664 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2665 Patch by Serge Demonchaux.
2667 NM/35 Bugzilla 39: Base64 decode bug fixes.
2668 Patch by Jakob Hirsch.
2670 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2672 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2674 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2676 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2682 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2683 ALT-N's libdkim that I have put here:
2684 http://duncanthrax.net/exim-experimental/
2686 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2687 were called 'true' and it seems that C99 defines that as a reserved
2688 keyword to be used with 'bool' variable types. That means you could
2689 not include C99-style headers which use bools without triggering
2690 build errors in sieve.c.
2692 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2693 as mailq or other aliases. Changed the --help handling significantly
2694 to do whats expected. exim_usage() emits usage/help information.
2696 SC/01 Added the -bylocaldomain option to eximstats.
2698 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2700 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2702 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2708 PH/01 Another patch from the Sieve maintainer.
2710 PH/02 When an IPv6 address is converted to a string for single-key lookup
2711 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2712 dots are used instead of colons so that keys in lsearch files need not
2713 contain colons. This was done some time before quoting was made available
2714 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2715 (notated using the quote facility) so as to distinguish them from IPv4
2716 keys. This meant that lookups for IP addresses in host lists did not work
2717 for iplsearch lookups.
2719 This has been fixed by arranging for IPv6 addresses to be expressed with
2720 colons if the lookup type is iplsearch. This is not incompatible, because
2721 previously such lookups could never work.
2723 The situation is now rather anomalous, since one *can* have colons in
2724 ordinary lsearch keys. However, making the change in all cases is
2725 incompatible and would probably break a number of configurations.
2727 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2730 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2731 conversion specification without a maximum field width, thereby enabling
2732 a rogue spamd server to cause a buffer overflow. While nobody in their
2733 right mind would setup Exim to query an untrusted spamd server, an
2734 attacker that gains access to a server running spamd could potentially
2735 exploit this vulnerability to run arbitrary code as the Exim user.
2737 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2738 $primary_hostname instead of what libspf2 thinks the hosts name is.
2740 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2741 a directory entry by the name of the lookup key. Previously, if a
2742 symlink pointed to a non-existing file or a file in a directory that
2743 Exim lacked permissions to read, a lookup for a key matching that
2744 symlink would fail. Now it is enough that a matching directory entry
2745 exists, symlink or not. (Bugzilla 503.)
2747 PH/03 The body_linecount and body_zerocount variables are now exported in the
2750 PH/04 Added the $dnslist_matched variable.
2752 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2753 This means they are set thereafter only if the connection becomes
2756 PH/06 Added the client_condition to authenticators so that some can be skipped
2757 by clients under certain conditions.
2759 PH/07 The error message for a badly-placed control=no_multiline_responses left
2760 "_responses" off the end of the name.
2762 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2764 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2765 (without spaces) instead of just copying the configuration text.
2767 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2769 PH/11 Added $max_received_linelength.
2771 PH/12 Added +ignore_defer and +include_defer to host lists.
2773 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2774 way in which PCRE > 7.0 is built.
2776 PH/14 Implemented queue_only_load_latch.
2778 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2779 MAIL command. The effect was to mangle the value on 64-bit systems.
2781 PH/16 Another patch from the Sieve maintainer.
2783 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2785 PH/18 If a system quota error occurred while trying to create the file for
2786 a maildir delivery, the message "Mailbox is full" was not appended to the
2787 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2788 only to a quota excession during the actual writing of the file.
2790 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2791 characters?) which causes problems in log lines. The DN values are now
2792 passed through string_printing() before being added to log lines.
2794 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2795 and InterBase are left for another time.)
2797 PH/21 Added message_body_newlines option.
2799 PH/22 Guard against possible overflow in moan_check_errorcopy().
2801 PH/23 POSIX allows open() to be a macro; guard against that.
2803 PH/24 If the recipient of an error message contained an @ in the local part
2804 (suitably quoted, of course), incorrect values were put in $domain and
2805 $local_part during the evaluation of errors_copy.
2811 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2812 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2815 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2816 issue a MAIL command.
2818 PH/02 In an ACL statement such as
2820 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2822 if a client was not listed at all, or was listed with a value other than
2823 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2824 the condition was not true (as it should be), so access was not denied.
2825 The bug was that the ! inversion was incorrectly passed on to the second
2826 item. This has been fixed.
2828 PH/03 Added additional dnslists conditions == and =& which are different from
2829 = and & when the dns lookup returns more than one IP address.
2831 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2832 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2834 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2835 FSYNC, which compiles an option called disable_fsync that allows for
2836 bypassing fsync(). The documentation is heavily laced with warnings.
2838 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2840 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2841 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2842 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2843 including adding "make clean"; (3) Added -fPIC when compiling the test
2844 dynamically loaded module, to get rid of a warning.
2846 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2847 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2848 = 0s. The bug is otherwise harmless.
2850 PH/07 There was a bug in the dovecot authenticator such that the value of
2851 $auth1 could be overwritten, and so not correctly preserved, after a
2852 successful authentication. This usually meant that the value preserved by
2853 the server_setid option was incorrect.
2855 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2857 PH/09 Installed PCRE release 7.0.
2859 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2860 run for batched SMTP input. It is now run at the start of every message
2861 in the batch. While fixing this I discovered that the process information
2862 (output by running exiwhat) was not always getting set for -bs and -bS
2863 input. This is fixed, and it now also says "batched" for BSMTP.
2865 PH/11 Added control=no_pipelining.
2867 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2868 patch, slightly modified), and move the expansion of helo_data till after
2869 the connection is made in the smtp transport (so it can use these
2872 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2874 PH/14 Added log_selector = +pid.
2876 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2878 PH/16 Add ${if forany and ${if forall.
2880 PH/17 Added dsn_from option to vary the From: line in DSNs.
2882 PH/18 Flush SMTP output before performing a callout, unless control =
2883 no_callout_flush is set.
2885 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2886 was true (the default) a successful delivery failed to delete the retry
2887 item, thus causing premature timeout of the address. The bug is now
2890 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2892 PH/21 Long custom messages for fakedefer and fakereject are now split up
2893 into multiline responses in the same way that messages for "deny" and
2894 other ACL rejections are.
2896 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2897 with slight modification.
2899 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2900 draft, changing the syntax and factoring some duplicate code".
2902 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2903 for deliveries of the second and subsequent messages over the same SMTP
2906 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2907 ${reduce, with only minor "tidies".
2909 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2911 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2912 expansion side effects.
2914 PH/27 When a message times out after an over-quota error from an Exim-imposed
2915 quota, the bounce message says "mailbox is full". This message was not
2916 being given when it was a system quota that was exceeded. It now should
2919 MH/03 Made $recipients available in local_scan(). local_scan() already has
2920 better access to the recipient list through recipients_list[], but
2921 $recipients can be useful in postmaster-provided expansion strings.
2923 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2924 in the case of a MAIL command with additional options following the
2925 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2926 were accidentally chopped off.
2928 PH/29 SMTP synchronization checks are implemented when a command is read -
2929 there is a check that no more input is waiting when there shouldn't be
2930 any. However, for some commands, a delay in an ACL can mean that it is
2931 some time before the response is written. In this time, more input might
2932 arrive, invalidly. So now there are extra checks after an ACL has run for
2933 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2934 pipelining has not been advertised.
2936 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2938 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2939 returned in the SMTP response when smtp_return_error_details was false.
2940 This has been fixed.
2942 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2943 instead of the C I/O that was originally supplied, because problems were
2944 reported on Solaris.
2946 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2947 Exim which did not show up earlier: it was assuming that a call to
2948 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2949 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2950 was a macro that became an assignment, so it seemed to work. This has
2951 changed to a proper function call with a void return, hence the compile
2952 error. Exim's code has been fixed.
2954 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2957 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2959 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2961 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2962 because it turns out to be too big or there is a timeout) there is no
2963 "Completed" line in the log. When some messages of this type were
2964 selected by exigrep, they were listed as "not completed". Others were
2965 picked up by some special patterns. I have improved the selection
2966 criteria to be more general.
2968 PH/38 The host_find_failed option in the manualroute router can now be set
2969 to "ignore", to completely ignore a host whose IP address cannot be
2970 found. If all hosts are ignored, the behaviour is controlled by the new
2971 host_all_ignored option.
2973 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2974 homing or because of multiple MX records with /mx) generated more than
2975 one IP address, and the following item turned out to be the local host,
2976 all the secondary addresses of the first item were incorrectly removed
2977 from the list, along with the local host and any following hosts (which
2978 is what is supposed to happen).
2980 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2981 whoever called Exim into the -H file. In the case of the daemon it was
2982 behaving confusingly. When first started, it used values for whoever
2983 started the daemon, but after a SIGHUP it used the Exim user (because it
2984 calls itself on a restart). I have changed the code so that it now always
2987 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2988 message are rejected with the same error (e.g. no authentication or bad
2989 sender address), and a DATA command is nevertheless sent (as can happen
2990 with PIPELINING or a stupid MUA), the error message that was given to the
2991 RCPT commands is included in the rejection of the DATA command. This is
2992 intended to be helpful for MUAs that show only the final error to their
2995 PH/42 Another patch from the Sieve maintainer.
2997 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2998 Eximstats - Fixed some broken HTML links and added missing column headers
3000 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3001 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3003 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3005 PH/43 Yet another patch from the Sieve maintainer.
3007 PH/44 I found a way to check for a TCP/IP connection going away before sending
3008 the response to the final '.' that terminates a message, but only in the
3009 case where the client has not sent further data following the '.'
3010 (unfortunately, this is allowed). However, in many cases there won't be
3011 any further data because there won't be any more messages to send. A call
3012 to select() can be used: if it shows that the input is "ready", there is
3013 either input waiting, or the socket has been closed. An attempt to read
3014 the next input character can distinguish the two cases. Previously, Exim
3015 would have sent an OK response which the client would never have see.
3016 This could lead to message repetition. This fix should cure that, at
3017 least in a lot of common cases.
3019 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3020 advertised in response to EHLO.
3026 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3027 fixed by 4.65/MH/01 (is this a record?) are fixed:
3029 (i) An empty string was always treated as zero by the numeric comparison
3030 operators. This behaviour has been restored.
3032 (ii) It is documented that the numeric comparison operators always treat
3033 their arguments as decimal numbers. This was broken in that numbers
3034 starting with 0 were being interpreted as octal.
3036 While fixing these problems I realized that there was another issue that
3037 hadn't been noticed. Values of message_size_limit (both the global option
3038 and the transport option) were treated as octal if they started with 0.
3039 The documentation was vague. These values are now always treated as
3040 decimal, and I will make that clear in the documentation.
3046 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3047 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3050 MH/01 Don't check that the operands of numeric comparison operators are
3051 integers when their expansion is in "skipping" mode (fixes bug
3052 introduced by 4.64-PH/07).
3054 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3055 child addresses, Exim now panics and dies. Previously, because the count
3056 is held in a short int, deliveries were likely to be lost. As such a
3057 large number of recipients for a single message is ridiculous
3058 (performance will be very, very poor), I have chosen to impose a limit
3059 rather than extend the field.
3065 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3066 leftover -K file (the existence of which was triggered by #402).
3067 While we were at it, introduced process PID as part of the -K
3068 filename. This should rule out race conditions when creating
3071 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3072 processing considerably. Previous code took too long for large mails,
3073 triggering a timeout which in turn triggers #401.
3075 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3076 in the DK code in transports.c. sendfile() is not really portable,
3077 hence the _LINUX specificness.
3079 TF/01 In the add_headers option to the mail command in an Exim filter,
3080 there was a bug that Exim would claim a syntax error in any
3081 header after the first one which had an odd number of characters
3084 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3085 callout verification, Exim cached a "reject" for the entire domain. This
3086 is correct for most verifications, but it is not correct for a recipient
3087 verification with use_sender or use_postmaster set, because in that case
3088 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3089 case of MAIL FROM:<> rejection from other early rejections (e.g.
3090 rejection of HELO). When verifying a recipient using a non-null MAIL
3091 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3092 Whatever the result of the callout, the value of the domain cache is
3093 left unchanged (for any other kind of callout, getting as far as trying
3094 RCPT means that the domain itself is ok).
3096 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3099 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3100 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3101 session, the daemon ignores ECONNECT errors and logs others; it now
3102 ignores EPIPE as well.
3104 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3105 (quoted-printable decoding).
3107 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3108 later the small subsequent patch to fix an introduced bug.
3110 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3112 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3114 PH/08 An error is now given if message_size_limit is specified negative.
3116 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3117 to be given (somewhat) arbitrary names.
3119 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3122 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3123 miscellaneous code fixes
3125 PH/10 Added the log_reject_target ACL modifier to specify where to log
3128 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3129 hostname. This is wrong, because it relates to the incoming message (and
3130 probably the interface on which it is arriving) and not to the outgoing
3131 callout (which could be using a different interface). This has been
3132 changed to use the value of the helo_data option from the smtp transport
3133 instead - this is what is used when a message is actually being sent. If
3134 there is no remote transport (possible with a router that sets up host
3135 addresses), $smtp_active_hostname is used.
3137 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3138 tweaks were necessary in order to get it to work (see also 21 below):
3139 (a) The code assumed that strncpy() returns a negative number on buffer
3140 overflow, which isn't the case. Replaced with Exim's string_format()
3142 (b) There were several signed/unsigned issues. I just did the minimum
3143 hacking in of casts. There is scope for a larger refactoring.
3144 (c) The code used strcasecmp() which is not a standard C function.
3145 Replaced with Exim's strcmpic() function.
3146 (d) The code set only $1; it now sets $auth1 as well.
3147 (e) A simple test gave the error "authentication client didn't specify
3148 service in request". It would seem that Dovecot has changed its
3149 interface. Fortunately there's a specification; I followed it and
3150 changed what the client sends and it appears to be working now.
3152 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3155 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3156 address A is aliased to B and C, where B exists and C does not. Without
3157 -v the output is "A verified" because verification stops after a
3158 successful redirection if more than one address is generated. However,
3159 with -v the child addresses are also verified. Exim was outputting "A
3160 failed to verify" and then showing the successful verification for C,
3161 with its parentage. It now outputs "B failed to verify", showing B's
3162 parentage before showing the successful verification of C.
3164 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3165 look up a TXT record in a specific list after matching in a combined
3168 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3169 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3170 they consult the DNS. I had assumed they would set it the way they
3171 wanted; and indeed my experiments on Linux seem to show that in some
3172 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3173 To be on the safe side, however, I have now made the interface to
3174 host_find_byname() similar to host_find_bydns(), with an argument
3175 containing the DNS resolver options. The host_find_byname() function now
3176 sets these options at its start, just as host_find_bydns() does. The smtp
3177 transport options dns_qualify_single and dns_search_parents are passed to
3178 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3179 of host_find_byname() use the default settings of RES_DEFNAMES
3180 (qualify_single) but not RES_DNSRCH (search_parents).
3182 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3183 spool_read_header() do less string testing, by means of a preliminary
3184 switch on the second character of optional "-foo" lines. (This is
3185 overdue, caused by the large number of possibilities that now exist.
3186 Originally there were few.) While I was there, I also converted the
3187 str(n)cmp tests so they don't re-test the leading "-" and the first
3188 character, in the hope this might squeeze out yet more improvement.
3190 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3191 flag allowing group syntax was set by the header_syntax check but not
3192 turned off, possible causing trouble later; (2) The flag was not being
3193 set at all for the header_verify test, causing "group"-style headers to
3194 be rejected. I have now set it in this case, and also caused header_
3195 verify to ignore an empty address taken from a group. While doing this, I
3196 came across some other cases where the code for allowing group syntax
3197 while scanning a header line wasn't quite right (mostly, not resetting
3198 the flag correctly in the right place). These bugs could have caused
3199 trouble for malformed header lines. I hope it is now all correct.
3201 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3202 with the "reply" argument non-NULL. The code, however (which originally
3203 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3204 but it didn't always do it. This confused somebody who was copying the
3205 code for some other use. I have removed all the tests.
3207 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3208 feature that was used to support insecure browsers during the U.S. crypto
3209 embargo. It requires special client support, and Exim is probably the
3210 only MTA that supported it -- and would never use it because real RSA is
3211 always available. This code has been removed, because it had the bad
3212 effect of slowing Exim down by computing (never used) parameters for the
3213 RSA_EXPORT functionality.
3215 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3216 authenticator to fail if there's a tab character in the incoming data
3217 (there should never be unless someone is messing about, as it's supposed
3218 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3219 if the connection is using TLS or if the remote IP is the same as the
3220 local IP, and the "valid-client-cert option" if a client certificate has
3223 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3224 authenticators. This can be used for authorization after authentication
3225 succeeds. (In the case of plaintext, it servers for both authentication
3228 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3229 if any retry times were supplied.
3231 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3232 connection, where there is no client IP address to check. In this
3233 situation, the verify now always succeeds.
3235 PH/25 Applied John Jetmore's -Mset patch.
3237 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3239 PH/27 In a string expansion for a processed (not raw) header when multiple
3240 headers of the same name were present, leading whitespace was being
3241 removed from all of them, but trailing whitespace was being removed only
3242 from the last one. Now trailing whitespace is removed from each header
3243 before concatenation. Completely empty headers in a concatenation (as
3244 before) are ignored.
3246 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3247 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3249 PH/29 [Removed. This was a change that I later backed out, and forgot to
3250 correct the ChangeLog entry (that I had efficiently created) before
3251 committing the later change.]
3253 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3254 address errors (4xx response to RCPT) over the same connection as other
3255 messages routed to the same hosts. Such deliveries are always "forced",
3256 so retry times are not inspected. This resulted in far too many retries
3257 for the affected addresses. The effect occurred only when there were more
3258 hosts than the hosts_max_try setting in the smtp transport when it had
3259 the 4xx errors. Those hosts that it had tried were not added to the list
3260 of hosts for which the message was waiting, so if all were tried, there
3261 was no problem. Two fixes have been applied:
3263 (i) If there are any address or message errors in an SMTP delivery, none
3264 of the hosts (tried or untried) are now added to the list of hosts
3265 for which the message is waiting, so the message should not be a
3266 candidate for sending over the same connection that was used for a
3267 successful delivery of some other message. This seems entirely
3268 reasonable: after all the message is NOT "waiting for some host".
3269 This is so "obvious" that I'm not sure why it wasn't done
3270 previously. Hope I haven't missed anything, but it can't do any
3271 harm, as the worst effect is to miss an optimization.
3273 (ii) If, despite (i), such a delivery is accidentally attempted, the
3274 routing retry time is respected, so at least it doesn't keep
3275 hammering the server.
3277 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3278 in ${readsocket because some servers need this prod.
3280 PH/32 Added some extra debug output when updating a wait-xxx database.
3282 PH/33 The hint "could be header name not terminated by colon", which has been
3283 given for certain expansion errors for a long time, was not being given
3284 for the ${if def:h_colon_omitted{... case.
3286 PH/34 The spec says: "With one important exception, whenever a domain list is
3287 being scanned, $domain contains the subject domain." There was at least
3288 one case where this was not true.
3290 PH/35 The error "getsockname() failed: connection reset by peer" was being
3291 written to the panic log as well as the main log, but it isn't really
3292 panic-worthy as it just means the connection died rather early on. I have
3293 removed the panic log writing for the ECONNRESET error when getsockname()
3296 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3297 runs only) independently of the message's sender address. This meant
3298 that, if the 4xx error was in fact related to the sender, a different
3299 message to the same recipient with a different sender could confuse
3300 things. In particular, this can happen when sending to a greylisting
3301 server, but other circumstances could also provoke similar problems.
3302 I have changed the default so that the retry time for these errors is now
3303 based a combination of the sender and recipient addresses. This change
3304 can be overridden by setting address_retry_include_sender=false in the
3307 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3308 remote server are returned as part of bounce messages. This was not
3309 happening for LMTP over a pipe (the lmtp transport), but now it is the
3310 same for both kinds of LMTP.
