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 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
64 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
65 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
66 now-deprecated earlier definitions used only specified the range up to TLS
67 1.2 (in the older-version library docs).
69 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
71 JH/15 Rework TLS client-side context management. Stop using a global, and
72 explicitly pass a context around. This enables future use of TLS for
73 connections to service-daemons (eg. malware scanning) while a client smtp
74 connection is using TLS; with cutthrough connections this is quite likely.
76 JH/16 Fix ARC verification to do AS checks in reverse order.
78 JH/17 Support a "tls" option on the ${readsocket } expansion item.
80 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
81 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
82 Previously the "utf8" would be re-prepended for every additional message.
84 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
85 Previously thery were accepted, resulting in issues when attempting to
86 forward messages to a non-supporting MTA.
88 PP/02 Let -n work with printing macros too, not just options.
90 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
91 one parent address was copied, and bogus data was used at delivery-logging
92 time. Either a crash (after delivery) or bogus log data could result.
93 Discovery and analysis by Tim Stewart.
95 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
96 Previously if the string ended mid-character, we did not insert the
97 promised '?' replacement.
99 PP/04 Documentation: current string operators work on bytes, not codepoints.
105 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
106 When redis_servers is set to a list of > 1 element, and the Redis servers
107 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
108 case of MOVED into a DEFER case instead, thus moving the query onto the
109 next server in the list. For a cluster of N elements, all N servers must
110 be defined in redis_servers.
112 GF/02 Catch and remove uninitialized value warning in exiqsumm
113 Check for existence of @ARGV before looking at $ARGV[0]
115 JH/01 Replace the store_release() internal interface with store_newblock(),
116 which internalises the check required to safely use the old one, plus
117 the allocate and data copy operations duplicated in both (!) of the
118 extant use locations.
120 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
121 modifier. This matches the restriction on the commandline.
123 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
124 Previously only the last row was returned.
126 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
127 we assumed that tags in the header were well-formed, and parsed the
128 element content after inspecting only the first char of the tag.
129 Assumptions at that stage could crash the receive process on malformed
132 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
133 While running the DKIM ACL we operate on the Permanent memory pool so that
134 variables created with "set" persist to the DATA ACL. Also (at any time)
135 DNS lookups that fail create cache records using the Permanent pool. But
136 expansions release any allocations made on the current pool - so a dnsdb
137 lookup expansion done in the DKIM ACL releases the memory used for the
138 DNS negative-cache, and bad things result. Solution is to switch to the
139 Main pool for expansions.
140 While we're in that code, add checks on the DNS cache during store_reset,
141 active in the testsuite.
142 Problem spotted, and debugging aided, by Wolfgang Breyha.
144 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
145 When none of the hosts presented to a transport match an already-open
146 connection, close it and proceed with the list. Previously we would
147 queue the message. Spotted by Lena with Yahoo, probably involving
150 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
151 Previously a spurious "250 OK id=" response was appended to the proper
154 JH/08 The "support for" informational output now, which built with Content
155 Scanning support, has a line for the malware scanner interfaces compiled
156 in. Interface can be individually included or not at build time.
158 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
159 by the template makefile "src/EDITME". The "STREAM" support for an older
160 ClamAV interface method is removed.
162 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
163 rows affected is given instead).
165 JH/11 The runtime Berkeley DB library version is now additionally output by
166 "exim -d -bV". Previously only the compile-time version was shown.
168 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
169 SMTP connection. Previously, when one had more receipients than the
170 first, an abortive onward connection was made. Move to full support for
171 multiple onward connections in sequence, handling cutthrough connection
172 for all multi-message initiating connections.
174 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
175 routers. Previously, a multi-recipient message would fail to match the
176 onward-connection opened for the first recipient, and cause its closure.