3312 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3313 in header lines that were in fact CNAMEs. This is no longer the case.
3315 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3316 and queue runs started by the daemon processed all messages. This has
3317 been fixed so that -R and -S can now usefully be given with -q<time>.
3319 PH/40 Import PCRE release 6.7 (fixes some bugs).
3321 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3323 PH/42 Give an error if -q is specified more than once.
3325 PH/43 Renamed the variables $interface_address and $interface_port as
3326 $received_ip_address and $received_port, to make it clear that these
3327 values apply to message reception, and not to the outgoing interface when
3328 a message is delivered. (The old names remain recognized, of course.)
3330 PH/44 There was no timeout on the connect() call when using a Unix domain
3331 socket in the ${readsocket expansion. There now is.
3333 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3334 be meaningful with "accept".
3336 SC/01 Eximstats V1.43
3337 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3339 SC/02 Eximstats V1.44
3340 Use a glob alias rather than an array ref in the generated
3341 parser. This improves both readability and performance.
3343 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3344 Collect SpamAssassin and rejection statistics.
3345 Don't display local sender or destination tables unless
3346 there is data to show.
3347 Added average volumes into the top table text output.
3349 SC/04 Eximstats V1.46
3350 Collect data on the number of addresses (recipients)
3351 as well as the number of messages.
3353 SC/05 Eximstats V1.47
3354 Added 'Message too big' to the list of mail rejection
3355 reasons (thanks to Marco Gaiarin).
3357 SC/06 Eximstats V1.48
3358 Mainlog lines which have GMT offsets and are too short to
3359 have a flag are now skipped.
3361 SC/07 Eximstats V1.49 (Alain Williams)
3362 Added the -emptyok flag.
3364 SC/08 Eximstats V1.50
3365 Fixes for obtaining the IP address from reject messages.
3367 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3368 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3369 whitespace changes from 4.64-PH/27
3371 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3374 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3375 are found, allow negative numbers in numeric criteria)
3377 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3379 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3380 to match changes made in 4.64-PH/43
3382 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3384 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3385 "domain" is actually an IP address, and force a failure. This locks out
3386 those revolvers/nameservers that support "A-for-A" lookups, in
3387 contravention of the specifications.
3389 PH/48 When a host name was looked up from an IP address, and the subsequent
3390 forward lookup of the name timed out, the host name was left in
3391 $sender_host_name, contrary to the specification.
3393 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3394 restricted to single-key lookups, Exim was not diagnosing an error if
3395 * or *@ was used with a query-style lookup.
3397 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3399 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3400 long ago, but noone interested enough thought of it. Let's just say that
3401 the "1.1" means that there are some new functions that weren't there at
3402 some point in the past.
3404 PH/51 Error processing for expansion failure of helo_data from an smtp
3405 transport during callout processing was broken.
3407 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3408 tested/used via the -bh/-bhc/-bs options.
3410 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3411 bug, fixed in subsequent PCRE releases).
3413 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3414 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3416 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3422 SC/01 Use a glob alias rather than an array ref in eximstats generated
3423 parser. This improves both readability and performance.
3425 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3426 Don't display local sender or destination tables in eximstats unless
3427 there is data to show.
3428 Added average volumes into the eximstats top table text output.
3430 SC/03 Collect data on the number of addresses (recipients) as well
3431 as the number of messages in eximstats.
3433 TF/01 Correct an error in the documentation for the redirect router. Exim
3434 does (usually) call initgroups() when daemonizing.
3436 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3437 with consistent privilege compared to when running as a daemon.
3439 TF/03 Note in the spec that $authenticated_id is not set for local
3440 submissions from trusted users.
3442 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3443 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3445 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3446 by adding some example configuration directives to the default
3447 configuration file. A little bit of work is required to uncomment the
3448 directives and define how usernames and passwords are checked, but
3449 there is now a framework to start from.
3451 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3452 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3453 without this. I don't know how relevant this is to other LDAP libraries.
3455 PH/02 Add the verb name to the "unknown ACL verb" error.
3457 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3459 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3461 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3462 directory not expanded when it should be if an expanded home directory
3463 was set for the address (which is overridden by the transport).
3465 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3468 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3469 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3470 because it is too late at that time, and has no effect.
3472 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3473 security issue with \' (bugzilla #107). I could not use the
3474 PQescapeStringConn() function, because it needs a PGconn value as one of
3477 PH/08 When testing addresses using -bt, indicate those final addresses that
3478 are duplicates that would not cause an additional delivery. At least one
3479 person was confused, thinking that -bt output corresponded to deliveries.
3480 (Suppressing duplicates isn't a good idea as you lose the information
3481 about possibly different redirections that led to the duplicates.)
3483 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3484 systems where poll() doesn't work, in particular OS X.
3486 PH/10 Added more information to debugging output for retry time not reached.
3488 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3489 operations in malware.c.
3491 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3494 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3495 syslog_duplication set false, log lines that would normally be written
3496 both the the main log and to the reject log were not written to syslog at
3499 PH/14 In the default configuration, change the use of "message" in ACL warn
3500 statements to "add_header".
3502 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3503 not followed by a command (e.g. "seen endif").
3505 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3506 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3509 PH/17 Added extra conditions to the default value of delay_warning_condition
3513 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3514 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3515 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3518 The Auto-Submitted: and various List- headers are standardised, whereas I
3519 don't think Precedence: ever was.
3521 PH/18 Refactored debugging code in route_finduser() to show more information,
3522 in particular, the error code if getpwnam() issues one.
3524 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3525 This is apparently needed in addition to the PH/07 change above to avoid
3526 any possible encoding problems.
3528 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3529 but not after initializing Perl.
3531 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3532 output them only if debugging. By default they are written stderr,
3533 apparently, which is not desirable.
3535 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3538 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3541 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3543 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3544 authenticated or an ident call has been made. Suppress the default
3545 values for $authenticated_id and $authenticated_sender (but permit -oMai
3546 and -oMas) when testing with -bh.
3548 PH/24 Re-jigged the order of the tests in the default configuration so that the
3549 tests for valid domains and recipients precede the DNS black list and CSA
3550 tests, on the grounds that those ones are more expensive.
3552 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3553 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3554 This bug exists in every version of Exim that I still have, right back to
3557 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3558 However, an attempt to turn on case-sensitivity in a regex key by
3559 including (?-i) didn't work because the subject string was already
3560 lowercased, and the effects were non-intuitive. It turns out that a
3561 one-line patch can be used to allow (?-i) to work as expected.
3567 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3568 other effects) broke the use of negated acl sub-conditions.
3570 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3573 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3574 "Deny" causes Exim to reject the incoming connection with a 554 error.
3575 Unfortunately, if there is a major crisis, such as a disk failure,
3576 tcp-wrappers gives "deny", whereas what one would like would be some
3577 kind of temporary error. A kludge has been added to help with this.
3578 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3579 554 error is used if errno is still zero or contains ENOENT (which occurs
3580 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3583 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3585 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3586 errors. Otherwise a message that provokes a temporary error (when other
3587 messages do not) can cause a whole host to time out.
3589 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3590 addresses were routed directly to files or pipes from a redirect router.
3591 File deliveries just didn't batch; pipe deliveries might have suffered
3594 PH/06 A failure to get a lock for a hints database would erroneously always say
3595 "Failed to get write lock", even when it was really a read lock.
3597 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3598 of 0600. This has been changed to use the value of the lockfile_mode
3599 option (which defaults to 0600).
3601 PH/08 Applied small patch from the Sieve maintainer.
3603 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3604 folder from quota calculations, a direct delivery into this folder messed
3605 up the contents of the maildirsize file. This was because the regex was
3606 used only to exclude .Trash (or whatever) when the size of the mailbox
3607 was calculated. There was no check that a delivery was happening into an
3608 excluded directory. This bug has been fixed by ignoring all quota
3609 processing for deliveries into excluded directories.
3611 PH/10 Added the maildirfolder_create_regex option to appendfile.
3617 PH/01 The code for finding all the local interface addresses on a FreeBSD
3618 system running IPv6 was broken. This may well have applied to all BSD
3619 systems, as well as to others that have similar system calls. The broken
3620 code found IPv4 interfaces correctly, but gave incorrect values for the
3621 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3622 that it would not match correctly against @[] and not recognize the IPv6
3625 PH/02 The ipliteral router was not recognizing addresses of the form user@
3626 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3628 PH/03 Added disable_ipv6.
3630 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3631 decoded form, because it is most often used to construct To: headers
3632 lines in autoreplies, and the decoded form may well be syntactically
3633 invalid. However, $reply_address has leading white space removed, and all
3634 newlines turned into spaces so that the autoreply transport does not
3637 PH/05 If group was specified without a user on a router, and no group or user
3638 was specified on a transport, the group from the router was ignored.
3640 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3641 for visible compile-time settings that can be used to change these
3642 numbers, for those that want even more. Backwards compatibility with old
3643 spool files has been maintained. However, going back to a previous Exim
3644 release will lost any variables that are in spool files.
3646 PH/07 Two small changes when running in the test harness: increase delay when
3647 passing a TCP/IP connection to a new process, in case the original
3648 process has to generate a bounce, and remove special handling of
3649 127.0.0.2 (sic), which is no longer necessary.
3651 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3652 be the same on different OS.
3654 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3657 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3658 whether --show-vars was specified or not
3660 JJ/02 exipick: Added support for new ACL variable spool format introduced
3663 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3664 syntactically invalid From: or Reply-to: line, and a filter used this to
3665 generate an autoreply, and therefore failed to obtain an address for the
3666 autoreply, Exim could try to deliver to a non-existent relative file
3667 name, causing unrelated and misleading errors. What now happens is that
3668 it logs this as a hard delivery error, but does not attempt to create a
3671 PH/11 The exinext utility has a -C option for testing purposes, but although
3672 the given file was scanned by exinext itself; it wasn't being passed on
3673 when Exim was called.
3675 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3676 an end-of-file indication when reading a command response.
3678 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3679 compiled. In many other places in Exim, IPv6 addresses are always
3680 recognized, so I have changed this. It also means that IPv4 domain
3681 literals of the form [IPV4:n.n.n.n] are now always recognized.
3683 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3684 used if the router is not running as root, for example, when verifying at
3685 ACL time, or when using -bh. The debugging output from this situation was
3686 non-existent - all you got was a failure to exec. I have made two
3689 (a) Failures to set uid/gid, the current directory, or a process leader
3690 in a subprocess such as that created by queryprogram now generate
3691 suitable debugging output when -d is set.
3693 (b) The queryprogram router detects when it is not running as root,
3694 outputs suitable debugging information if -d is set, and then runs
3695 the subprocess without attempting to change uid/gid.
3697 PH/15 Minor change to Makefile for building test_host (undocumented testing
3700 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3701 additional section of a DNS packet that returns MX or SRV records.
3702 Instead, it always explicitly searches for A/AAAA records. This avoids
3703 major problems that occur when a DNS server includes only records of one
3704 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3705 fixed another bug: if SRV records were looked up and the corresponding
3706 address records were *not* found in the additional section, the port
3707 values from the SRV records were lost.
3709 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3710 using the correct key (the original address) when searching the retry
3711 rules in order to find which one to use for generating the retry hint.
3713 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3714 adding the default one. Similarly, if it contains a Reply-To: header, the
3715 errors_reply_to option, if set, is not used.
3717 PH/19 When calculating a retry time, Exim used to measure the "time since
3718 failure" by looking at the "first failed" field in the retry record. Now
3719 it does not use this if it is later than than the arrival time of the
3720 message. Instead it uses the arrival time. This makes for better
3721 behaviour in cases where some deliveries succeed, thus re-setting the
3722 "first failed" field. An example is a quota failure for a huge message
3723 when small messages continue to be delivered. Without this change, the
3724 "time since failure" will always be short, possible causing more frequent
3725 delivery attempts for the huge message than are intended.
3726 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3728 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3729 $1, $2, $3) because the numerical variables can be reset during some
3730 expansion items (e.g. "match"), thereby losing the authentication data.
3732 PH/21 Make -bV show the size of off_t variables so that the test suite can
3733 decide whether to run tests for quotas > 2G.
3735 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3736 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3737 filecount value is greater than 2G or if a quota value is greater than 2G
3738 on a system where the size of off_t is not greater than 4, a panic error
3741 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3742 never match. The debug and -bh output now contains an explicit error
3743 message indicating a malformed IPv4 address or mask.
3745 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3746 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3747 PH/23 above applies.
3749 PH/25 Do not write to syslog when running in the test harness. The only
3750 occasion when this arises is a failure to open the main or panic logs
3751 (for which there is an explicit test).
3753 PH/26 Added the /no_tell option to "control=freeze".
3755 PH/27 If a host name lookup failed very early in a connection, for example, if
3756 the IP address matched host_lookup and the reverse lookup yielded a name
3757 that did not have a forward lookup, an error message of the form "no IP
3758 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3759 could be logged. Now it outputs the IP address instead of "NULL".
3761 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3762 allows the sender and the authenticated sender to be set when
3763 submitting a message from within Exim. Since child_open_exim() is
3764 documented for local_scan(), the new function should be too.
3766 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3767 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3768 results in an empty string is now treated as unset.
3770 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3772 PH/31 Added sender_verify_fail logging option.
3774 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3775 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3776 tidied the source and removed it altogether.
3778 PH/33 When a queue run was abandoned because the load average was too high, a
3779 log line was always written; now it is written only if the queue_run log
3780 selector is set. In addition, the log line for abandonment now contains
3781 information about the queue run such as the pid. This is always present
3782 in "start" and "stop" lines but was omitted from the "abandon" line.
3784 PH/34 Omit spaces between a header name and the colon in the error message that
3785 is given when verify = headers_syntax fails (if there are lots of them,
3786 the message gets confusing).
3788 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3789 names, as there are now some PTR records that contain slashes. This check
3790 is only to protect against broken name servers that fall over on strange
3791 characters, so the fact that it applies to all lookups doesn't matter.
3793 PH/36 Now that the new test suite is complete, we can remove some of the
3794 special code in Exim that was needed for the old test suite. For example,
3795 sorting DNS records because real resolvers return them in an arbitrary
3796 order. The new test suite's fake resolver always returns records in the
3799 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3800 bounces) except when queue_only is set, to avoid logging races between
3801 the different processes.
3803 PH/38 Panic-die if .include specifies a non-absolute path.
3805 PH/39 A tweak to the "H" retry rule from its user.
3807 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3808 a label. They prevented compilation on older perls.
3810 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3811 a warning to be raised on newish perls.
3813 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3814 on queue. Changes to match documented behaviour of showing count of
3815 messages matching specified criteria.
3817 PH/40 Changed the default ident timeout from 30s to 5s.
3819 PH/41 Added support for the use of login_cap features, on those BSD systems
3820 that have them, for controlling the resources used by pipe deliveries.
3822 PH/42 The content-scanning code uses fopen() to create files in which to put
3823 message data. Previously it was not paying any attention to the mode of
3824 the files. Exim runs with umask(0) because the rest of the code creates
3825 files with open(), and sets the required mode explicitly. Thus, these
3826 files were ending up world-writeable. This was not a big issue, because,
3827 being within the spool directory, they were not world-accessible. I have
3828 created a function called modefopen, which takes an additional mode
3829 argument. It sets umask(777), creates the file, chmods it to the required
3830 mode, then resets the umask. All the relevant calls to fopen() in the
3831 content scanning code have been changed to use this function.
3833 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3834 to 24 hours. This avoids potential overflow problems when processing G
3835 and H retry rules. I suspect nobody ever tinkers with this value.
3837 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3839 PH/45 When the plaintext authenticator is running as a client, the server's
3840 challenges are checked to ensure they are valid base64 strings. By
3841 default, the authentication attempt is cancelled if an invalid string is
3842 received. Setting client_ignore_invalid_base64 true ignores these errors.
3843 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3844 they are received. Thus, the responses can be made to depend on the
3845 challenges. If an invalid string is ignored, an empty string is placed in
3848 PH/46 Messages that are created by the autoreply transport now contains a
3849 References: header, in accordance with RFCs 2822 and 3834.
3851 PH/47 Added authenticated_sender_force to the smtp transport.
3853 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3855 PH/49 Installed latest patch from the Sieve maintainer.
3857 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3858 was also set, the appendfile transport was unnecessarily scanning a
3859 directory of message files (e.g. for maildir delivery) to find the count
3860 of files (along with the size), even though it did not need this
3861 information. It now does the scan only if it needs to find either the
3862 size of the count of files.
3864 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3866 PH/52 Two bugs concerned with error handling when the smtp transport is
3869 (i) Exim was not creating retry information for temporary errors given
3870 for individual recipients after the DATA command when the smtp transport
3871 was used in LMTP mode. This meant that they could be retried too
3872 frequently, and not timed out correctly.
3874 (ii) Exim was setting the flag that allows error details to be returned
3875 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3876 recipients that were returned after the DATA command.
3878 PH/53 This is related to PH/52, but is more general: for any failing address,
3879 when detailed error information was permitted to be returned to the
3880 sender, but the error was temporary, then after the final timeout, only
3881 "retry timeout exceeded" was returned. Now it returns the full error as
3882 well as "retry timeout exceeded".
3884 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3885 do this, and (what is worse) MTAs that accept it.
3887 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3888 will now be deprecated.
3890 PH/56 New os.c-cygwin from the Cygwin maintainer.
3892 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3893 formats (previously only available in exim formats via -bpr, -bpru,
3894 and -bpra. Now also available in native and exiqgrep formats)
3896 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3897 with very large, slow to parse queues
3899 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3901 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3903 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3904 responses to authentication challenges, though it was showing the
3905 challenges; (ii) I've removed the CR characters from the debug output for
3908 PH/58 Allow for the insertion of a newline as well as a space when a string
3909 is turned into more than one encoded-word during RFC 2047 encoding. The
3910 Sieve code now uses this.
3912 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3913 data_4xx, lost_connection, tls_required.
3915 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3916 message was being sent as an SMTP response.
3918 PH/61 Add -l and -k options to exicyclog.
3920 PH/62 When verifying, if an address was redirected to one new address, so that
3921 verification continued, and the new address failed or deferred after
3922 having set something in $address_data, the value of $address_data was not
3923 passed back to the ACL. This was different to the case when no
3924 redirection occurred. The value is now passed back in both cases.
3926 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3927 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3928 use login_cap.h, so on its own it isn't the distinguishing feature. The
3929 new name refers directly to the setclassresources() function.
3931 PH/65 Added configuration files for NetBSD3.
3933 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3935 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3936 is preferred over IPv4.
3938 PH/68 The bounce_return_message and bounce_return_body options were not being
3939 honoured for bounces generated during the reception of non-SMTP messages.
3940 In particular, this applied to messages rejected by the ACL. This bug has
3941 been fixed. However, if bounce_return_message is true and bounce_return_
3942 body is false, the headers that are returned for a non-SMTP message
3943 include only those that have been read before the error was detected.
3944 (In the case of an ACL rejection, they have all been read.)
3946 PH/69 The HTML version of the specification is now built in a directory called
3947 spec_html instead of spec.html, because the latter looks like a path with
3948 a MIME-type, and this confuses some software.
3950 PH/70 Catch two compiler warnings in sieve.c.
3952 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3953 function verify_get_ident() calls ip_connect() to connect a socket, but
3954 if the "connect()" function timed out, ip_connect() used to close the
3955 socket. However, verify_get_ident() also closes the socket later, and in
3956 between Exim writes to the log, which may get opened at this point. When
3957 the socket was closed in ip_connect(), the log could get the same file
3958 descriptor number as the socket. This naturally causes chaos. The fix is
3959 not to close the socket in ip_connect(); the socket should be closed by
3960 the function that creates it. There was only one place in the code where
3961 this was missing, in the iplookup router, which I don't think anybody now
3962 uses, but I've fixed it anyway.
3964 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3965 well as to direct DNS lookups. Otherwise the handling of names in host
3966 lists is inconsistent and therefore confusing.
3972 PH/01 Two changes to the default runtime configuration:
3974 (1) Move the checks for relay_from_hosts and authenticated clients from
3975 after to before the (commented out) DNS black list checks.
3977 (2) Add control=submission to the relay_from_hosts and authenticated
3978 clients checks, on the grounds that messages accepted by these
3979 statements are most likely to be submissions.
3981 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3983 (1) Generate an error if the third argument for the ${prvs expansion is
3986 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3989 (3) Reset the variables that are obtained from the first argument of
3990 ${prvscheck and used in the second argument before leaving the code,
3991 because their memory is reclaimed, so using them afterwards may do
3994 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3995 one (it's much easier than Tom thought :-).
3997 (5) Because of (4), we can now allow for the use of $prvscheck_result
3998 inside the third argument.
4000 PH/03 For some reason, the default setting of PATH when running a command from
4001 a pipe transport was just "/usr/bin". I have changed it to
4004 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4005 anything to be listed in the output from -bV.
4007 PH/05 When a filter generated an autoreply, the entire To: header line was
4008 quoted in the delivery log line, like this:
4010 => >A.N.Other <ano@some.domain> <original@ddress> ...
4012 This has been changed so that it extracts the operative address. There
4013 may be more than one such address. If so, they are comma-separated, like
4016 => >ano@some.domain,ona@other.domain <original@ddress> ...
4018 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4019 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4020 not being looked up in the rDNS to get a host name, Exim was showing the
4021 IP address twice in Received: lines, even though the IP addresses were
4022 identical. For example:
4024 Received: from [1.2.3.4] (helo=[1.2.3.4])
4026 However, if the real host name was known, it was omitting the HELO data
4027 if it matched the actual IP address. This has been tidied up so that it
4028 doesn't show the same IP address twice.
4030 PH/07 When both +timestamp and +memory debugging was on, the value given by
4031 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4032 called by the debug printing, thereby overwriting the timestamp buffer.
4033 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4035 PH/08 When the original message was included in an autoreply transport, it
4036 always said "this is a copy of the message, including all the headers",
4037 even if body_only or headers_only was set. It now gives an appropriate
4040 PH/09 Applied a patch from the Sieve maintainer which:
4042 o fixes some comments
4043 o adds the (disabled) notify extension core
4044 o adds some debug output for the result of if/elsif tests
4045 o points to the current vacation draft in the documentation
4046 and documents the missing references header update
4050 o fixes a bug in processing the envelope test (when testing
4051 multiple envelope elements, the last element determined the
4054 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4055 Electronic Mail") by including:
4057 Auto-submitted: auto-generated
4059 in the messages that it generates (bounce messages and others, such as
4060 warnings). In the case of bounce messages for non-SMTP messages, there was
4061 also a typo: it was using "Auto_submitted" (underscore instead of
4062 hyphen). Since every message generated by Exim is necessarily in response
4063 to another message, thes have all been changed to:
4065 Auto-Submitted: auto-replied
4067 in accordance with these statements in the RFC:
4069 The auto-replied keyword:
4071 - SHOULD be used on messages sent in direct response to another
4072 message by an automatic process,
4074 - MUST NOT be used on manually-generated messages,
4076 - MAY be used on Delivery Status Notifications (DSNs) and Message
4077 Disposition Notifications (MDNs),
4079 - MUST NOT be used on messages generated by automatic or periodic
4080 processes, except for messages which are automatic responses to
4083 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4084 to the default Received: header definition.
4086 PH/12 Added log selector acl_warn_skipped (default on).
4088 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4089 variables because (a) they are in the wrong storage pool and (b) even if
4090 they were copied, it wouldn't work properly because of the caching.
4092 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4093 checking when decoding. Apparently there are clients that generate
4094 overlong encoded strings. Why am I not surprised?
4096 PH/15 If the first argument of "${if match_address" was not empty, but did not
4097 contain an "@" character, Exim crashed. Now it writes a panic log message
4098 and treats the condition as false.
4100 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4102 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4103 extension "envelope-auth". The code is finished and in agreement with
4104 other implementations, but there is no documentation so far and in fact,
4105 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4106 not changing the active code.
4108 Print executed "if" and "elsif" statements when debugging is used. This
4109 helps a great deal to understand what a filter does.
4111 Document more things not specified clearly in RFC3028. I had all this
4112 sorted out, when out of a sudden new issues came to my mind. Oops."
4114 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4117 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4118 canonical form (as documented). However, after a host name lookup from
4119 the IP address, check_host() was doing a simple string comparison with
4120 addresses acquired from the DNS when checking that the found name did
4121 have the original IP as one of its addresses. Since any found IPv6
4122 addresses are likely to be in abbreviated form, the comparison could
4123 fail. Luckily, there already exists a function for doing the comparison
4124 by converting both addresses to binary, so now that is used instead of
4125 the text comparison.
4127 PH/20 There was another similar case to PH/19, when a complete host name was
4128 given in a host list; looking up its IP address could give an abbreviated
4129 form, whereas the current host's name might or might not be abbreviated.
4130 The same fix has been applied.
4136 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4137 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4140 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4141 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4143 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4145 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4146 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4147 Also, for "Auto-Submitted", treat anything other than "no" as
4148 non-personal, in accordance with RFC 3834. (Previously it treated
4149 anything starting "auto-" as non-personal.)
4151 TF/01 The control=submission/name=... option had a problem with syntax
4152 errors if the name included a slash character. The /name= option
4153 now slurps the rest of the string, so it can include any characters
4154 but it must come last in the list of options (after /sender_retain
4157 PH/05 Some modifications to the interface to the fake nameserver for the new
4165 TK/01 Added the "success_on_redirect" address verification option. See
4166 NewStuff for rationale and an example.
4168 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4170 PH/02 Patch to exigrep to allow it to work on syslog lines.
4172 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4173 fread() to skip over the body file's header line, because in Cygwin the
4174 header line is locked and is inaccessible.
4176 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4177 co-exist for some time) to make it clear that it is the Exim ID that is
4178 referenced, not the Message-ID: header line.
4180 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4181 string_format() function, because snprintf() does not exist on all
4184 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4185 ${stat: expansion item.
4187 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4188 protocol synchronization error", to keep the pedants happy.
4190 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4191 well as for IRIX systems, when gcc is being used. See the host.c source
4194 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4196 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4199 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4200 transport and to the smtp transport in LMTP mode.
4202 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4204 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4205 run a filter in a subprocess. This could lead to confusion in subsequent
4206 lookups in the parent process. There should also be a search_tidyup() at
4207 the end of the subprocess.
4209 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4210 only if the host matched helo_try_verify_hosts, which caused the
4211 verification to occur when the EHLO/HELO command was issued. The ACL just
4212 tested the remembered result. Now, if a previous verification attempt has
4213 not happened, "verify = helo" does it there and then.
4215 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4217 TK/03 Fix log output including CR from clamd.
4219 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4220 exist provoked a memory error which could cause a segfault.
4222 PH/15 Installed PCRE 6.2
4224 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4226 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4227 of the problem. Specifically, suggested +O2 rather than +O1 for the
4230 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4232 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4233 record, the port was not correctly logged when the outgoing_port log
4234 selector was set (it logged the transort's default port).
4236 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4237 fallback_hosts, and "hosts" in the smtp transport.
4239 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4240 host errors such as "Connection refused".
4242 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4243 authentication with radiusclient 0.4.9:
4245 - Error returned from rc_read_config was caught wrongly
4246 - Username/password not passed on to radius server due to wrong length.
4248 The presumption is that some radiusclient API changes for 4.51/PH/17
4249 were not taken care of correctly. The code is still untested by me (my
4250 Linux distribution still has 0.3.2 of radiusclient), but it was
4251 contributed by a Radius user.
4253 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4254 expanding the "port" option of the smtp transport.
4256 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4257 while reading a MIME header. Thanks to Tom Hughes for a patch.
4259 PH/24 Include config.h inside local_scan.h so that configuration settings are
4262 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4263 that in an ACL for RCPT (for example), you can examine exactly what was
4266 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4267 commands, but it was not correctly comparing the address with the actual
4268 client host address. Thus, it would show the EHLO address in Received:
4269 header lines when this was not necessary.
4271 PH/27 Added the % operator to ${eval:}.
4273 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4274 it should be ignoring failures (because with -C, for example, it has lost
4275 privilege). It wasn't ignoring creation failures other than "already
4278 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4281 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4282 because an input line was too long, either on its own, or by virtue of
4283 too many continuations, the temporary file was not being removed, and the
4284 return code was incorrect.
4286 PH/31 Missing "BOOL" in function definition in filtertest.c.
4288 PH/32 Applied Sieve patches from the maintainer.
4290 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4292 PH/33 Added "verify = not_blind".
4294 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4295 Local/Makefile (with some defaults set). These are used in built scripts
4296 such as exicyclog, but they have never been used in the exim_install
4297 script (though there are many overriding facilities there). I have
4298 arranged that the exim_install script now takes note of these two
4301 PH/35 Installed configuration files for Dragonfly.
4303 PH/36 When a locally submitted message by a trusted user did not contain a
4304 From: header, and the sender address was obtained from -f or from an SMTP
4305 MAIL command, and the trusted user did not use -F to supply a sender
4306 name, $originator_name was incorrectly used when constructing a From:
4307 header. Furthermore, $originator_name was used for submission mode
4308 messages from external hosts without From: headers in a similar way,
4309 which is clearly wrong.
4311 PH/37 Added control=suppress_local_fixups.
4313 PH/38 When log_selector = +received_sender was set, and the addition of the
4314 sender made the log line's construction buffer exactly full, or one byte
4315 less than full, an overflow happened when the terminating "\n" was
4318 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4319 when the result of a list match is failure because a DNS lookup failed.
4321 PH/40 RM_COMMAND is now used in the building process.
4323 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4324 the "build-* directories that it finds.
4326 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4327 address in a domain literal was a prefix of an interface address.
4329 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4330 when verifying a sender address, unless rewrite_headers is false.
4332 PH/44 Wrote a long comment about why errors_to addresses are verified as
4333 recipients, not senders.
4335 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4336 the ratelimit ACL was added.
4338 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4340 PH/46 Added extra information about PostgreSQL errors to the error string.
4342 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4343 avoiding the need to install special zones in a real server. This is
4344 backwards compatible; if it can't find the fake resolver, it drops back.
4345 Thus, both old and new test suites can be run.
4347 TF/02 Added util/ratelimit.pl
4349 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4350 clock is set back in time.
4352 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4353 Candler <B.Candler@pobox.com>.
4355 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4356 OK for addresses that are the result of redirections.
4358 PH/48 A number of further additions for the benefit of the new test suite,
4359 including a fake gethostbyname() that interfaces to the fake DNS resolver
4362 TF/06 The fix for widen_domains has also been applied to qualify_single and
4363 search_parents which are the other dnslookup options that can cause
4366 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4369 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4371 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4372 service" if it encountered an MX record with an empty target hostname.
4373 The message is now "an MX or SRV record indicated no SMTP service".
4375 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4376 if the DNS of the sending site is misconfigured. This is quite a
4377 common situation. This change restores the behaviour of treating a
4378 helo verification defer as a failure.
4380 PH/51 If self=fail was set on a router, the bounce message did not include the
4381 actual error message.
4387 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4389 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4390 command itself ended in error, the underlying message about the transport
4391 filter timeout was being overwritten with the pipe command error. Now the
4392 underlying error message should be appended to the second error message.
4394 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4396 PH/02 Reduce the amount of output that "make" produces by default. Full output
4397 can still be requested.
4399 PH/03 The warning log line about a condition test deferring for a "warn" verb
4400 was being output only once per connection, rather than after each
4401 occurrence (because it was using the same function as for successful
4402 "warn" verbs). This seems wrong, so I have changed it.
4404 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4405 it should not have, which might have caused a crash in the right
4406 circumstances, but probably never did.
4408 PH/04 Installed a modified version of Tony Finch's patch to make submission
4409 mode fix the return path as well as the Sender: header line, and to
4410 add a /name= option so that you can make the user's friendly name appear
4413 TF/03 Added the control = fakedefer ACL modifier.
4415 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4416 Mark Lowes for thorough testing.
4418 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4420 TK/03 Merged latest SRS patch from Miles Wilton.
4422 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4423 with the definition in sysexits.h (which is #included earlier).
4424 Fortunately, Exim does not actually use EX_OK. The code used to try to
4425 preserve the sysexits.h value, by assuming that macro definitions were
4426 scanned for macro replacements. I have been disabused of this notion,
4427 so now the code just undefines EX_OK before #including unistd.h.
4429 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4430 in the smtp transport. When a block could not be written in a single
4431 write() function, the timeout was being re-applied to each part-write.
4432 This seems wrong - if the receiver was accepting one byte at a time it
4433 would take for ever. The timeout is now adjusted when this happens. It
4434 doesn't have to be particularly precise.
4436 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4437 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4439 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4440 without a domain if the check to <postmaster@domain> fails.
4442 SC/01 Eximstats: added -xls and the ability to specify output files
4443 (patch written by Frank Heydlauf).
4445 SC/02 Eximstats: use FileHandles for outputting results.
4447 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4449 SC/04 Eximstats: fixed display of large numbers with -nvr option
4451 SC/05 Eximstats: fixed merging of reports with empty tables.
4453 SC/06 Eximstats: added the -include_original_destination flag
4455 SC/07 Eximstats: removed tabs and trailing whitespace.
4457 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4459 TK/06 MBOX spool code: Add real "From " MBOX separator line
4460 so the .eml file is really in mbox format (even though
4461 most programs do not really care). Patch from Alex Miller.
4463 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4464 The latter is generated from $received_to and is only set if the
4465 message has one envelope recipient. SA can use these headers,
4466 obviously out-of-the-box. Patch from Alex Miller.
4468 PH/08 The ${def test on a variable was returning false if the variable's
4469 value was "0", contrary to what the specification has always said!
4470 The result should be true unless the variable is empty.
4472 PH/09 The syntax error of a character other than { following "${if
4473 def:variable_name" (after optional whitespace) was not being diagnosed.
4474 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4475 accidental colon was present, for example, could give incorrect results.
4477 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4478 result is used (not including appendfile, where other changes are about
4481 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4482 This involved changing a lot of size variables from int to off_t. It
4483 should work with maildirs and everything.
4485 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4486 spamd dying while we are connected to it.
4488 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4491 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4492 function for BDB 4.3.
4494 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4496 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4497 the sources for PCRE have been re-arranged and more files are now
4500 PH/15 The code I had for printing potentially long long variables in PH/11
4501 above was not the best (it lost precision). The length of off_t variables
4502 is now inspected at build time, and an appropriate printing format (%ld
4503 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4504 to be "long long int" or "long int". This is needed for the internal
4505 formatting function string_vformat().
4507 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4508 the configuration file to be ":syslog", then the script "guesses" where
4509 the logs files are, rather than using the compiled in default. In our
4510 case the guess is not the same as the compiled default, so the script
4511 suddenly stopped working when I started to use syslog. The patch checks
4512 to see if log_file_path is "". If so, it attempts to read it from exim
4513 with no configuration file to get the compiled in version, before it
4514 falls back to the previous guessing code."
4516 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4517 implementing BATV in an Exim configuration. See NewStuff for the gory
4520 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4521 Makefile that are specific to HP-UX.
4523 PH/18 If the "use_postmaster" option was set for a recipient callout together
4524 with the "random" option, the postmaster address was used as the MAIL
4525 FROM address for the random test, but not for the subsequent recipient
4526 test. It is now used for both.
4528 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4529 patch removes a few documentation additions to RFC 3028, because the
4530 latest draft now contains them. It adds the new en;ascii-case comparator
4531 and a new error check for 8bit text in MIME parts. Comparator and
4532 require names are now matched exactly. I enabled the subaddress
4533 extension, but it is not well tested yet (read: it works for me)."
4535 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4536 rework some of the code of TK/09 above to avoid the hardwired use of
4537 "%lld" and "long long". Replaced the call to snprintf() with a call to
4540 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4541 records point to non-existent hosts", "retry timeout exceeded", and
4542 "retry time not reached for any host after a long failure period".
4544 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4545 experimental DomainKeys support:
4547 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4548 (2) On an error such as an illegally used "control", the wrong name for
4549 the control was given.
4551 These problems did NOT occur unless DomainKeys support was compiled.
4553 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4555 PH/24 Added ${if match_ip condition.
4557 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4558 databases so that it will be absolutely obvious if a crash occurs in the
4559 DB library. This is a regular occurrence (often caused by mis-matched
4562 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4563 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4564 picked up on a user's system that detects such things. There doesn't seem
4565 to be a gcc warning option for this - only an attribute that has to be
4566 put on the function's prototype. It seems that in Fedora Core 4 they have
4567 set this on a number of new functions. No doubt there will be more in due
4570 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4571 specify a transport. However, if an address that was verified by such a
4572 router was the subject of a callout, Exim crashed because it tried to
4573 read the rcpt_include_affixes from the non-existent transport. Now it
4574 just assumes that the setting of that option is false. This bug was
4575 introduced by 4.51/PH/31.
4577 PH/28 Changed -d+all to exclude +memory, because that information is very
4578 rarely of interest, but it makes the output a lot bigger. People tend to
4579 do -d+all out of habit.
4581 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4582 code in os-type was giving problems when libc.so lives in lib64, like on
4585 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4586 aren't the modern standard, and it seems that some systems' include files
4587 don't always have them. Exim was already checking for some of the newer
4588 ones like T_AAAA, and defining it itself. I've added checks for all the
4589 record types that Exim uses.
4591 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4592 not automatically generating a new one, as it is supposed to. This
4593 prevented TLS from working. If the file did exist, but contained invalid
4594 data, a new version was generated, as expected. It was only the case of a
4595 non-existent file that was broken.
4597 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4598 with a change in libdomainkeys > 0.64.
4600 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4601 from DNS. If the selector record carries the flag, it now has
4602 precedence over the domain-wide flag.
4604 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4606 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4607 the use of a port name that isn't defined in /etc/services) occurred, the
4608 message was deferred as in a normal delivery, and thus remained on the
4609 spool, instead of being failed because of the mua_wrapper setting. This
4610 is now fixed, and I tidied up some of the mua_wrapper messages at the
4613 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4614 the messages in a hash of arrays rather than using individual hashes.
4615 This is a bit cleaner and results in dramatic memory savings, albeit
4616 at a slight CPU cost.
4618 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4619 as requested by Marc Sherman.
4621 SC/10 Eximstats: added histograms for user specified patterns as requested
4624 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4626 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4627 fopen() in the content-scanning modules that did not already have it.
4633 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4634 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4636 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4638 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4640 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4641 to test Sieve filters that use "vacation".
4643 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4644 that changes the way the GnuTLS parameters are stored in the cache file.
4645 The new format can be generated externally. For backward compatibility,
4646 if the data in the cache doesn't make sense, Exim assumes it has read an
4647 old-format file, and it generates new data and writes a new file. This
4648 means that you can't go back to an older release without removing the
4651 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4652 work if there are any delivery delays because "one_time" forces the
4653 parent to be marked "delivered", so its unseen clone is never tried
4654 again. For this reason, Exim now forbids the simultaneous setting of
4657 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4658 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4659 entry if you want to know the details. The fix, however, neglected to
4660 consider the case when local delivery batching is involved. The test for
4661 "previously delivered" was not happening when checking to see if an
4662 address could be batched with a previous (undelivered) one; under
4663 certain circumstances this could lead to multiple deliveries to the same
4666 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4667 in its include files, and this causes problems building Exim.