178 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
179 a timeout on read on a GnuTLS initiating connection, resulting in the
180 initiating connection being dropped. This mattered most when the callout
181 was marked defer_ok. Fix to keep the two timeout-detection methods
184 JH/15 Relax results from ACL control request to enable cutthrough, in
185 unsupported situations, from error to silently (except under debug)
186 ignoring. This covers use with PRDR, frozen messages, queue-only and
189 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
191 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
192 metadata, resulting in a crash in free().
194 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
195 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
196 Broken also in d185889f4, with init system revamp.
198 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
199 Previously we abruptly closed the connection after reading a malware-
200 found indication; now we go on to read the "scan ok" response line,
203 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
204 ACL. Previously, a crash would result.
206 JH/19 Speed up macro lookups during configuration file read, by skipping non-
207 macro text after a replacement (previously it was only once per line) and
208 by skipping builtin macros when searching for an uppercase lead character.
210 JH/20 DANE support moved from Experimental to mainline. The Makefile control
211 for the build is renamed.
213 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
214 was allocated for every new TLS startup, meaning one per message. Fix
215 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
217 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
218 reported the original. Fix to report (as far as possible) the ACL
219 result replacing the original.
221 JH/23 Fix memory leak during multi-message connections using STARTTLS under
222 OpenSSL. Certificate information is loaded for every new TLS startup,
223 and the resources needed to be freed.
225 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
227 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
230 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
231 DATA response info to the (existing) per-recipient response info for
232 the "C=" log element. It can have useful tracking info from the
233 destination system. Patch from Simon Arlott.
235 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
236 length value. Previously this would segfault.
238 HS/02 Support Avast multiline protoocol, this allows passing flags to
239 newer versions of the scanner.
241 JH/28 Ensure that variables possibly set during message acceptance are marked
242 dead before release of memory in the daemon loop. This stops complaints
243 about them when the debug_store option is enabled. Discovered specifically
244 for sender_rate_period, but applies to a whole set of variables.
245 Do the same for the queue-runner and queue-list loops, for variables set
246 from spool message files. Do the same for the SMTP per-message loop, for
247 certain variables indirectly set in ACL operations.
249 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
250 as a multi-recipient message from a mailinglist manager). The coding had
251 an arbitrary cutoff number of characters while checking for more input;
252 enforced by writing a NUL into the buffer. This corrupted long / fast
253 input. The problem was exposed more widely when more pipelineing of SMTP
254 responses was introduced, and one Exim system was feeding another.
255 The symptom is log complaints of SMTP syntax error (NUL chars) on the
256 receiving system, and refused recipients seen by the sending system
257 (propating to people being dropped from mailing lists).
258 Discovered and pinpointed by David Carter.
260 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
261 replaced by the ${authresults } expansion.
263 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
265 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
266 allows proper process termination in container environments.
268 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
269 Previously the "final dot" had a newline after it; ensure it is CR,LF.
271 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
272 and "err_perm", deprecated since 4.83 when the RFC-defined words
273 "temperror" and "permerror" were introduced.
275 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
276 transport-filters or DKIM-signing. The restriction was lost in the
277 consolidation of verify-callout and delivery SMTP handling.
278 Extend the restriction to also cover ARC-signing.
280 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
281 in defer=pass mode supply a 450 to the initiator. Previously the message
284 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
285 tls_require_ciphers is used as before.
287 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
288 "pass_unscanned". Only tmpfails from the scanner are written to
289 the paniclog, as they may require admin intervention (permission
290 denied, license issues). Other scanner errors (like decompression
291 bombs) do not cause a paniclog entry.
293 JH/36 Fix reinitialisation of DKIM logging variable between messages.
294 Previously it was possible to log spurious information in receive log
297 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
298 triggered odd behaviour from Outlook Express clients.
300 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
303 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
304 since the IETF WG has not yet settled on that versus the original
305 "bare" representation.
307 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
308 Previously the millisecond value corrupted the output.
309 Fix also for syslog_pid=no and log_selector +pid, for which the pid
310 corrupted the output.