4669 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4670 header_syntax) but Exim was just ignoring anything given after a slash.
4671 In particular, this caused confusion with an attempt to use "verify =
4672 reverse_host_lookup/defer_ok". An error is now given when options are
4673 supplied for verify items that do not have them. (Maybe reverse_host_
4674 lookup should have a defer_ok option, but that's a different point.)
4676 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4677 defined by RFC 821) to 2048, because there were problems with some AUTH
4678 commands, and RFC 1869 says the size should be increased for extended
4679 SMTP commands that take arguments.
4681 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4684 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4685 "unknown" error; now it says that the functionality isn't in the binary.
4687 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4688 an address' error message when a string expansion fails (syntax or
4689 whatever). Otherwise the password may appear in the log. Following change
4690 PH/42 below, there is no longer a chance of it appearing in a bounce
4693 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4695 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4696 of its addresses was ever tried. (Bugzilla bug #2.)
4698 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4699 the result incorrectly in the debug output. (It correctly added a newline
4700 to what was transported.)
4702 TF/01 Added $received_time.
4704 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4705 commented out examples of how to interface to a virus scanner and to
4706 SpamAssassin. Also added commented examples of av_scanner and
4707 spamd_address settings.
4709 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4710 and controls are allowed in which ACLs. There were a couple of minor
4711 errors. Some of the entries in the conditions table (which is a table of
4712 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4713 negation of where the condition IS allowed.
4715 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4717 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4718 header file does not have a version number, so I've had to invent a new
4719 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4720 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4721 radiusclient), but it was contributed by a Radius user.
4723 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4724 files or directories, for OpenSSL.
4726 PH/19 When an Exim process that is running as root has to create an Exim log
4727 file, it does so in a subprocess that runs as exim:exim so as to get the
4728 ownership right at creation (otherwise, other Exim processes might see
4729 the file with the wrong ownership). There was no test for failure of this
4730 fork() call, which would lead to the process getting stuck as it waited
4731 for a non-existent subprocess. Forks do occasionally fail when resources
4732 run out. I reviewed all the other calls to fork(); they all seem to check
4735 PH/20 When checking for unexpected SMTP input at connect time (before writing
4736 the banner), Exim was not dealing correctly with a non-positive return
4737 from the read() function. If the client had disconnected by this time,
4738 the result was a log entry for a synchronization error with an empty
4739 string after "input=" when read() returned zero. If read() returned -1
4740 (an event I could not check), uninitialized data bytes were printed.
4741 There were reports of junk text (parts of files, etc) appearing after
4744 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4746 PH/22 Added support for macro redefinition, and (re)definition in between
4747 driver and ACL definitions.
4749 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4750 forgetting to use the resulting value; it was using the unexpanded value.
4752 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4753 hadn't been configured. The fix is from Juergen Kreileder, who
4754 understands it better than I do:
4756 "Here's what I see happening with three configured cyrus_sasl
4757 authenticators configured (plain, login, cram-md5):
4759 On startup auth_cyrus_sasl_init() gets called for each of these.
4760 This means three calls to sasl_listmech() without a specified mech_list.
4761 => SASL tests which mechs of all available mechs actually work
4762 => three warnings about OTP not working
4763 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4765 With the patch, sasl_listmech() also gets called three times. But now
4766 SASL's mech_list option is set to the server_mech specified in the the
4767 authenticator. Or in other words, the answer from sasl_listmech()
4768 gets limited to just the mech you're testing for (which is different
4770 => the return list contains just 'plain' or 'login', 'cram-md5' or
4771 nothing depending on the value of ob->server_mech.
4773 I've just tested the patch: Authentication still works fine,
4774 unavailable mechs specified in the exim configuration are still
4775 caught, and the auth.log warnings about OTP are gone."
4777 PH/25 When debugging is enabled, the contents of the command line are added
4778 to the debugging output, even when log_selector=+arguments is not
4781 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4782 answer is "GNU", and only if the return is "GNU/something" is the answer
4785 PH/27 $acl_verify_message is now set immediately after the failure of a
4786 verification in an ACL, and so is available in subsequent modifiers. In
4787 particular, the message can be preserved by coding like this:
4789 warn !verify = sender
4790 set acl_m0 = $acl_verify_message
4792 Previously, $acl_verify_message was set only while expanding "message"
4793 and "log_message" when a very denied access.
4795 PH/28 Modified OS/os.c-Linux with
4797 -#ifndef OS_LOAD_AVERAGE
4798 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4800 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4801 nomenclature these days.)
4803 PH/29 Installed patch from the Sieve maintainer that adds the options
4804 sieve_useraddress and sieve_subaddress to the redirect router.
4806 PH/30 In these circumstances:
4807 . Two addresses routed to the same list of hosts;
4808 . First host does not offer TLS;
4809 . First host accepts first address;
4810 . First host gives temporary error to second address;
4811 . Second host offers TLS and a TLS session is established;
4812 . Second host accepts second address.
4813 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4814 and peerdn, if requested) that were in fact used only for the second
4817 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4818 attention to any local part prefix or suffix that was matched by the
4819 router that accepted the address. It now behaves in the same way as it
4820 does for delivery: the affixes are removed from the local part unless
4821 rcpt_include_affixes is set on the transport.
4823 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4824 timeout during the DATA phase of an incoming message.
4826 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4827 applied a patch sanctioned by the Sieve maintainer.
4829 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4830 the uid or gid is negative. A case of a negative gid caused this to be
4831 noticed. The fix allows for either to be negative.
4833 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4834 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4837 PH/36 The change PH/12 above was broken. Fixed it.
4839 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4840 the grounds that routing the same address twice would always produce the
4841 same answer. This might have been true once, but it is certainly no
4842 longer true now. Routing a child address may depend on the previous
4843 routing that produced that child. Some complicated redirection strategies
4844 went wrong when messages had multiple recipients, and made Exim's
4845 behaviour dependent on the order in which the addresses were given.
4847 I have moved the duplicate checking until after the routing is complete.
4848 Exim scans the addresses that are assigned to local and remote
4849 transports, and removes any duplicates. This means that more work will be
4850 done, as duplicates will always all be routed, but duplicates are
4851 presumably rare, so I don't expect this is of any significance.
4853 For deliveries to pipes, files, and autoreplies, the duplicate checking
4854 still happens during the routing process, since they are not going to be
4857 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4858 It corrects a timeout issue with spamd. This is Ian's comment: "The
4859 background is that sometimes spamd either never reads data from a
4860 connection it has accepted, or it never writes response data. The exiscan
4861 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4862 blindly assumes that writes won't block so it may never time out."
4864 PH/39 Allow G after quota size as well as K and M.
4866 PH/40 The value set for $authenticated_id in an authenticator may not contain
4867 binary zeroes or newlines because the value is written to log lines and
4868 to spool files. There was no check on this. Now the value is run through
4869 the string_printing() function so that such characters are converted to
4870 printable escape sequences.
4872 PH/41 $message_linecount is a new variable that contains the total number of
4873 lines in the message. Compare $body_linecount, which is the count for the
4876 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4877 bounce and delay warning messages, except in certain special cases, which
4880 (a) An SMTP error message from a remote host;
4881 (b) A message specified in a :fail: redirection;
4882 (c) A message specified in a "fail" command in a system filter;
4883 (d) A message specified in a FAIL return from the queryprogram router;
4884 (e) A message specified by the cannot_route_message router option.
4886 In these cases only, Exim does include the error details in bounce and
4887 warning messages. There are also a few cases where bland messages such
4888 as "unrouteable address" or "local delivery error" are given.
4890 PH/43 $value is now also set for the "else" part of a ${run expansion.
4892 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4893 being worked on, but at least Exim now implements the latest version to
4896 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4897 process to complete was treated as a delivery failure, a timeout while
4898 writing the message to the pipe was logged, but erroneously treated as a
4899 successful delivery. Such timeouts include transport filter timeouts. For
4900 consistency with the overall process timeout, these timeouts are now
4901 treated as errors, giving rise to delivery failures by default. However,
4902 there is now a new Boolean option for the pipe transport called
4903 timeout_defer, which, if set TRUE, converts the failures into defers for
4904 both kinds of timeout. A transport filter timeout is now identified in
4907 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4908 systems where "make" and "gmake" are different, calling "gmake" at top
4909 level broke things. I've arranged for the value of $(MAKE) to be passed
4910 from the Makefile to this script so that it can call the same version of
4914 A note about Exim versions 4.44 and 4.50
4915 ----------------------------------------
4917 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4918 changes of various kinds. As a consequence, a big documentation update was
4919 needed. This delayed the release for rather longer than seemed good, especially
4920 in the light of a couple of (minor) security issues. Therefore, the changes
4921 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4922 release. So 4.44 and 4.50 are in effect two different branches that both start
4925 I have left the 4.50 change log unchanged; it contains all the changes since
4926 4.43. The change log for 4.44 is below; many of its items are identical to
4927 those for 4.50. This seems to be the most sensible way to preserve the
4928 historical information.
4934 1. Minor wording change to the doc/README.SIEVE file.
4936 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4937 computation of the current number of files was incorrect.
4939 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4940 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4943 4. Give more explanation in the error message when the command for a transport
4944 filter fails to execute.
4946 5. There are several places where Exim runs a non-Exim command in a
4947 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4948 was being done only for the command run by the queryprogram router. It is
4949 now done for all such subprocesses. The other cases are: ${run, transport
4950 filters, and the commands run by the lmtp and pipe transports.
4952 6. Added CONFIGURE_GROUP build-time option.
4954 7. Some older OS have a limit of 256 on the maximum number of file
4955 descriptors. Exim was using setrlimit() to set 1000 as a large value
4956 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4957 systems. I've change it so that if it can't get 1000, it tries for 256.
4959 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4960 was an oversight, and furthermore, ever since the addition of extra
4961 controls (e.g. 4.43/32), the checks on when to allow different forms of
4962 "control" were broken. There should now be diagnostics for all cases when a
4963 control that does not make sense is encountered.
4965 9. Added the /retain_sender option to "control=submission".
4967 10. $recipients is now available in the predata ACL (oversight).
4969 11. Tidy the search cache before the fork to do a delivery from a message
4970 received from the command line. Otherwise the child will trigger a lookup
4971 failure and thereby defer the delivery if it tries to use (for example) a
4972 cached ldap connection that the parent has called unbind on.
4974 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4975 of $address_data from the recipient verification was clobbered by the
4976 sender verification.
4978 13. The value of address_data from a sender verification is now available in
4979 $sender_address_data in subsequent conditions in the ACL statement.
4981 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4983 15. Added a new option "connect=<time>" to callout options, to set a different
4986 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4987 was its contents. (It was OK if the option was not defined at all.)
4989 17. A "Completed" log line is now written for messages that are removed from
4990 the spool by the -Mrm option.
4992 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4993 information about exactly what failed.
4995 19. Added -dd to debug only the daemon process.
4997 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4998 handles timeouts, both on the server side and network timeouts. Renamed the
4999 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5001 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5002 It is now set to "smtps".
5004 22. $host_address is now set to the target address during the checking of
5005 ignore_target_hosts.
5007 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5008 being passed; this would have caused $sender_host_name to have been used if
5009 matching the list had actually called for a host name (not very likely,
5010 since this list is usually IP addresses). A host name is now passed as
5013 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5014 code that specifies a non-restarting handler (typically sigaction() in
5015 modern systems) in an attempt to fix a rare and obscure crash bug.
5017 25. Narrowed the window for a race in the daemon that could cause it to ignore
5018 SIGCHLD signals. This is not a major problem, because they are used only to
5019 wake it up if nothing else does.
5021 26. A malformed maildirsize file could cause Exim to calculate negative values
5022 for the mailbox size or file count. Odd effects could occur as a result.
5023 The maildirsize information is now recalculated if the size or filecount
5026 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5027 support for a long time. Removed HAVE_SYS_VFS_H.
5029 28. Installed the latest version of exipick from John Jetmore.
5031 29. In an address list, if the pattern was not a regular expression, an empty
5032 subject address (from a bounce message) matched only if the pattern was an
5033 empty string. Non-empty patterns were not even tested. This was the wrong
5034 because it is perfectly reasonable to use an empty address as part of a
5035 database query. An empty address is now tested by patterns that are
5036 lookups. However, all the other forms of pattern expect the subject to
5037 contain a local part and a domain, and therefore, for them, an empty
5038 address still always fails if the pattern is not itself empty.
5040 30. Exim went into a mad DNS loop when attempting to do a callout where the
5041 host was specified on an smtp transport, and looking it up yielded more
5042 than one IP address.
5044 31. Re-factored the code for checking spool and log partition space into a
5045 function that finds that data and another that does the check. The former
5046 is then used to implement four new variables: $spool_space, $log_space,
5047 $spool_inodes, and $log_inodes.
5049 32. The RFC2047 encoding function was originally intended for short strings
5050 such as real names; it was not keeping to the 75-character limit for
5051 encoded words that the RFC imposes. It now respects the limit, and
5052 generates multiple encoded words if necessary. To be on the safe side, I
5053 have increased the buffer size for the ${rfc2047: expansion operator from
5056 33. It is now permitted to omit both strings after an "if" condition; if the
5057 condition is true, the result is "true". As before, when the second string
5058 is omitted, a false condition yields an empty string. This makes it less
5059 cumbersome to write custom ACL and router conditions.
5061 34. Failure to deliver a bounce message always caused it to be frozen, even if
5062 there was an errors_to setting on the router. The errors_to setting is now
5065 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5066 canonical form (fully expanded) before being placed in
5067 $sender_host_address.
5069 36. The table in the code that translates DNS record types into text (T_A to
5070 "A" for instance) was missing entries for NS and CNAME. It is just possible
5071 that this could have caused confusion if both these types were looked up
5072 for the same domain, because the text type is used as part of Exim's
5073 per-process caching. But the chance of anyone hitting this buglet seems
5076 37. The dnsdb lookup has been extended in a number of ways.
5078 (1) There is a new type, "zns", which walks up the domain tree until it
5079 finds some nameserver records. It should be used with care.
5081 (2) There is a new type, "mxh", which is like "mx" except that it returns
5082 just the host names, not the priorities.
5084 (3) It is now possible to give a list of domains (or IP addresses) to be
5085 looked up. The behaviour when one of the lookups defers can be
5086 controlled by a keyword.
5088 (4) It is now possible to specify the separator character for use when
5089 multiple records are returned.
5091 38. The dnslists ACL condition has been extended: it is now possible to supply
5092 a list of IP addresses and/or domains to be looked up in a particular DNS
5095 39. Added log_selector=+queue_time_overall.
5097 40. When running the queue in the test harness, wait just a tad after forking a
5098 delivery process, to get repeatability of debugging output.
5100 41. Include certificate and key file names in error message when GnuTLS fails
5101 to set them up, because the GnuTLS error message doesn't include the name
5102 of the failing file when there is a problem reading it.
5104 42. Allow both -bf and -bF in the same test run.
5106 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5108 44. The "Exiscan patch" is now merged into the mainline Exim source.
5110 45. Sometimes the final signoff response after QUIT could fail to get
5111 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5112 < 0 before doing a fflush(). This bug looks as though it goes back to the
5113 introduction of TLS in release 3.20, but "sometimes" must have been rare
5114 because the tests only now provoked it.
5116 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5117 (this can affect the format of dates).
5119 47. exim_tidydb, when checking for the continued existence of a message for
5120 which it has found a message-specific retry record, was not finding
5121 messages that were in split spool directories. Consequently, it was
5122 deleting retry records that should have stayed in existence.
5124 48. Steve fixed some bugs in eximstats.
5126 49. The SPA authentication driver was not abandoning authentication and moving
5127 on to the next authenticator when an expansion was forced to fail,
5128 contradicting the general specification for all authenticators. Instead it
5129 was generating a temporary error. It now behaves as specified.
5131 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5132 (the order specifies the preference for clients). The order is now AES256,
5133 AES128, 3DES, ARCFOUR128.
5135 51. Small patch to Sieve code - explicitly set From: when generating an
5138 52. Exim crashed if a remote delivery caused a very long error message to be
5139 recorded - for instance if somebody sent an entire SpamAssassin report back
5140 as a large number of 550 error lines. This bug was coincidentally fixed by
5141 increasing the size of one of Exim's internal buffers (big_buffer) that
5142 happened as part of the Exiscan merge. However, to be on the safe side, I
5143 have made the code more robust (and fixed the comments that describe what
5146 53. Now that there can be additional text after "Completed" in log lines (if
5147 the queue_time_overall log selector is set), a one-byte patch to exigrep
5148 was needed to allow it to recognize "Completed" as not the last thing in
5151 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5152 patch that reportedly fixes this has been added. I am not expert enough to
5153 create a test for it. This is what the patch creator wrote:
5155 "I found a little strange behaviour of ldap code when working with
5156 Windows 2003 AD Domain, where users was placed in more than one
5157 Organization Units. When I tried to give exim partial DN, the exit code
5158 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5159 But simultaneously result of request was absolutely normal ldap result,
5160 so I produce this patch..."
5162 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5163 REFERENCE, so I have modified the code to exclude the patch when that macro
5166 55. Some experimental protocols are using DNS PTR records for new purposes. The
5167 keys for these records are domain names, not reversed IP addresses. The
5168 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5169 leaves it alone. Component reversal etc. now happens only for IP addresses.
5172 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5174 57. Double the size of the debug message buffer (to 2048) so that more of very
5175 long debug lines gets shown.
5177 58. The exicyclog utility now does better if the number of log files to keep
5178 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5180 59. Two changes related to the smtp_active_hostname option:
5182 (1) $smtp_active_hostname is now available as a variable.
5183 (2) The default for smtp_banner uses $smtp_active_hostname instead
5184 of $primary_hostname.
5186 60. The host_aton() function is supposed to be passed a string that is known
5187 to be a valid IP address. However, in the case of IPv6 addresses, it was
5188 not checking this. This is a hostage to fortune. Exim now panics and dies
5189 if the condition is not met. A case was found where this could be provoked
5190 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5191 components; fortuitously, this particular loophole had already been fixed
5192 by change 4.50/55 above.
5194 If there are any other similar loopholes, the new check in host_aton()
5195 itself should stop them being exploited. The report I received stated that
5196 data on the command line could provoke the exploit when Exim was running as
5197 exim, but did not say which command line option was involved. All I could
5198 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5199 running as the user.
5202 61. There was a buffer overflow vulnerability in the SPA authentication code
5203 (which came originally from the Samba project). I have added a test to the
5204 spa_base64_to_bits() function which I hope fixes it.
5207 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5208 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5210 63. The daemon start-up calls getloadavg() while still root for those OS that
5211 need the first call to be done as root, but it missed one case: when
5212 deliver_queue_load_max is set with deliver_drop_privilege. This is
5213 necessary for the benefit of the queue runner, because there is no re-exec
5214 when deliver_drop_privilege is set.
5216 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5217 This has been fixed.
5219 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5220 was in use, was not putting the data itself into the right store pool;
5221 consequently, it could be overwritten for a subsequent message in the same
5222 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5225 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5227 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5228 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5229 the function were treating the return as a boolean value, which happened to
5230 work because 0=false and not-0=true, but is not correct code.
5232 68. The host_aton() function was not handling scoped IPv6 addresses (those
5233 with, for example, "%eth0" on the end) correctly.
5235 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5236 negated items (that is, ~something) in unsigned ints. Some compilers
5237 apparently mutter when there is no cast.
5239 70. If an address verification called from an ACL failed, and did not produce a
5240 user-specific message (i.e. there was only a "system" message), nothing was
5241 put in $acl_verify_message. In this situation, it now puts the system
5244 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5245 session; change 4.31/43 added the unwanted input to the log line - except
5246 that it did not do this in the start of session case. It now does.
5248 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5249 This is wrong and can cause the other end to generate a synchronization
5250 error if it is another Exim or anything else that does the synchronization
5251 check. A QUIT command is no longer sent after a timeout.
5253 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5254 during host lookups.
5256 74. The defer_ok option of callout verification was not working if it was used
5257 when verifying addresses in header lines, that is, for this case:
5259 verify = header_sender/callout=defer_ok
5261 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5262 those file descriptors could be used for SMTP connections. If anything
5263 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5264 could be sent to the SMTP client, causing chaos. The daemon now opens
5265 stdin, stdout, and stderr to /dev/null when it puts itself into the
5268 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5269 log by default. The user can override this with suitable Perl magic.
5271 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5272 the log message when discard triggers, was not working for the DATA ACL or
5273 for the non-SMTP ACL.
5275 78. Error message wording change in sieve.c.
5277 79. If smtp_accept_max_per_host was set, the number of connections could be
5278 restricted to fewer than expected, because the daemon was trying to set up
5279 a new connection before checking whether the processes handling previous
5280 connections had finished. The check for completed processes is now done
5281 earlier. On busy systems, this bug wouldn't be noticed because something
5282 else would have woken the daemon, and it would have reaped the completed
5285 80. If a message was submitted locally by a user whose login name contained one
5286 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5287 It caused a spool format error. I have fixed the spool reading code. A
5288 related problem was that the "from" clause in the Received: line became
5289 illegal because of the space(s). It is now covered by ${quote_local_part.
5291 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5294 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5296 83. Patch from Sieve maintainer to fix unterminated string problem in
5297 "vacation" handling.
5299 84. Some minor changes to the Linux configuration files to help with other
5300 OS variants using glibc.
5302 85. One more patch for Sieve to update vacation handling to latest spec.
5305 ----------------------------------------------------
5306 See the note above about the 4.44 and 4.50 releases.
5307 ----------------------------------------------------
5313 1. Change 4.43/35 introduced a bug that caused file counts to be
5314 incorrectly computed when quota_filecount was set in an appendfile
5317 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5318 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5321 3. Give more explanation in the error message when the command for a transport
5322 filter fails to execute.
5324 4. There are several places where Exim runs a non-Exim command in a
5325 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5326 was being done only for the command run by the queryprogram router. It is
5327 now done for all such subprocesses. The other cases are: ${run, transport
5328 filters, and the commands run by the lmtp and pipe transports.
5330 5. Some older OS have a limit of 256 on the maximum number of file
5331 descriptors. Exim was using setrlimit() to set 1000 as a large value
5332 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5333 systems. I've change it so that if it can't get 1000, it tries for 256.
5335 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5336 was an oversight, and furthermore, ever since the addition of extra
5337 controls (e.g. 4.43/32), the checks on when to allow different forms of
5338 "control" were broken. There should now be diagnostics for all cases when a
5339 control that does not make sense is encountered.
5341 7. $recipients is now available in the predata ACL (oversight).
5343 8. Tidy the search cache before the fork to do a delivery from a message
5344 received from the command line. Otherwise the child will trigger a lookup
5345 failure and thereby defer the delivery if it tries to use (for example) a
5346 cached ldap connection that the parent has called unbind on.
5348 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5349 of $address_data from the recipient verification was clobbered by the
5350 sender verification.
5352 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5353 was its contents. (It was OK if the option was not defined at all.)
5355 11. A "Completed" log line is now written for messages that are removed from
5356 the spool by the -Mrm option.
5358 12. $host_address is now set to the target address during the checking of
5359 ignore_target_hosts.
5361 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5362 being passed; this would have caused $sender_host_name to have been used if
5363 matching the list had actually called for a host name (not very likely,
5364 since this list is usually IP addresses). A host name is now passed as
5367 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5368 code that specifies a non-restarting handler (typically sigaction() in
5369 modern systems) in an attempt to fix a rare and obscure crash bug.
5371 15. Narrowed the window for a race in the daemon that could cause it to ignore
5372 SIGCHLD signals. This is not a major problem, because they are used only to
5373 wake it up if nothing else does.
5375 16. A malformed maildirsize file could cause Exim to calculate negative values
5376 for the mailbox size or file count. Odd effects could occur as a result.
5377 The maildirsize information is now recalculated if the size or filecount
5380 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5381 support for a long time. Removed HAVE_SYS_VFS_H.
5383 18. Updated exipick to current release from John Jetmore.
5385 19. Allow an empty sender to be matched against a lookup in an address list.
5386 Previously the only cases considered were a regular expression, or an
5389 20. Exim went into a mad DNS lookup loop when doing a callout where the
5390 host was specified on the transport, if the DNS lookup yielded more than
5393 21. The RFC2047 encoding function was originally intended for short strings
5394 such as real names; it was not keeping to the 75-character limit for
5395 encoded words that the RFC imposes. It now respects the limit, and
5396 generates multiple encoded words if necessary. To be on the safe side, I
5397 have increased the buffer size for the ${rfc2047: expansion operator from
5400 22. Failure to deliver a bounce message always caused it to be frozen, even if
5401 there was an errors_to setting on the router. The errors_to setting is now
5404 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5405 canonical form (fully expanded) before being placed in
5406 $sender_host_address.
5408 24. Updated eximstats to version 1.33
5410 25. Include certificate and key file names in error message when GnuTLS fails
5411 to set them up, because the GnuTLS error message doesn't include the name
5412 of the failing file when there is a problem reading it.
5414 26. Expand error message when OpenSSL has problems setting up cert/key files.
5417 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5418 (this can affect the format of dates).
5420 28. exim_tidydb, when checking for the continued existence of a message for
5421 which it has found a message-specific retry record, was not finding
5422 messages that were in split spool directories. Consequently, it was
5423 deleting retry records that should have stayed in existence.
5425 29. eximstats updated to version 1.35
5426 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5427 1.35 - bugfix such that pie charts by volume are generated correctly
5429 30. The SPA authentication driver was not abandoning authentication and moving
5430 on to the next authenticator when an expansion was forced to fail,
5431 contradicting the general specification for all authenticators. Instead it
5432 was generating a temporary error. It now behaves as specified.
5434 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5435 (the order specifies the preference for clients). The order is now AES256,
5436 AES128, 3DES, ARCFOUR128.
5438 31. Small patch to Sieve code - explicitly set From: when generating an
5441 32. Exim crashed if a remote delivery caused a very long error message to be
5442 recorded - for instance if somebody sent an entire SpamAssassin report back
5443 as a large number of 550 error lines. This bug was coincidentally fixed by
5444 increasing the size of one of Exim's internal buffers (big_buffer) that
5445 happened as part of the Exiscan merge. However, to be on the safe side, I
5446 have made the code more robust (and fixed the comments that describe what
5449 33. Some experimental protocols are using DNS PTR records for new purposes. The
5450 keys for these records are domain names, not reversed IP addresses. The
5451 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5452 leaves it alone. Component reversal etc. now happens only for IP addresses.
5455 34. The host_aton() function is supposed to be passed a string that is known
5456 to be a valid IP address. However, in the case of IPv6 addresses, it was
5457 not checking this. This is a hostage to fortune. Exim now panics and dies
5458 if the condition is not met. A case was found where this could be provoked
5459 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5460 components; fortuitously, this particular loophole had already been fixed
5461 by change 4.50/55 or 4.44/33 above.
5463 If there are any other similar loopholes, the new check in host_aton()
5464 itself should stop them being exploited. The report I received stated that
5465 data on the command line could provoke the exploit when Exim was running as
5466 exim, but did not say which command line option was involved. All I could
5467 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5468 running as the user.
5471 35. There was a buffer overflow vulnerability in the SPA authentication code
5472 (which came originally from the Samba project). I have added a test to the
5473 spa_base64_to_bits() function which I hope fixes it.
5476 36. The daemon start-up calls getloadavg() while still root for those OS that
5477 need the first call to be done as root, but it missed one case: when
5478 deliver_queue_load_max is set with deliver_drop_privilege. This is
5479 necessary for the benefit of the queue runner, because there is no re-exec
5480 when deliver_drop_privilege is set.
5482 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5483 was in use, was not putting the data itself into the right store pool;
5484 consequently, it could be overwritten for a subsequent message in the same
5485 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5488 38. Sometimes the final signoff response after QUIT could fail to get
5489 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5490 < 0 before doing a fflush(). This bug looks as though it goes back to the
5491 introduction of TLS in release 3.20, but "sometimes" must have been rare
5492 because the tests only now provoked it.
5498 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5499 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5500 Now it returns an integer. A number of calls were still expecting a T/F
5501 return. Fortuitously, in all cases, the tests worked in OK situations,
5502 which is the norm. However, things would have gone wrong on any write
5503 failures on the smtp file descriptor. This function is used when sending
5504 messages over SMTP and also when doing verify callouts.
5506 2. When Exim is called to do synchronous delivery of a locally submitted
5507 message (the -odf or -odi options), it no longer closes stderr before doing
5510 3. Implemented the mua_wrapper option.
5512 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5514 5. Implemented the functions header_remove(), header_testname(),
5515 header_add_at_position(), and receive_remove_recipient(), and exported them
5518 6. If an ACL "warn" statement specified the addition of headers, Exim already
5519 inserted X-ACL-Warn: at the start if there was no header name. However, it
5520 was not making this test for the second and subsequent header lines if
5521 there were newlines in the string. This meant that an invalid header could
5522 be inserted if Exim was badly configured.
5524 7. Allow an ACL "warn" statement to add header lines at the start or after all
5525 the Received: headers, as well as at the end.
5527 8. Added the rcpt_4xx retry error code.
5529 9. Added postmaster_mailfrom=xxx to callout verification option.
5531 10. Added mailfrom=xxxx to the callout verification option, for verify=
5534 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5535 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5537 12. Inserted some casts to stop certain compilers warning when using pointer
5538 differences as field lengths or precisions in printf-type calls (mostly
5539 affecting debugging statements).
5541 13. Added optional readline() support for -be (dynamically loaded).
5543 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5544 same clock tick as a message's arrival, so that its received time was the
5545 same as the "first fail" time on the retry record, and that message
5546 remained on the queue past the ultimate address timeout, every queue runner
5547 would try a delivery (because it was past the ultimate address timeout) but
5548 after another failure, the ultimate address timeout, which should have then
5549 bounced the address, did not kick in. This was a "< instead of <=" error;
5550 in most cases the first failure would have been in the next clock tick
5551 after the received time, and all would be well.
5553 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5554 being recognized when the domain list was tested by the match_domain
5555 condition in an expansion string.
5557 16. Added the ${str2b64: operator.
5559 17. Exim was always calling setrlimit() to set a large limit for the number of
5560 processes, without checking whether the existing limit was already
5561 adequate. (It did check for the limit on file descriptors.) Furthermore,
5562 errors from getrlimit() and setrlimit() were being ignored. Now they are
5563 logged to the main and panic logs, but Exim does carry on, to try to do its
5564 job under whatever limits there are.
5566 18. Imported PCRE 5.0.
5568 19. Trivial typo in log message " temporarily refused connection" (the leading
5571 20. If the log selector return_path_on_delivery was set and an address was
5572 redirected to /dev/null, the delivery process crashed because it assumed
5573 that a return path would always be set for a "successful" delivery. In this
5574 case, the whole delivery is bypassed as an optimization, and therefore no
5577 21. Internal re-arrangement: the function for sending a challenge and reading
5578 a response while authentication was assuming a zero-terminated challenge
5579 string. It's now changed to take a pointer and a length, to allow for
5580 binary data in such strings.
5582 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5584 23. Exim was not respecting finduser_retries when seeking the login of the
5585 uid under which it was called; it was always trying 10 times. (The default
5586 setting of finduser_retries is zero.) Also, it was sleeping after the final
5587 failure, which is pointless.
5589 24. Implemented tls_on_connect_ports.
5591 25. Implemented acl_smtp_predata.
5593 26. If the domain in control=submission is set empty, Exim assumes that the
5594 authenticated id is a complete email address when it generates From: or
5595 Sender: header lines.
5597 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5598 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5599 chown and chgrp in /bin and hostname in /usr/bin.
5601 28. Exim was keeping the "process log" file open after each use, just as it
5602 does for the main log. This opens the possibility of it remaining open for
5603 long periods when the USR1 signal hits a daemon. Occasional processlog
5604 errors were reported, that could have been caused by this. Anyway, it seems
5605 much more sensible not to leave this file open at all, so that is what now
5608 29. The long-running daemon process does not normally write to the log once it
5609 has entered its main loop, and it closes the log before doing so. This is
5610 so that log files can straightforwardly be renamed and moved. However,
5611 there are a couple of unusual error situations where the daemon does write
5612 log entries, and I had neglected to close the log afterwards.
5614 30. The text of an SMTP error response that was received during a remote
5615 delivery was being truncated at 512 bytes. This is too short for some of
5616 the long messages that one sometimes sees. I've increased the limit to
5619 31. It is now possible to make retry rules that apply only when a message has a
5620 specific sender, in particular, an empty sender.
5622 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5623 it possible to be selective about when SMTP synchronization is enforced.
5625 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5627 32. Implemented hosts_connection_nolog.
5629 33. Added an ACL for QUIT.
5631 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5634 35. Added mailbox_size and mailbox_filecount to appendfile.
5636 36. Added control = no_multiline_responses to ACLs.
5638 37. There was a bug in the logic of the code that waits for the clock to tick
5639 in the case where the clock went backwards by a substantial amount such
5640 that the microsecond fraction of "now" was more than the microsecond
5641 fraction of "then" (but the whole seconds number was less).
5643 38. Added support for the libradius Radius client library this is found on
5644 FreeBSD (previously only the radiusclient library was supported).
5650 1. When certain lookups returned multiple values in the form name=value, the
5651 quoting of the values was not always being done properly. Specifically:
5652 (a) If the value started with a double quote, but contained no whitespace,
5654 (b) If the value contained whitespace other than a space character (i.e.
5655 tabs or newlines or carriage returns) it was not quoted.
5656 This fix has been applied to the mysql and pgsql lookups by writing a
5657 separate quoting function and calling it from the lookup code. The fix
5658 should probably also be applied to nisplus, ibase and oracle lookups, but
5659 since I cannot test any of those, I have not disturbed their existing code.
5661 2. A hit in the callout cache for a specific address caused a log line with no
5662 reason for rejecting RCPT. Now it says "Previous (cached) callout
5663 verification failure".
5665 3. There was an off-by-one bug in the queryprogram router. An over-long
5666 return line was truncated at 256 instead of 255 characters, thereby
5667 overflowing its buffer with the terminating zero. As well as fixing this, I
5668 have increased the buffer size to 1024 (and made a note to document this).
5670 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5671 when Exim is waiting for an SMTP response from a remote server, Exim
5672 restarts its select() call on the socket, thereby resetting its timeout.
5673 This is not a problem when such interrupts are rare. Somebody set up a cron
5674 job to run exiwhat every 2 minutes, which is less than the normal select()
5675 timeout (5 or 10 minutes). This meant that the select() timeout never
5676 kicked in because it was always reset. I have fixed this by comparing the
5677 time when an interrupt arrives with the time at the start of the first call
5678 to select(). If more time than the timeout has elapsed, the interrupt is
5679 treated as a timeout.
5681 5. Some internal re-factoring in preparation for the addition of Sieve
5682 extensions (by MH). In particular, the "personal" test is moved to a
5683 separate function, and given an option for scanning Cc: and Bcc: (which is
5684 not set for Exim filters).
5686 6. When Exim created an email address using the login of the caller as the
5687 local part (e.g. when creating a From: or Sender: header line), it was not
5688 quoting the local part when it contained special characters such as @.
5690 7. Installed new OpenBSD configuration files.
5692 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5693 try to make them clearer.
5695 9. Callout options, other than the timeout value, were being ignored when
5696 verifying sender addresses in header lines. For example, when using
5698 verify = header_sender/callout=no_cache
5700 the cache was (incorrectly) being used.
5702 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5703 only the Cygwin environment.
5705 11. When return_path_on_delivery was set as a log selector, if different remote
5706 addresses in the same message used different return paths and parallel
5707 remote delivery occurred, the wrong values would sometimes be logged.
5708 (Whenever a remote delivery process finished, the return path value from
5709 the most recently started remote delivery process was logged.)
5711 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5712 lines when AUTH and/or TLS are in use. This is the "received protocol"
5713 field. Exim used to use "asmtp" for authenticated SMTP, without any
5714 indication (in the protocol name) for TLS use. Now it follows the RFC and
5715 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5716 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5717 names appear in log lines as well as in Received: header lines.
5719 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5720 extensions, and comparison tests, and to fix some bugs.
5722 14. Changes to the "personal" filter test:
5724 (1) The test was buggy in that it was just doing the equivalent of
5725 "contains" tests on header lines. For example, if a user's address was
5726 anne@some.where, the "personal" test would incorrectly be true for
5728 To: susanne@some.where
5730 This test is now done by extracting each address from the header in turn,
5731 and checking the entire address. Other tests that are part of "personal"
5732 are now done using regular expressions (for example, to check local parts
5733 of addresses in From: header lines).
5735 (2) The list of non-personal local parts in From: addresses has been
5736 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5737 taken from the Sieve specification recommendations.
5739 (3) If the message contains any header line starting with "List-" it is
5740 treated as non-personal.
5742 (4) The test for "circular" in the Subject: header line has been removed
5743 because it now seems ill-conceived.
5745 15. Minor typos in src/EDITME comments corrected.
5747 16. Installed latest exipick from John Jetmore.
5749 17. If headers_add on a router specified a text string that was too long for
5750 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5751 of string_sprintf() is now avoided.
5753 18. $message_body_size was not set (it was always zero) when running the DATA
5754 ACL and the local_scan() function.
5756 19. For the "mail" command in an Exim filter, no default was being set for
5757 the once_repeat time, causing a random time value to be used if "once" was
5758 specified. (If the value happened to be <= 0, no repeat happened.) The
5759 default is now 0s, meaning "never repeat". The "vacation" command was OK
5760 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5761 (I found it when inspecting the code).
5763 20. There is now an overall timeout for performing a callout verification. It
5764 defaults to 4 times the callout timeout, which applies to individual SMTP
5765 commands during the callout. The overall timeout applies when there is more
5766 than one host that can be tried. The timeout is checked before trying the
5767 next host. This prevents very long delays if there are a large number of
5768 hosts and all are timing out (e.g. when the network connections are timing
5769 out). The value of the overall timeout can be changed by specifying an
5770 additional sub-option for "callout", called "maxwait". For example:
5772 verify = sender/callout=5s,maxwait=20s
5774 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5775 to the end before writing, but this should make it even safer).
5777 22. Exim was forgetting that it had advertised PIPELINING for the second and
5778 subsequent messages on an SMTP connection. It was also not resetting its
5779 memory on STARTTLS and an internal HELO.
5781 23. When Exim logs an SMTP synchronization error within a session, it now
5782 records whether PIPELINING has been advertised or not.
5784 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5785 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5786 rather than long int.
5788 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5790 26. Added the never_mail option to autoreply.
5796 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5797 crash if the getsockname() call failed; this can happen if a connection is
5798 closed very soon after it is established. The problem was simply in the
5799 order in which certain operations were done, causing Exim to try to write
5800 to the SMTP stream before it had set up the file descriptor. The bug has
5801 been fixed by making things happen in the correct order.
5807 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5808 before the connection was closed, thus losing the rejection response.
5810 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5811 some early Solaris releases, but causes trouble in current releases where
5812 socklen_t is defined.
5814 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5817 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5820 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5821 that releasing the top of it at the end releases what was used for sub-
5822 expansions (unless the block got too big). However, discard this block if
5823 the first thing is a variable or header, so that we can use its block when
5824 it is dynamic (useful for very large $message_headers, for example).