316 JH/01 Rework error string handling in TLS interface so that the caller in
317 more cases is responsible for logging. This permits library-sourced
318 string to be attached to addresses during delivery, and collapses
319 pairs of long lines into single ones.
321 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
322 during configuration. Wildcards are allowed and expanded.
324 JH/02 Rework error string handling in DKIM to pass more info back to callers.
325 This permits better logging.
327 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
328 do not close it down and have the child transport start it up again on
329 the passed-on TCP connection. Instead, proxy the child (and any
330 subsequent ones) for TLS via a unix-domain socket channel. Logging is
331 affected: the continued delivery log lines do not have any DNSSEC, TLS
332 Certificate or OCSP information. TLS cipher information is still logged.
334 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
335 identical IP addresses on different listening ports. Will also affect
338 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
339 add noisy ifdef guards to special-case this sillyness.
340 Patch from Bernd Kuhls.
342 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
343 than 255 are no longer allowed.
345 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
346 Disable the session-cache too, which might reduce our load. Since we
347 currrectly use a new context for every connection, both as server and
348 client, there is no benefit for these.
349 GnuTLS appears to not support tickets server-side by default (we don't
350 call gnutls_session_ticket_enable_server()) but client side is enabled
351 by default on recent versions (3.1.3 +) unless the PFS priority string
354 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
355 <https://reproducible-builds.org/specs/source-date-epoch/>.
357 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
358 the check for any unsuccessful recipients did not notice the limit, and
359 erroneously found still-pending ones.
361 JH/08 Pipeline CHUNKING command and data together, on kernels that support
362 MSG_MORE. Only in-clear (not on TLS connections).
364 JH/09 Avoid using a temporary file during transport using dkim. Unless a
365 transport-filter is involved we can buffer the headers in memory for
366 creating the signature, and read the spool data file once for the
367 signature and again for transmission.
369 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
370 4.77 as the kernel support then wasn't solid, having issues in 64bit
371 mode. Now, it's been long enough. Add support for FreeBSD also.
373 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
374 case where the routing stage had gathered several addresses to send to
375 a host before calling the transport for the first, we previously failed
376 to close down TLS in the old transport process before passing the TCP
377 connection to the new process. The new one sent a STARTTLS command
378 which naturally failed, giving a failed delivery and bloating the retry
379 database. Investigation and fix prototype from Wolfgang Breyha.
381 JH/12 Fix check on SMTP command input synchronisation. Previously there were
382 false-negatives in the check that the sender had not preempted a response
383 or prompt from Exim (running as a server), due to that code's lack of
384 awareness of the SMTP input buffering.
386 PP/04 Add commandline_checks_require_admin option.
387 Exim drops privileges sanely, various checks such as -be aren't a
388 security problem, as long as you trust local users with access to their
389 own account. When invoked by services which pass untrusted data to
390 Exim, this might be an issue. Set this option in main configuration
391 AND make fixes to the calling application, such as using `--` to stop
394 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
395 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
396 responses to those, into a single TLS record each way (this usually means
397 a single packet). As a side issue, smtp_enforce_sync now works on TLS
400 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
401 affects you only if you're dancing at the edge of the param size limits.
402 If you are, and this message makes sense to you, then: raise the
403 configured limit or use OpenSSL 1.1. Nothing we can do for older
406 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
407 cmdline element to get the documented default one. Previously it was
410 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
411 are unsuable for non-message-specific reasons - eg. connection timeout,
412 banner-time rejection.
414 JH/16 Fix logging of delivery remote port, when specified by router, under
417 PP/06 Repair manualroute's ability to take options in any order, even if one
418 is the name of a transport.
421 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
423 JH/17 Change the list-building routines interface to use the expanding-string
424 triplet model, for better allocation and copying behaviour.