5826 6. Lookups now cache *every* query, not just the most recent. A new, separate
5827 store pool is used for this. It can be recovered when all lookup caches are
5828 flushed. Lookups now release memory at the end of their result strings.
5829 This has involved some general refactoring of the lookup sources.
5831 7. Some code has been added to the store_xxx() functions to reduce the amount
5832 of flapping under certain conditions.
5834 8. log_incoming_interface used to affect only the <= reception log lines. Now
5835 it causes the local interface and port to be added to several more SMTP log
5836 lines, for example "SMTP connection from", and rejection lines.
5838 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5840 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5842 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5843 could be overwritten at the end of the current message (or the start of a
5844 new message if it was set in a HELO ACL). The value is now preserved for
5845 the duration of the SMTP connection.
5847 12. If a transport had a headers_rewrite setting, and a matching header line
5848 contained an unqualified address, that address was qualified, even if it
5849 did not match any rewriting rules. The underlying bug was that the values
5850 of the flags that permit the existence of unqualified sender and recipient
5851 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5852 non-local messages, and by -bnq for local messages) were not being
5853 preserved with the message after it was received.
5855 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5856 "next input=" as part of the text comprising the host identity instead of
5857 the correct text. The code was using the same buffer for two different
5858 strings. However, depending on which order the printing function evaluated
5859 its arguments, the bug did not always show up. Under Linux, for example, my
5860 test suite worked just fine.
5862 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5863 doesn't work with some older versions of Perl. It has been changed to "my",
5864 which in any case is probably the better facility to use.
5866 15. A really picky compiler found some instances of statements for creating
5867 error messages that either had too many or two few arguments for the format
5870 16. The size of the buffer for calls to the DNS resolver has been increased
5871 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5872 for addresses that have a lot of PTR records. This alleviates a problem; it
5873 does not fully solve it.
5875 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5876 buffer now truncates the list and logs the incident, which is the same
5877 action as happens when Exim is looking up a host name and its aliases.
5878 Previously in this situation something unpredictable would happen;
5879 sometimes it was "internal error: store_reset failed".
5881 18. If a server dropped the connection unexpectedly when an Exim client was
5882 using GnuTLS and trying to read a response, the client delivery process
5883 crashed while trying to generate an error log message.
5885 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5886 string, for example:
5888 warn message = H1: something\nH2: something
5890 the text was added as a single header line from Exim's point of view
5891 though it ended up OK in the delivered message. However, searching for the
5892 second and subsequent header lines using $h_h2: did not work. This has been
5893 fixed. Similarly, if a system filter added multiple headers in this way,
5894 the routers could not see them.
5896 20. Expanded the error message when iplsearch is called with an invalid key to
5897 suggest using net-iplsearch in a host list.
5899 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5900 ACLs are no longer actually imposed (and a message to that effect is
5903 22. If a "gecos" field in a passwd entry contained escaped characters, in
5904 particular, if it contained a \" sequence, Exim got it wrong when building
5905 a From: or a Sender: header from that name. A second bug also caused
5906 incorrect handling when an unquoted " was present following a character
5907 that needed quoting.
5909 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5910 was not being matched caselessly.
5912 24. Arranged for all hyphens in the exim.8 source to be escaped with
5915 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5916 properly. Recipient callout cache records were still being keyed to include
5917 the sender, even when use_sender was set false. This led to far more
5918 callouts that were necessary. The sender is no longer included in the key
5919 when use_sender is false.
5921 26. Added "control = submission" modifier to ACLs.
5923 27. Added the ${base62d: operator to decode base 62 numbers.
5925 28. dnsdb lookups can now access SRV records.
5927 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5928 the configuration file.
5930 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5931 (-v) mode. This makes the output for a verbose queue run more intelligible.
5933 31. Added a use_postmaster feature to recipient callouts.
5935 32. Added the $body_zerocount variable, containing the number of binary zero
5936 bytes in the message body.
5938 33. The time of last modification of the "new" subdirectory is now used as the
5939 "mailbox time last read" when there is a quota error for a maildir
5942 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5944 35. Added +ignore_unknown as a special item in host lists.
5946 36. Code for decoding IPv6 addresses in host lists is now included, even if
5947 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5948 address was recognized as an IP address, but was then not correctly decoded
5949 into binary, causing unexpected and incorrect effects when compared with
5956 1. Very minor rewording of debugging text in manualroute to say "list of
5957 hosts" instead of "hostlist".
5959 2. If verify=header_syntax was set, and a header line with an unqualified
5960 address (no domain) and a large number of spaces between the end of the
5961 name and the colon was received, the reception process suffered a buffer
5962 overflow, and (when I tested it) crashed. This was caused by some obsolete
5963 code that should have been removed. The fix is to remove it!
5965 3. When running in the test harness, delay a bit after writing a bounce
5966 message to get a bit more predictability in the log output.
5968 4. Added a call to search_tidyup() just before forking a reception process. In
5969 theory, someone could use a lookup in the expansion of smtp_accept_max_
5970 per_host which, without the tidyup, could leave open a database connection.
5972 5. Added the variables $recipient_data and $sender_data which get set from a
5973 lookup success in an ACL "recipients" or "senders" condition, or a router
5974 "senders" option, similar to $domain_data and $local_part_data.
5976 6. Moved the writing of debug_print from before to after the "senders" test
5979 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5980 problems for message scanning, either using a data ACL, or using
5981 local_scan() because the Received: header was not generated till after they
5982 were called (in order to set the time as the time of reception completion).
5983 I have revised the way this works. The header is now generated after the
5984 body is received, but before the ACL or local_scan() are called. After they
5985 are run, the timestamp in the header is updated.
5991 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5992 before starting a queue runner without re-exec. This happened only when
5993 deliver_drop_privilege was set or when the Exim user was set to root. The
5994 effect of the bug was that timeouts during subsequent deliveries caused
5995 crashes instead of being properly handled. The handler is now left at its
5996 default (and expected) setting.
5998 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5999 message, again when deliver_drop_privilege is set or Exim is run as root.
6000 The bug described in (1) was not present in this case, but the tidying up
6001 of the other signals was missing. I have made the two cases consistent.
6003 3. The ignore_target_hosts setting on a manualroute router was being ignored
6004 for hosts that were looked up using the /MX notation.
6006 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6009 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6010 operated on the sender address. After changing the $sender_address to <>
6011 for the sender address verify, Exim was re-instated it as the original
6012 (before rewriting) address, but remembering that it had rewritten it, so it
6013 wasn't rewriting it again. This bug also had the effect of breaking the
6014 sender address verification caching when the sender address was rewritten.
6016 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6017 This has been changed so that if the ip literal address matches
6018 ignore_target_hosts, the router declines.
6020 7. Added expansion conditions match_domain, match_address, and match_local_
6021 part (NOT match_host).
6023 8. The placeholder for the Received: header didn't have a length field set.
6025 9. Added code to Exim itself and to exim_lock to test for a specific race
6026 condition that could lead to file corruption when using MBX delivery. The
6027 issue is with the lockfile that is created in /tmp. If this file is removed
6028 after a process has opened it but before that process has acquired a lock,
6029 there is the potential for a second process to recreate the file and also
6030 acquire a lock. This could lead to two Exim processes writing to the file
6031 at the same time. The added code performs the same test as UW imapd; it
6032 checks after acquiring the lock that its file descriptor still refers to
6033 the same named file.
6035 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6036 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6039 11. Added the smtp_active_hostname option. If used, this will typically be made
6040 to depend on the incoming interface address. Because $interface_address is
6041 not set up until the daemon has forked a reception process, error responses
6042 that can happen earlier (such as "too many connections") no longer contain
6045 12. If an expansion in a condition on a "warn" statement fails because a lookup
6046 defers, the "warn" statement is abandoned, and the next ACL statement is
6047 processed. Previously this caused the whole ACL to be aborted.
6049 13. Added the iplsearch lookup type.
6051 14. Added ident_timeout as a log selector.
6053 15. Added tls_certificate_verified as a log selector.
6055 16. Added a global option tls_require_ciphers (compare the smtp transport
6056 option of the same name). This controls incoming TLS connections.
6058 17. I finally figured out how to make tls_require_ciphers do a similar thing
6059 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6060 before starting the TLS session.
6062 18. Tabs are now shown as \t in -bP output.
6064 19. If the log selector return_path_on_delivery was set, Exim crashed when
6065 bouncing a message because it had too many Received: header lines.
6067 20. If two routers both had headers_remove settings, and the first one included
6068 a superfluous trailing colon, the final name in the first list and the
6069 first name in the second list were incorrectly joined into one item (with a
6070 colon in the middle).
6076 1. Added -C and -D options to the exinext utility, mainly to make it easier
6077 to include in the automated testing, but these could be helpful when
6078 multiple configurations are in use.
6080 2. The exinext utility was not formatting the output nicely when there was
6081 an alternate port involved in the retry record key, nor when there was a
6082 message id as well (for retries that were specific to a specific message
6083 and a specific host). It was also confused by IPv6 addresses, because of
6084 the additional colons they contain. I have fixed the IPv4 problem, and
6085 patched it up to do a reasonable job for IPv6.
6087 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6088 delivery, the log line now contains "pipelined" if PIPELINING was used.
6090 4. An SMTP transport process used to panic and die if the bind() call to set
6091 an explicit outgoing interface failed. This has been changed; it is now
6092 treated in the same way as a connect() failure.
6094 5. A reference to $sender_host_name in the part of a conditional expansion
6095 that was being skipped was still causing a DNS lookup. This no longer
6098 6. The def: expansion condition was not recognizing references to header lines
6099 that used bh_ and bheader_.
6101 7. Added the _cache feature to named lists.
6103 8. The code for checking quota_filecount in the appendfile transport was
6104 allowing one more file than it should have been.
6106 9. For compatibility with Sendmail, the command line option
6114 and sets the incoming protocol and host name (for trusted callers). The
6115 host name and its colon can be omitted when only the protocol is to be set.
6116 Note the Exim already has two private options, -pd and -ps, that refer to
6117 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6118 "s", but I don't think that's a major issue.
6120 10. A number of refactoring changes to the code, none of which should affect
6123 (a) The number of logging options was getting close to filling up the
6124 32-bit word that was used as a bit map. I have split them into two classes:
6125 those that are passed in the argument to log_write(), and those that are
6126 only ever tested independently outside of that function. These are now in
6127 separate 32-bit words, so there is plenty of room for expansion again.
6128 There is no change in the user interface or the logging behaviour.
6130 (b) When building, for example, log lines, the code previously used a
6131 macro that called string_cat() twice, in order to add two strings. This is
6132 not really sufficiently general. Furthermore, there was one instance where
6133 it was actually wrong because one of the argument was used twice, and in
6134 one call a function was used. (As it happened, calling the function twice
6135 did not affect the overall behaviour.) The macro has been replaced by a
6136 function that can join an arbitrary number of extra strings onto a growing
6139 (c) The code for expansion conditions now uses a table and a binary chop
6140 instead of a serial search (which was left over from when there were very
6141 few conditions). Also, it now recognizes conditions like "pam" even when
6142 the relevant support is not compiled in: a suitably worded error message is
6143 given if an attempt is made to use such a condition.
6145 11. Added ${time_interval:xxxxx}.
6147 12. A bug was causing one of the ddress fields not to be passed back correctly
6148 from remote delivery subprocesses. The field in question was not being
6149 subsequently used, so this caused to problems in practice.
6151 13. Added new log selectors queue_time and deliver_time.
6153 14. Might have fixed a bug in maildirsizefile handling that threw up
6154 "unexpected character" debug warnings, and recalculated the data
6155 unnecessarily. In any case, I expanded the warning message to give more
6158 15. Added the message "Restricted characters in address" to the statements in
6159 the default ACL that block characters like @ and % in local parts.
6161 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6162 Three changes have been made:
6164 (a) There was a serious bug; a negative response to MAIL caused the whole
6165 recipient domain to be cached as invalid, thereby blocking all messages
6166 to all local parts at the same domain, from all senders. This bug has
6167 been fixed. The domain is no longer cached after a negative response to
6168 MAIL if the sender used is not empty.
6170 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6173 (c) A new callout option, "use_sender" has been added for people who want
6174 the modified behaviour.
6180 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6183 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6184 indeed breaks things for older releases.
6186 3. Added additional logging to the case where there is a problem reading data
6187 from a filter that is running in a subprocess using a pipe, in order to
6188 try to track down a specific problem.
6190 4. Testing facility fudge: when running in the test harness and attempting
6191 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6192 getting "No route to host". Convert this to a timeout.
6194 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6197 6. Some OS don't have socklen_t but use size_t instead. This affects the
6198 fifth argument of getsockopt() amongst other things. This is now
6199 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6200 can be set for individual OS. I have set it for SunOS5, OSF1, and
6201 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6202 some earlier ones do not.
6204 7. Change 4.30/15 was not doing the test caselessly.
6206 8. The standard form for an IPv6 address literal was being rejected by address
6207 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6208 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6209 this, as well as the form without the "IPv6" on the front (but only when
6210 address literals are enabled, of course).
6212 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6214 10. Exim crashed if a message with an empty sender address specified by -f
6215 encountered a router with an errors_to setting. This could be provoked only
6216 by a command such as
6220 where an empty string was supplied; "<>" did not hit this bug.
6222 11. Installed PCRE release 4.5.
6224 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6225 remained set. It is now erased.
6227 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6228 times from message ids (which are base 36 rather than the normal 62).
6230 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6231 were being counted as actual protocol errors, and logged if the log
6232 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6233 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6234 and DATA following a set of rejected RCPTs do not count as protocol errors.
6235 In other words, Exim assumes they were pipelined, though this may not
6236 actually be the case. Of course, in all cases the client gets an
6237 appropriate error code.
6239 15. If a lookup fails in an ACL condition, a message about the failure may
6240 be available; it is used if testing the ACL cannot continue, because most
6241 such messages specify what the cause of the deferral is. However, some
6242 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6243 that caused an old message to be retained and used if a later statement
6244 caused a defer, replacing the real cause of the deferral.
6246 16. If an IP address had so many PTR records that the DNS lookup buffer
6247 was not large enough to hold them, Exim could crash while trying to process
6248 the truncated data. It now detects and logs this case.
6250 17. Further to 4.21/58, another change has been made: if (and only if) the
6251 first line of a message (the first header line) ends with CRLF, a bare LF
6252 in a subsequent header line has a space inserted after it, so as not to
6253 terminate the header.
6255 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6256 unnecessarily, used atoi() instead of strtol(), and didn't check the
6257 termination when getting file sizes from file names by regex.
6259 19. Completely re-implemented the support for maildirsize files, in the light
6260 of a number of problems with the previous contributed implementation
6261 (4.30/29). In particular:
6263 . If the quota is zero, the maildirsize file is maintained, but no quota is
6266 . If the maildir directory does not exist, it is created before any attempt
6267 to write a maildirsize file.
6269 . The quota value in the file is just a cache; if the quota is changed in
6270 the transport, the new value overrides.
6272 . A regular expression is available for excluding directories from the
6275 20. The autoreply transport checks the characters in options that define the
6276 message's headers; it allows continued headers, but it was checking with
6277 isspace() after an embedded newline instead of explicitly looking for a
6280 21. If all the "regular" hosts to which an address was routed had passed their
6281 expiry times, and had not reached their retry times, the address was
6282 bounced, even if fallback hosts were defined. Now Exim should go on to try
6285 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6286 equivalent code in the SMTP transport. Some hosts send humungous responses
6287 to HELO/EHLO, more than 1024 it seems.
6289 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6290 gives compiler warnings in some environments. I've now done it "properly",
6293 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6294 (because of problems with the built-in one) was declared to return uschar *
6295 instead of char *, causing compiler failure.
6297 25. Fixed a file descriptor leak when processing alias/forward files.
6299 26. Fixed a minor format string issue in dbfn.c.
6301 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6303 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6304 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6307 29. When a sender address is verified, it is cached, to save repeating the test
6308 when there is more than one recipient in a message. However, when the
6309 verification involves a callout, it is possible for different callout
6310 options to be set for different recipients. It is too complicated to keep
6311 track of this in the cache, so now Exim always runs a verification when a
6312 callout is required, relying on the callout cache for the optimization.
6313 The overhead is duplication of the address routing, but this should not be
6316 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6317 to be verified with callout=postmaster, and the main callout worked but the
6318 postmaster check failed, the verification correctly failed. However, if a
6319 subsequent RCPT command asked for sender verification *without* the
6320 postmaster check, incorrect caching caused this verification also to fail,
6323 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6324 it was not caching the DNS options (qualify_single, search_parents) that
6325 were used when the lookup failed. A subsequent lookup with different
6326 options therefore always gave the same answer, though there were cases
6327 where it should not have. (Example: a "domains = !$mx_any" option on a
6328 dnslookup router: the "domains" option is always processed without any
6329 widening, but the router might have qualify_single set.) Now Exim uses the
6330 cached value only when the same options are set.
6332 32. Added John Jetmore's "exipick" utility to the distribution.
6334 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6335 than a timeout (e.g. a certificate is required, and is not provided), an
6336 Exim server now closes the connection immediately. Previously it waited for
6337 the client to close - but if the client is SSL, it seems that they each
6338 wait for each other, leading to a delay before one of them times out.
6340 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6341 maintained 0.8.x compatibility because I don't think many are using it, and
6342 it is clearly obsolete.
6344 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6347 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6348 client certificate was expired. A simple patch fixes this, though I don't
6349 understand the full logic of why the verify callback is called multiple
6352 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6353 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6354 which causes problems with some clients (such as the Certicom SSL Plus
6355 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6356 disables the coutermeasure allowing Eudora to connect."
6358 38. Exim was not checking that a write() to a log file succeeded. This could
6359 lead to Bad Things if a log got too big, in particular if it hit a file
6360 size limit. Exim now panics and dies if it cannot write to a log file, just
6361 as it does if it cannot open a log file.
6363 39. Modified OS/Makefile-Linux so that it now contains
6365 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6367 The two -D definitions ensure that Exim is compiled with large file
6368 support, which makes it possible to handle log files that are bigger than
6371 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6372 instance) a domain was checked against a named list that involved a lookup,
6373 causing $domain_data to be set, then another domain was checked against the
6374 same list, then the first domain was re-checked, the value of $domain_data
6375 after the final check could be wrong. In particular, if the second check
6376 failed, it could be set empty. This bug probably also applied to
6379 41. The strip_trailing_dot option was not being applied to the address given
6380 with the -f command-line option.
6382 42. The code for reading a message's header from the spool was incrementing
6383 $received_count, but never initializing it. This meant that the value was
6384 incorrect (doubled) while delivering a message in the same process in which
6385 it was received. In the most common configuration of Exim, this never
6386 happens - a fresh exec is done - but it can happen when
6387 deliver_drop_privilege is set.
6389 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6390 it now includes up to 150 characters of the unexpected data in the log
6393 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6394 and building data strings. The size of both of these buffers was 10 000
6395 bytes - far larger than anybody would *ever* want, thought I. Needless to
6396 say, somebody hit the limit. I have increased the maximum line length to
6397 20 000 and the maximum data length of concatenated lines to 100 000. I have
6398 also fixed two bugs, because there was no checking on these buffers. Tsk,
6399 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6400 buffer is too small.
6402 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6403 lsearch lookups. Now it does.
6405 46. When parsing a route_list item in a manualroute router, a fixed-length
6406 buffer was used for the list of hosts. I made this 1024 bytes long,
6407 thinking that nobody would ever have a list of hosts that long. Wrong.
6408 Somebody had a whole pile of complicated expansion conditions, and the
6409 string was silently truncated, leading to an expansion error. It turns out
6410 that it is easier to change to an unlimited length (owing to other changes
6411 that have happened since this code was originally written) than to build
6412 structure for giving a limitation error. The length of the item that
6413 expands into the list of hosts is now unlimited.