426 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
427 Previously it was constructed the first time a possibly-matching string
428 was met in the configuration file input during startup; now it is done
431 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
432 compatible one, to avoid the (poorly documented) possibility of a config
433 file in the working directory redirecting the DB files, possibly correpting
434 some existing file. CVE-2017-10140 assigned for BDB.
436 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
437 cache-hot. Previously, although the result was properly cached, the
438 initial verify call returned a defer.
440 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
441 the main verify for receipient in uncached-mode.
443 JH/22 Retire historical build files to an "unsupported" subdir. These are
444 defined as "ones for which we have no current evidence of testing".
446 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
447 if present. Previously it was ignored.
449 JH/24 Start using specified-initialisers in C structure init coding. This is
450 a C99 feature (it's 2017, so now considered safe).
452 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
453 if was a fixed-sized field and bitmask ops via macros; it is now more
456 PP/07 GitHub PR 56: Apply MariaDB build fix.
457 Patch provided by Jaroslav Škarvada.
459 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
460 during Coverity cleanups [4.87 JH/47]
461 Diagnosis and fix provided by Michael Fischer v. Mollard.
463 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
464 the right size to place the terminating semicolon on its own folded
465 line, the header hash was calculated to an incorrect value thanks to
466 the (relaxed) space the fold became.
468 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
469 and confused the parent.
471 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
472 which could crash as a result. This could lead to undeliverable messages.
474 JH/28 Logging: "next input sent too soon" now shows where input was truncated
477 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
478 matters on fast-turnover and PID-randomising systems, which were getting
479 out-of-order delivery.
481 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
482 a possibly-overlapping copy. The symptom was that "Remote host closed
483 connection in response to HELO" was logged instead of the actual 4xx
486 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
487 Previously only that bufferd was discarded, resulting in SYMTP command
490 JH/32 DKIM: when a message has multiple signatures matching an identity given
491 in dkim_verify_signers, run the dkim acl once for each. Previously only
492 one run was done. Bug 2189.
494 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
495 "panic the current process" to "deliberately defer". The panic log is
496 still written with the problem list name; the mail and reject logs now
497 get a temp-reject line for the message that was being handled, saying
498 something like "domains check lookup or other defer". The SMTP 451
499 message is still "Temporary local problem".
501 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
502 A crafted sequence of BDAT commands could result in in-use memory beeing
503 freed. CVE-2017-16943.
505 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
506 from SMTP input. Previously it was always done; now only done for DATA
507 and not BDAT commands. CVE-2017-16944.
509 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
510 to the message (such as an overlong header line). Previously this was
511 not done and we did not exit BDAT mode. Followon from the previous item
512 though a different problem.
518 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
519 than -2003 did; needs libidn2 in addition to libidn.
521 JH/02 The path option on a pipe transport is now expanded before use.
523 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
524 Patch provided by "Björn", documentation fix added too.
526 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
527 missing a wire-to-host endian conversion.
529 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
530 close after a BDAT command line could be taken as a following command,
531 giving a synch failure. Fix by only checking for synch immediately
532 before acknowledging the chunk.
534 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
535 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
536 macro. Patches provided by Josh Soref.
538 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
539 Previously we did not; the RFC seems ambiguous and VRFY is not listed
540 by IANA as a service extension. However, John Klensin suggests that we
543 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
544 the dkim code may be unix-mode line endings rather than smtp wire-format
545 CRLF, so prepend a CR to any bare LF.
547 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
548 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
550 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
551 the first were themselves being wrongly included in the feed into dkim
552 processing; with most chunk sizes in use this resulted in an incorrect
553 body hash calculated value.
555 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
556 DKIM signature block, for verification. Although advised against by
557 standards it is specifically not ruled illegal.
559 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
561 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
562 missing a body hash (the bh= tag).
564 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
565 It seems that HAProxy sends the Proxy Protocol information in clear and
566 only then does a TLS startup, so do the same.