6415 47. The lsearch lookup could not handle data where the length of text line was
6416 more than 4095 characters. Such lines were truncated, leading to shortened
6417 data being returned. It should now handle lines of any length.
6419 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6420 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6423 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6424 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6425 no longer happens in comment lines. A list of replacements is now placed
6426 at the head of all of the source files, except those whose only change is
6427 to replace PERL_COMMAND in the very first #! line.
6429 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6430 messages on the queue, with a bottom-up merge sort, using code contributed
6431 by Michael Haardt. This should make operations like -bp somewhat faster on
6432 large queues. It won't affect queue runners, except when queue_run_in_order
6435 51. Installed eximstats 1.31 in the distribution.
6437 52. Added support for SRV lookups to the dnslookup router.
6439 53. If an ACL referred to $message_body or $message_body_end, the value was not
6440 reset for any messages that followed in the same SMTP session.
6442 54. The store-handling optimization for building very long strings was not
6443 differentiating between the different store pools. I don't think this
6444 actually made any difference in practice, but I've tidied it.
6446 55. While running the routers to verify a sender address, $sender_address
6447 was still set to the sender address. This is wrong, because when routing to
6448 send a bounce to the sender, it would be empty. Therefore, I have changed
6449 it so that, while verifying a sender address, $sender_address is set to <>.
6450 (There is no change to what happens when verifying a recipient address.)
6452 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6453 target A or AAAA records (if not already returned) without resetting the
6454 qualify_single or search_parents options of the DNS resolver. These are
6455 inappropriate in this case because the targets of MX and SRV records must
6456 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6457 target that, when qualified, matched something in the local domain. These
6458 two options are now turned off when doing these lookups.
6460 57. It seems that at least some releases of Reiserfs (which does not have the
6461 concept of a fixed number of inodes) returns zero and not -1 for the
6462 number of available inodes. This interacted badly with check_spool_inodes,
6463 which assumed that -1 was the "no such thing" setting. What I have done is
6464 to check that the total number of inodes is greater than zero before doing
6465 the test of how many are available.
6467 58. When a "warn" ACL statement has a log_message modifier, the message is
6468 remembered, and not repeated. This is to avoid a lot of repetition when a
6469 message has many recipients that cause the same warning to be written.
6470 However, Exim was preserving the list of already written lines for an
6471 entire SMTP session, which doesn't seem right. The memory is now reset if a
6472 new message is started.
6474 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6475 debugging output unless log_rewrite was also set.
6477 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6478 of (int)(handle) when we know that handle contains (void *)(-1).
6480 61. The Exim daemon panic-logs an error return when it closes the incoming
6481 connection. However "connection reset by peer" seems to be common, and
6482 isn't really an error worthy of noting specially, so that particular error
6485 62. When Exim is trying to find all the local interfaces, it used to panic and
6486 die if the ioctl to get the interface flags failed. However, it seems that
6487 on at least one OS (Solaris 9) it is possible to have an interface that is
6488 included in the list of interfaces, but for which you get a failure error
6489 for this call. This happens when the interface is not "plumbed" into a
6490 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6491 failure of the "get flags" call assumes that the interface is down.
6493 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6494 makes life easier for people who are doing arithmetic on fields extracted
6495 from dates, where you often get leading zeros that should not be
6496 interpreted as octal.
6498 64. Added qualify_domain to the redirect router, to override the global
6501 65. If a pathologically long header line contained very many addresses (the
6502 report of this problem mentioned 10 000) and each of them was rewritten,
6503 Exim could use up a very large amount of memory. (It kept on making new
6504 copies of the header line as it rewrote, and never released the old ones.)
6505 At the expense of a bit more processing, the header rewriting function has
6506 been changed so that it no longer eats memory in this way.
6508 66. The generation of the Received: header has been moved from the time that a
6509 message starts to be received, to the time that it finishes. The timestamp
6510 in the Received: header should now be very close to that of the <= log
6511 line. There are two side-effects of this change:
6513 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6514 logged header lines no longer include the local Received: line, because
6515 it has not yet been created. The same applies to a copy of the message
6516 that is returned to a non-SMTP sender when a message is rejected.
6518 (b) When a filter file is tested using -bf, no additional Received: header
6519 is added to the test message. After some thought, I decided that this
6522 This change does not affect the value of $received_for. It is still set
6523 after address rewriting, but before local_scan() is called.
6525 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6527 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6528 gave an unhelpful panic error message, and a defer error. I have managed to
6529 change this behaviour so that it now rejects any supplied certificate,
6530 which seems right, as the list of acceptable certificates is empty.
6532 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6533 gave an unhelpful defer error. I have not managed to make this reject any
6534 supplied certificates, but the error message it gives is "no certificate
6535 supplied", which is not helpful.
6537 70. exigrep's output now also includes lines that are not associated with any
6538 message, but which match the given pattern. Implemented by a patch from
6539 Martin Sluka, which also tidied up the Perl a bit.
6541 71. Recipient callout verification, like sender verification, was using <> in
6542 the MAIL FROM command. This isn't really the right thing, since the actual
6543 sender may affect whether the remote host accepts the recipient or not. I
6544 have changed it to use the actual sender in the callout; this means that
6545 the cache record is now keyed on a recipient/sender pair, not just the
6546 recipient address. There doesn't seem to be a real danger of callout loops,
6547 since a callout by the remote host to check the sender would use <>.
6548 [SEE ABOVE: changed after hitting problems.]
6550 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6551 temporary errors. However, in the case of such a code being given after
6552 the end of a data transmission (i.e. after ".") Exim was failing to write
6553 a retry record for the message. (Yes, there was some broken host that was
6554 actually sending 8xx at this point.)
6556 73. An unknown lookup type in a host list could cause Exim to panic-die when
6557 the list was checked. (An example that provoked this was putting <; in the
6558 middle of a list instead of at the start.) If this happened during a DATA
6559 ACL check, a -D file could be left lying around. This kind of configuration
6560 error no longer causes Exim to die; instead it causes a defer error. The
6561 incident is still logged to the main and panic logs.
6563 74. Buglet left over from Exim 3 conversion. The message "too many messages
6564 in one connection" was written to the rejectlog but not the mainlog, except
6565 when address rewriting (yes!) was being logged.
6567 75. Added write_rejectlog option.
6569 76. When a system filter was run not as root (that is, when system_filter_user
6570 was set), the values of the $n variables were not being returned to the
6571 main process; thus, they were not subsequently available in the $sn
6574 77. Added +return_path_on_delivery log selector.
6576 78. A connection timeout was being treated differently from recipients deferred
6577 when testing hosts_max_try with a message that was older than the host's
6578 retry timeout. (The host should not be counted, thus allowing all hosts to
6579 be tried at least once before bouncing.) This may have been the cause of an
6580 occasionally reported bug whereby a message would remain on the queue
6581 longer than the retry timeout, but would be bounced if a delivery was
6582 forced. I say "may" because I never totally pinned down the problem;
6583 setting up timeout/retry tests is difficult. See also the next item.
6585 79. The ultimate address timeout was not being applied to errors that involved
6586 a combination of host plus message (for example, a timeout on a MAIL
6587 command). When an address resolved to a number of possible hosts, and they
6588 were not all tried for each delivery (e.g. because of hosts_max_try), a
6589 message could remain on the queue longer than the retry timeout.
6591 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6594 81. Fixed an obscure SMTP outgoing bug which required at least the following
6595 conditions: (a) there was another message waiting for the same server;
6596 (b) the server returned 5xx to all RCPT commands in the first message so
6597 that the message was not completed; (c) the server dropped the connection
6598 or gave a negative response to the RSET that Exim sends to abort the
6599 transaction. The observed case was a dropped connection after DATA that had
6600 been sent in pipelining mode. That is, the server had advertised PIPELINING
6601 but was not implementing it correctly. The effect of the bug was incorrect
6602 behaviour, such as trying another host, and this could lead to a crash.
6608 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6609 and daemon.c were passed as pointers to ints; they should have been
6610 pointers to socklen_t variables (which are typically unsigned ints).
6612 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6615 3. Fixed a really odd bug that affected only the testing scheme; patching a
6616 certain fixed string in the binary changed the value of another string that
6617 happened to be identical to the end of the original first string.
6619 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6620 name", it returns that address as the IP address. On some operating
6621 systems (e.g. Solaris), it also passes back the IP address string as the
6622 "host name". However, on others (e.g. Linux), it passes back an empty
6623 string. Exim wasn't checking for this, and was changing the host name to an
6624 empty string, assuming it had been canonicalized.
6626 5. Although rare, it is permitted to have more than one PTR record for a given
6627 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6628 all the names associated with an address, because they do in Solaris.
6629 However, it seems that they do not in Linux for data that comes from the
6630 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6631 I found this out when I moved to a new Linux workstation and tried to run
6632 the Exim test suite.
6634 To get round this problem I have changed the code so that it now does its
6635 own call to the DNS to look up PTR records when searching for a host name.
6636 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6637 addresses that are only in /etc/hosts are still found.
6639 This behaviour is, however, controlled by an option called host_lookup_
6640 order, which defaults to "bydns:byaddr". If people want to use the other
6641 order, or indeed, just use one or the other means of lookup, they can
6642 specify it in this variable.
6644 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6645 some operating systems, this comes back from gethostbyaddr() as an empty
6646 string, and this is what Exim used to test for. However, it seems that in
6647 other systems, "." is yielded. Exim now tests for this case too.
6649 7. The values of check_spool_space and check_log_space are now held internally
6650 as a number of kilobytes instead of an absolute number of bytes. If a
6651 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6652 kilobyte. This means that much larger values can be stored.
6654 8. Exim monitor: an attempt to get the action menu when not actually pointing
6655 at a message produces an empty menu entitled "No message selected". This
6656 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6657 no entries in it ("Shell widget menu has zero width and/or height"). So I
6658 have added a single, blank menu entry in this case.
6660 9. Added ${quote_local_part.
6662 10. MIME decoding is now applied to the contents of Subject: header lines when
6665 11. Now that a reference to $sender_host_address automatically causes a reverse
6666 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6667 host lookup before query-style lookups in lists that might use this
6668 variable. This has therefore been abolished, and the "net-" prefix is no
6669 longer necessary for query-style lookups.
6671 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6672 have been a typo for LFLAGS, so it has been changed.
6674 13. The install script calls Exim with "-C /dev/null" in order to find the
6675 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6676 to be output. However, since Exim outputs its version number before the
6677 error, it didn't break the script. It just looked ugly. I fixed this by
6678 always allowing "-C /dev/null" if the caller is root.
6680 14. Ignore overlarge ACL variable number when reading spool file - insurance
6681 against a later release with more variables having written the file.
6683 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6684 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6685 this, as well as the form without the "IPv6" on the front.
6687 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6688 OS/Makefile-Darwin file.
6690 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6691 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6693 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6694 variable. However, some OS use ints for inodes. I've added cast to long int
6695 to get rid of the compiler warning.
6697 19. I had forgotten to lock out "/../" in configuration file names when
6698 ALT_CONFIG_PREFIX was set.
6700 20. Routers used for verification do not need to specify transports. However,
6701 if such a router generated a host list, and callout was configured, Exim
6702 crashed, because it could not find a port number from the (non-existent)
6703 transport. It now assumes port 25 in this circumstance.
6705 21. Added the -t option to exigrep.
6707 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6708 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6709 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6710 them is set, LOOKUP_LSEARCH is forced.
6712 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6714 24. Added sender and host information to the "rejected by local_scan()" log
6715 line; previously there was no indication of these.
6717 25. Added .include_if_exists.
6719 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6720 Linux. It turns out that not all file systems support this. Apparently some
6721 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6722 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6723 not supported on the file descriptor, is now ignored when Exim is trying to
6724 sync a directory. This applies only to Linux.
6726 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6728 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6729 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6732 29. Added support for maildirsize files from supplied patch (modified a bit).
6734 30. The use of :fail: followed by an empty string could lead Exim to respond to
6735 sender verification failures with (e.g.):
6737 550 Verification failed for <xxx>
6738 550 Sender verify failed
6740 where the first response line was missing the '-' that indicates it is not
6741 the final line of the response.
6743 31. The loop for finding the name of the user that called Exim had a hardwired
6744 limit of 10; it now uses the value of finduser_retries, which is used for
6745 all other user lookups.
6747 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6750 33. Exim was neglecting to zero errno before one call of strtol() when
6751 expanding a string and expecting an integer value. On some systems this
6752 resulted in spurious "integer overflow" errors. Also, it was casting the
6753 result into an int without checking.
6755 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6756 did not work. The code looks as if it has *never* worked, though it appears
6757 to have been documented since at least release 1.62. I have made it work.
6759 35. The "timeout_DNS" error in retry rules, also documented since at least
6760 1.62, also never worked. As it isn't clear exactly what this means, and
6761 clearly it isn't a major issue, I have abolished the feature by treating it
6762 as "timeout", and writing a warning to the main and panic logs.
6764 36. The display of retry rules for -brt wasn't always showing the error code
6767 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6768 timeout_connect_A, timeout_connect_MX.
6770 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6771 to the empty sender.
6773 39. The daemon was not analysing the content of -oX till after it had closed
6774 stderr and disconnected from the controlling terminal. This meant that any
6775 syntax errors were only noted on the panic log, and the return code from
6776 the command was 0. By re-arranging the code a little, I've made the
6777 decoding happen first, so such errors now appear on stderr, and the return
6778 code is 1. However, the actual setting up of the sockets still happens in
6779 the disconnected process, so errors there are still only recorded on the
6782 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6783 connections (as happens on some IP stacks) was logged at start up time as
6784 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6785 it from "IPv6 and IPv4", which means that two separate sockets are being
6788 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6789 says whether AF_INET or AF_INET6 was passed as an argument.
6791 42. Exiwhat output was messed up when time zones were included in log
6794 43. Exiwhat now gives more information about the daemon's listening ports,
6795 and whether -tls-on-connect was used.
6797 44. The "port" option of the smtp transport is now expanded.
6799 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6800 silently ignored. Now an error message is written to the main and panic
6803 46. There's a new ACL modifier called "logwrite" which writes to a log file
6804 as soon as it is encountered.
6806 47. Added $local_user_uid and $local_user_gid at routing time.
6808 48. Exim crashed when trying to verify a sender address that was being
6811 49. Exim was recognizing only a space character after ".include". It now also
6812 recognizes a tab character.
6814 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6815 extracting the relevant information from the specification. The man page no
6816 longer contains scrambled data for the -d option, and I've added a section
6817 at the front about calling Exim under different names.
6819 51. Added "extra_headers" argument to the "mail" command in filter files.
6821 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6824 53. Installed eximstats 1.29.
6826 54. Added transport_filter_timeout as a generic transport option.
6828 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6829 Cc: header lines. This was required by RFC 822, but it not required by RFC
6832 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6833 incoming messages that did not have them. Now it does so only if the
6834 message originates locally, that is, if there is no associated remote host
6835 address. When Resent- header lines are present, this applies to the Resent-
6836 lines rather than the non-Resent- lines.
6838 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6839 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6841 58. Messages for configuration errors now include the name of the main
6842 configuration file - useful now that there may be more than one file in a
6843 list (.included file names were always shown).
6845 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6846 for those rare installations that do not start the daemon as root or run it
6847 setuid root. I've cut out the call to initgroups() if the daemon is not
6850 60. The Exim user and group can now be bound into the binary as text strings
6851 that are looked up at the start of Exim's processing.
6853 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6855 62. Added $mailstore_basename variable.
6857 63. Installed patch to sieve.c from Michael Haardt.
6859 64. When Exim failed to open the panic log after failing to open the main log,
6860 the original message it was trying to log was written to stderr and debug
6861 output, but if they were not available (the usual case in production), it
6862 was lost. Now it is written to syslog before the two lines that record the
6863 failures to open the logs.
6865 65. Users' Exim filters run in subprocesses under the user's uid. It is
6866 possible for a "deliver" command or an alias in a "personal" command to
6867 provoke an address rewrite. If logging of address rewriting is configured,
6868 this fails because the process is not running as root or exim. There may be
6869 a better way of dealing with this, but for the moment (because 4.30 needs
6870 to be released), I have disabled address rewrite logging when running a
6871 filter in a non-root, non-exim process.
6877 1. The buildconfig auxiliary program wasn't quoting the value set for
6878 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6879 not defined. This bug was masked in 4.22 by the effect that was fixed in
6882 2. Some messages that were rejected after a message id was allocated were
6883 shown as "incomplete" by exigrep. It no longer does this for messages that
6884 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6886 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6887 have allow_domain_literals set, the ID did not get logged in the <= line.
6888 Domain literals are now always recognized in Message-ID: header lines.
6890 4. The first argument for a ${extract expansion item is the key name or field
6891 number. Leading and trailing spaces in this item were not being ignored,
6892 causing some misleading effects.
6894 5. When deliver_drop_privilege was set, single queue runner processes started
6895 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6896 same command in the process it spins off) were not dropping privilege.
6898 6. When the daemon running as "exim" started a queue runner, it always
6899 re-executed Exim in the spun-off process. This is a waste of effort when
6900 deliver_drop_privilege is set. The new process now just calls the
6901 queue-runner function directly.
6907 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6910 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6911 ignored. Though the use of -oP was forcing the writing of a pid file, it
6912 was always written to the default place.
6914 3. If the message "no IP address found for host xxxx" is generated during
6915 incoming verification, it is now followed by identification of the incoming
6916 connection (so you can more easily find what provoked it).
6918 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6920 5. Added some features to "harden" Exim a bit more against certain attacks:
6922 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6923 be put in Local/Makefile. This is like the never_users runtime option,
6924 but it cannot be overridden. The default setting is "root".
6926 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6927 prefix string with which any file named in a -C command line option
6930 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6931 is retained for -C and -D only if the caller of Exim is root. Without
6932 it, the exim user may also use -C and -D and retain privilege.
6934 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6935 command line option is disabled.
6937 6. Macro names set by the -D option must start with an upper case letter, just
6938 like macro names defined in the configuration file.
6940 7. Added "dereference=" facility to LDAP.
6942 8. Two instances of the typo "uknown" in the source files are fixed.
6944 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6945 the Configure-Makefile script screwed up while processing it.
6947 10. Incorporated PCRE 4.4.
6949 11. The SMTP synchronization check was not operating right at the start of an
6950 SMTP session. For example, it could not catch a HELO sent before the client
6951 waited for the greeting. There is now a check for outstanding input at the
6952 point when the greeting is written. Because of the duplex, asynchronous
6953 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6954 way, but not yet received, when the check is performed.
6956 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6957 on TCP/IP sockets, because this apparently causes some broken clients to
6960 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6961 unchanged) from the Cygwin maintainer.
6963 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6964 was supported instead of testing for mbx. Effectively a typo.
6966 15. The spa authenticator server code was not checking that the input it
6967 received was valid base64.
6969 16. The debug output line for the "set" modifier in ACLs was not showing the
6970 name of the variable that was being set.
6972 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6974 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6975 lookup on its own. Something else was needed to trigger the lookup. For
6976 example, a match in host_lookup or the need for a host name in a host list.
6977 Now, if $sender_host_name is referenced and the host name has not yet been
6978 looked up, a lookup is performed. If the lookup fails, the variable remains
6979 empty, and $host_lookup_failed is set to "1".
6981 19. Added "eqi" as a case-independent comparison operator.
6983 20. The saslauthd authentication condition could segfault if neither service
6984 nor realm was specified.
6986 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6987 error message that was logged was misleading, and incoming SMTP
6988 connections were dropped. The message is now more accurate, and temporary
6989 errors are given to SMTP connections.