568 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
569 TCP connections (such as for Spamd) unless the daemon successfully set
570 Fast Open mode on its listening sockets. This fixes breakage seen on
571 too-old kernels or those not configured for Fast Open, at the cost of
572 requiring both directions being enabled for TFO, and TFO never being used
573 by non-daemon-related Exim processes.
575 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
576 endings, at least on the first header line. Try to canonify any that get
577 past that check, despite the cost.
579 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
580 now limited to an arbitrary five deep, while parsing addresses with the
581 strip_excess_angle_brackets option enabled.
583 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
584 instead leave the unprompted TLS handshake in socket buffer for the
585 TLS library to consume.
587 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
589 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
591 JH/16 Drop variables when they go out of scope. Memory management drops a whole
592 region in one operation, for speed, and this leaves assigned pointers
593 dangling. Add checks run only under the testsuite which checks all
594 variables at a store-reset and panics on a dangling pointer; add code
595 explicitly nulling out all the variables discovered. Fixes one known
596 bug: a transport crash, where a dangling pointer for $sending_ip_address
597 originally assigned in a verify callout, is re-used.
599 PP/06 Drop '.' from @INC in various Perl scripts.
601 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
603 PP/08 Reduce a number of compilation warnings under clang; building with
604 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
605 should be warning-free.
607 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
609 HS/01 Fix portability problems introduced by PP/08 for platforms where
610 realloc(NULL) is not equivalent to malloc() [SunOS et al].
612 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
613 chunk. This allows us to accept broken chunked messages. We need a more
614 general solution here.
616 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
617 already-broken messages in the queue.
619 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
621 JH/19 Fix reference counting bug in routing-generated-address tracking.
627 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
628 supports it and a size is available (ie. the sending peer gave us one).
630 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
631 years of being deprecated). The replacements are the ACLs
632 acl_smtp_mime and acl_not_smtp_mime.
634 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
635 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
636 or even an in-clear connection were permitted. Now, if the host lookup
637 was dnssec and dane was requested then the host is only used if the
638 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
639 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
640 if one fails this test.
641 This means that a poorly-configured remote DNS will make it incommunicado;
642 but it protects against a DNS-interception attack on it.
644 JH/04 Bug 1810: make continued-use of an open smtp transport connection
645 non-noisy when a race steals the message being considered.
647 JH/05 If main configuration option tls_certificate is unset, generate a
648 self-signed certificate for inbound TLS connections.
650 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
651 in rewrites and routers.
653 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
654 and logged a warning sing 4.83; now they are a configuration file error.
656 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
657 (lacking @domain). Apply the same qualification processing as RCPT.
659 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
661 JH/10 Support ${sha256:} applied to a string (as well as the previous
664 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
665 a cutthrough deliver is pending, as we always want to make a connection.
666 This also avoids re-routing the message when later placing the cutthrough
667 connection after a verify cache hit.
668 Do not update it with the verify result either.
670 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
671 when routing results in more than one destination address.
673 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
674 signing (which inhibits the cutthrough capability). Previously only
675 the presence of an option was tested; now an expansion evaluating as
676 empty is permissible (obviously it should depend only on data available
677 when the cutthrough connection is made).
679 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
680 the relevant preceding SMTP command did not note the pipelining mode.
682 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
683 Previously they were not counted.
685 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
686 as one having no matching records. Previously we deferred the message
687 that needed the lookup.
689 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
690 distinguished as "(=".
692 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
693 for missing MX records. Previously it only worked for missing A records.
695 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
697 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
698 after the data-go-ahead and data-ack. Patch from Jason Betts.
700 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
701 even for a "none" policy. Patch from Tony Meyer.
703 JH/22 Fix continued use of a connection for further deliveries. If a port was
704 specified by a router, it must also match for the delivery to be
707 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
708 When one of the recipients of a message was unsuitable for the connection
709 (has no matching addresses), we lost track of needing to mark it
710 deferred. As a result mail would be lost.
712 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
714 JH/25 Decoding ACL controls is