6991 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6992 (except RSET) with a 421 error, until QUIT is received. However, it was
6993 failing to send a response to QUIT.
6995 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6996 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6997 ACL is now run after the helo_try_verify_hosts code.
6999 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7000 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7001 case-independent, so other case variants are also recognized). Apparently
7002 some systems use these upper case variants.
7004 25. If more than two messages were waiting for the same host, and a transport
7005 filter was specified for the transport, Exim sent two messages over the
7006 same TCP/IP connection, and then failed with "socket operation on non-
7007 socket" when it tried to send the third.
7009 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7011 27. The extern definition of crypt16() in expand.c was not being excluded when
7012 the OS had its own crypt16() function.
7014 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7015 as a preferred synonym for return_size_limit, both as an option and as an
7018 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7020 30. Changed the default configuration ACL to relax the local part checking rule
7021 for addresses that are not in any local domains. For these addresses,
7022 slashes and pipe symbols are allowed within local parts, but the sequence
7023 /../ is explicitly forbidden.
7025 31. SPA server authentication was not clearing the challenge buffer before
7028 32. log_message in a "warn" ACL statement was writing to the reject log as
7029 well as to the main log, which contradicts the documentation and doesn't
7030 seem right (because no rejection is happening). So I have stopped it.
7032 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7033 However, I am unable to do any testing of this.
7035 34. Fixed an infelicity in the appendfile transport. When checking directories
7036 for a mailbox, to see if any needed to be created, it was accidentally
7037 using path names with one or more superfluous leading slashes; tracing
7038 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7040 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7041 contents are added to the log line that is written for every discarded
7042 recipient. (Previously a log_message setting was ignored.)
7044 36. The ${quote: operator now quotes the string if it is empty.
7046 37. The install script runs exim in order to find its version number. If for
7047 some reason other than non-existence or emptiness, which it checks, it
7048 could not run './exim', it was installing it with an empty version number,
7049 i.e. as "exim-". This error state is now caught, and the installation is
7052 38. An argument was missing from the function that creates an error message
7053 when Exim fails to connect to the socket for saslauthd authentication.
7054 This could cause Exim to crash, or give a corrupted message.
7056 39. Added isip, isip4, and isip6 to ${if conditions.
7058 40. The ACL variables $acl_xx are now saved with the message, and can be
7059 accessed later in routers, transports, and filters.
7061 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7062 strings in the file are not string-expanded.
7064 42. If a MAIL command specified a SIZE value that was too large to fit into an
7065 int variable, the check against message_size_limit failed. Such values are
7066 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7067 day this will have to be increased, but I don't think I want to be around
7068 when emails are that large.
7075 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7076 iconv() is not standard in FreeBSD.
7078 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7079 IPv6 enabled. The observed symptom was a segmentation fault on return from
7080 the function os_common_find_running_interfaces() in src/os.c.
7082 3. In the check_special_case() function in daemon.c I had used "errno" as an
7083 argument name, which causes warnings on some systems. This was basically a
7084 typo, since it was named "eno" in the comments!
7086 4. The code that waits for the clock to tick (at a resolution of some fraction
7087 of a second) so as to ensure message-id uniqueness was always waiting for
7088 at least one whole tick, when it could have waited for less. [This is
7089 almost certainly not relevant at current processor speeds, where it is
7090 unlikely to ever wait at all. But we try to future-proof.]
7092 5. The function that sleeps for a time interval that includes fractions of a
7093 second contained a race. It did not block SIGALRM between setting the
7094 timer, and suspending (a couple of lines later). If the interval was short
7095 and the sigsuspend() was delayed until after it had expired, the suspension
7096 never ended. On busy systems this could lead to processes getting stuck for
7099 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7100 process, before it forks any delivery processes. The open lookup caching
7101 mechanism meant that the open file or database connection was passed into
7102 the delivery process. The problem was that delivery processes always tidy
7103 up cached lookup data. This could cause a problem for the next delivery
7104 process started by the queue runner, because the external queue runner
7105 process does not know about the closure. So the next delivery process
7106 still has data in the lookup cache. In the case of a file lookup, there was
7107 no problem because closing a file descriptor in a subprocess doesn't affect
7108 the parent. However, if the lookup was caching a connection to a database,
7109 the connection was closed, and the second delivery process was likely to
7110 see errors such as "PGSQL: query failed: server closed the connection
7111 unexpectedly". The problem has been fixed by closing all cached lookups
7112 in a queue runner before running a delivery process.
7114 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7115 seem to have the "const" qualifier which it has on other OS. I've
7118 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7119 *of the same type* (client or server) with the same public name that an
7120 error should be diagnosed.
7122 9. When Exim looked up a host name for an IP address, but failed to find the
7123 original IP address when looking up the host name (a safety check), it
7124 output the message "<ip address> does not match any IP for NULL", which was
7125 confusing, to say the least. The bug was that the host name should have
7126 appeared instead of "NULL".
7128 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7129 user that is built into the binary, and the -C or -D options is used, root
7130 privilege is dropped before the configuration file is read. In addition,
7131 logging is switched to stderr instead of the normal log files. If the
7132 configuration then re-defines the Exim user, the unprivileged environment
7133 is probably not what is expected, so Exim logs a panic warning message (but
7136 However, if deliver_drop_privilege is set, the unprivileged state may well
7137 be exactly what is intended, so the warning has been cut out in that case,
7138 and Exim is allowed to try to write to its normal log files.
7144 1. smtp_return_error_details was not giving details for temporary sender
7145 or receiver verification errors.
7147 2. Diagnose a configuration error if two authenticators have the same public
7150 3. Exim used not to create the message log file for a message until the first
7151 delivery attempt. This could be confusing when incoming messages were held
7152 for policy or load reasons. The message log file is now created at the time
7153 the message is received, and an initial "Received" line is written to it.
7155 4. The automatically generated man page for command line options had a minor
7156 bug that caused no ill effects; however, a more serious problem was that
7157 the procedure for building the man page automatically didn't always
7158 operate. Consequently, release 4.20 contains an out-of-date version. This
7159 shouldn't happen again.
7161 5. When building Exim with embedded Perl support, the script that builds the
7162 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7163 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7165 6. The freeze_tell option was not being used for messages that were frozen on
7166 arrival, either by an ACL or by local_scan().
7168 7. Added the smtp_incomplete_transaction log selector.
7170 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7171 was accepting AUTH without a new EHLO.
7173 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7174 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7177 10. Logging of TCP/IP connections (when configured) now happens in the main
7178 daemon process instead of the child process, so that the TCP/IP connection
7179 count is more accurate (but it can never be perfect).
7181 11. The use of "drop" in a nested ACL was not being handled correctly in the
7182 outer ACL. Now, if condition failure induced by the nested "drop" causes
7183 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7184 or "require"), the connection is dropped.
7186 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7187 that yield "discard" can now be used with an "accept" or a "discard" verb,
7188 but an error is generated for any others (because I can't see a useful way
7189 to define what should happen).
7191 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7192 are now processed in the same way as lines in the Exim configuration file.
7193 In particular, continuation lines are supported.
7195 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7197 15. Added -ti meaning -t -i.
7199 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7200 domains, and warn by logging if others are found.
7202 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7203 returned from GIFCONF when Exim is trying to find the list of interfaces on
7204 a host. The code in os.c has been modified to copy each ifreq to an aligned
7205 structure in all cases.
7207 Also, in some cases, the returned ifreq's were being copied to a 'struct
7208 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7209 means the last couple of bytes of an IPv6 address could be chopped if the
7210 ifreq contained only a normal sockaddr (14 bytes storage).
7212 18. Named domain lists were not supported in the hosts_treat_as_local option.
7213 An entry such as +xxxx was not recognized, and was treated as a literal
7216 19. Ensure that header lines added by a DATA ACL are included in the reject log
7217 if the ACL subsequently rejects the message.
7219 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7220 MD5 (which is deprecated).
7222 21. When testing a filter file using -bf, Exim was writing a message when it
7223 took the sender from a "From " line in the message, but it was not doing so
7224 when it took $return_path from a Return-Path: header line. It now does.
7226 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7227 with a valid header line field name (a series of printing characters
7228 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7230 23. Changed "disc" in the source to "disk" to conform to the documentation and
7231 the book and for uniformity.
7233 24. Ignore Sendmail's -Ooption=value command line item.
7235 25. When execve() failed while trying to run a command in a pipe transport,
7236 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7237 could be confused with a return value of 69 from the command itself. This
7238 has been changed to 127, the value the shell returns if it is asked to run
7239 a non-existent command. The wording for the related log line suggests a
7240 non-existent command as the problem.
7242 26. If received_header_text expands to an empty string, do not add a Received:
7243 header line to the message. (Well, it adds a token one on the spool, but
7244 marks it "old" so that it doesn't get used or transmitted.)
7246 27. Installed eximstats 1.28 (addition of -nt option).
7248 28. There was no check for failure on the call to getsockname() in the daemon
7249 code. This can fail if there is a shortage of resources on the system, with
7250 ENOMEM, for example. A temporary error is now given on failure.
7252 29. Contrary to the C standard, it seems that in some environments, the
7253 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7254 program. Exim now does this explicitly; it affects the formatting of
7255 timestamps using strftime().
7257 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7258 complaints. I've now initialized all the variables, to avoid this.
7260 32. Header lines added by a system filter were not being "seen" during
7261 transport-time rewrites.
7263 33. The info_callback() function passed to OpenSSL is set up with type void
7264 (*)(SSL *, int, int), as described somewhere. However, when calling the
7265 function (actually a macro) that sets it up, the type void(*)() is
7266 expected. I've put in a cast to prevent warnings from picky compilers.
7268 34. If a DNS black list lookup found a CNAME record, but there were no A
7269 records associated with the domain it pointed at, Exim crashed.
7271 35. If a DNS black list lookup returned more than one A record, Exim ignored
7272 all but the first. It now scans all returned addresses if a particular IP
7273 value is being sought. In this situation, the contents of the
7274 $dnslist_value variable are a list of all the addresses, separated by a
7277 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7278 to accept a host name and all its aliases if the forward lookup for any of
7279 them yielded the IP address of the incoming connection. Now it accepts only
7280 those names whose forward lookup yields the correct IP address. Any other
7281 names are discarded. This closes a loophole whereby a rogue DNS
7282 administrator could create reverse DNS records to break through a
7283 wildcarded host restriction in an ACL.
7285 37. If a user filter or a system filter that ran in a subprocess used any of
7286 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7287 the wrong values were passed to the pipe command ($thisaddress had the
7288 value of $0, $0 had the value of $1, etc). This bug was introduced by
7289 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7291 38. Improved the line breaking for long SMTP error messages from ACLs.
7292 Previously, if there was no break point between 40 and 75 characters, Exim
7293 left the rest of the message alone. Two changes have been made: (a) I've
7294 reduced the minimum length to 35 characters; (b) if it can't find a break
7295 point between 35 and 75 characters, it looks ahead and uses the first one
7296 that it finds. This may give the occasional overlong line, but at least the
7297 remaining text gets split now.
7299 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7300 file descriptors might be low, and that setting 1000 would always raise it.
7301 It turns out that in some environments, the limit is already over 1000 and
7302 that lowering it causes trouble. So now Exim takes care not to decrease it.
7304 40. When delivering a message, the value of $return_path is set to $sender_
7305 address at the start of routing (routers may change the value). By an
7306 oversight, this default was not being set up when an address was tested by
7307 -bt or -bv, which affected the outcome if any router or filter referred to
7310 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7311 log only when "message" or "log_message" are set. However, if one of the
7312 conditions was an address verification, or a call to a nested ACL, the
7313 messages generated by the underlying test were being passed through. This
7314 no longer happens. The underlying message is available in $acl_verify_
7315 message for both "message" and "log_message" expansions, so it can be
7316 passed through if needed.
7318 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7319 new expansion $bh_ to give the encoded byte string without charset
7320 translation. Translation happens only if iconv() is available; HAVE_ICONV
7321 indicates this at build time. HEADERS_CHARSET gives the charset to
7322 translate to; headers_charset can change it in the configuration, and
7323 "headers charset" can change it in an individual filter file.
7325 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7326 that creates RFC 2047 encoded "words" labels them as that charset instead
7327 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7328 expansion operator; (ii) when Exim creates a From: line for a local
7329 message; (iii) when a header line is rewritten to include a "phrase" part.
7331 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7332 buggy, causing it to skip the first lines of messages whose message ID
7333 ended in 'D'. This would not have bitten before Exim release 4.14, because
7334 message IDs were unlikely to end in 'D' before then. The effect was to have
7335 incorrect size information for certain domains.
7337 45. #include "config.h" was missing at the start of the crypt16.c module. This
7338 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7341 46. If there was a timeout during a "random" callout check, Exim treated it as
7342 a failure of the random address, and carried on sending RSET and the real
7343 address. If the delay was just some slowness somewhere, the response to the
7344 original RCPT would be taken as a response to RSET and so on, causing
7345 mayhem of various kinds.
7347 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7348 when I implemented it. It didn't allow for the fact that some option values
7349 may legitimately be negative (e.g. size_addition), and it didn't even do
7350 the right test for positive values.
7352 48. Domain names in DNS records are case-independent. Exim always looks them up
7353 in lower case. Some resolvers return domain names in exactly the case they
7354 appear in the zone file, that is, they may contain uppercase letters. Not
7355 all resolvers do this - some return always lower case. Exim was treating a
7356 change of case by a resolver as a change of domain, similar to a widening
7357 of a domain abbreviation. This triggered its re-routing code and so it was
7358 trying to route what was effectively the same domain again. This normally
7359 caused routing to fail (because the router wouldn't handle the domain
7360 twice). Now Exim checks for this case specially, and just changes the
7361 casing of the domain that it ultimately uses when it transmits the message
7364 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7367 50. If a filter generated a file delivery with a non-absolute name (possible if
7368 no home directory exists for the router), the forbid_file option was not
7371 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7372 the existing equality matching.
7374 52. Exim was using ints instead of ino_t variables in some places where it was
7375 dealing with inode numbers.
7377 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7378 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7379 called TMPDIR, and if it finds it is different, it changes its value.
7381 54. The smtp_printf() function is now made available to local_scan() so
7382 additional output lines can be written before returning. There is also an
7383 smtp_fflush() function to enable the detection of a dropped connection.
7384 The variables smtp_input and smtp_batched_input are exported to
7387 55. Changed the default runtime configuration: the message "Unknown user"
7388 has been removed from the ACL, and instead placed on the localuser router,
7389 using the cannot_route_message feature. This means that any verification
7390 failures that generate their own messages won't get overridden. Similarly,
7391 the "Unrouteable address" message that was in the ACL for unverifiable
7392 relay addresses has also been removed.
7394 56. Added hosts_avoid_esmtp to the smtp transport.
7396 57. The exicyclog script was not checking for the esoteric option
7397 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7398 will work only if exicyclog is run under the appropriate euid.
7400 58. Following a discussion on the list, the rules by which Exim recognises line
7401 endings on incoming messages have been changed. The -dropcr and drop_cr
7402 options are now no-ops, retained only for backwards compatibility. The
7403 following line terminators are recognized: LF CRLF CR. However, special
7404 processing applies to CR:
7406 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7407 nor a local message in the state where . is a terminator.
7409 (ii) If a bare CR is encountered in a header line, an extra space is added
7410 after the line terminator so as not to end the header. The reasoning
7411 behind this is that bare CRs in header lines are most likely either
7412 to be mistakes, or people trying to play silly games.
7414 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7415 was being incorrectly given as 18 bytes larger than it should have been.
7416 This is a VOB (very old bug).
7418 60. This may never have affected anything current, but just in case it has:
7419 When the local host is found other than at the start of a list of hosts,
7420 the local host, those with the same MX, and any that follow, are discarded.
7421 When the list in question was part of a longer list of hosts, the following
7422 hosts (not currently being processed) were also being discarded. This no
7423 longer happens. I'm not sure if this situation could ever has previously
7426 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7429 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7430 header. This is something that is recommended in a new Internet Draft, and
7431 is something that is documented as being done by Sendmail. There are two
7432 possible values. For messages generated by the autoreply transport, Exim
7435 Auto-Submitted: auto-replied
7437 whereas for all other generated messages (e.g. bounces) it adds
7439 Auto-Submitted: auto-generated
7441 63. The "personal" condition in filters now includes a test for the
7442 Auto-Submitted: header. If it contains the string "auto-" the message it
7443 not considered personal.
7445 64. Added rcpt_include_affixes as a generic transport option.
7447 65. Added queue_only_override (default true).
7449 66. Added the syslog_duplication option.
7451 67. If what should have been the first header line of a message consisted of
7452 a space followed by a colon, Exim was mis-interpreting it as a header line.
7453 It isn't of course - it is syntactically invalid and should therefore be
7454 treated as the start of the message body. The misbehaviour could have
7455 caused a number of strange effects, including loss of data in subsequent
7456 header lines, and spool format errors.
7458 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7459 client host had authenticated. This control can now be exercised by an ACL
7460 for more flexibility.
7462 69. By default, callouts do not happen when testing with -bh. There is now a
7463 variant, -bhc, which does actually run the callout code, including
7464 consulting and updating the callout cache.
7466 70. Added support for saslauthd authentication, courtesy of Alexander
7469 71. If statvfs() failed on the spool or log directories while checking their
7470 size for availability, Exim confusingly gave the error "space shortage".
7471 Furthermore, in debugging mode it crashed with a floating point exception.
7472 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7473 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7474 serious problem, Exim now writes to the main and panic logs when this
7475 happens, with details of the failure. It then refuses to accept the
7476 incoming message, giving the message "spool directory problem" or "log
7477 directory problem" with a 421 code for SMTP messages.
7479 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7480 0, 1, and 2 exist, because some OS complain for execs without them (see
7481 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7482 descriptors. However, the code omitted to check that the open succeeded,
7483 causing mysterious errors if for some reason the permissions on /dev/null
7484 got screwed. Now Exim writes a message to the main and panic logs, and
7485 bombs out if it can't open /dev/null.
7487 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7488 interact so that it is all more flexible. It is supposed to remain
7489 backwards compatible. Also added extra_local_interfaces.
7491 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7492 to bomb out with an assertion failure - to the client this appears as a
7493 connection drop. This problem occurs in the part of the code that was taken
7494 from the Samba project. Fortunately, the assertion is in a very simple
7495 function, so I have fixed this by reproducing the function inline in the
7496 one place where it is called, and arranging for authentication to fail
7497 instead of killing the process with assert().
7499 75. The SPA client code was not working when the server requested OEM rather
7500 than Unicode encoding.
7502 76. Added code to make require_files with a specific uid setting more usable in
7503 the case where statting the file as root fails - usually a non-root-mounted
7504 NFS file system. When this happens and the failure is EACCES, Exim now
7505 forks a subprocess and does the per-uid checking as the relevant uid.
7507 77. Added process_log_path.
7509 78. If log_file_path was not explicitly set, a setting of check_log_space or
7510 check_log_inodes was ignored.
7512 79. If a space check for the spool or log partitions fails, the incident is now
7513 logged. Of course, in the latter case the data may get lost...
7515 80. Added the %p formatting code to string_format() so that it can be used to
7516 print addresses in debug_print(). Adjusted all the address printing in the
7517 debugging in store.c to use %p rather than %d.
7519 81. There was a concern that a line of code in smtp_in.c could overflow a
7520 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7521 initially expressed, the concern was not well-founded, because trailing
7522 spaces are removed early. However, if the trailing spaces were followed by
7523 a NULL, they did not get removed, so the overflow was possible. Two fixes
7526 (a) I re-wrote the offending code in a cleaner fashion.
7527 (b) If an incoming SMTP command contains a NULL character, it is rejected
7530 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7531 runs initgroups(), so that if the Exim user is in any additional groups,
7532 they will be used during message reception.
7538 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob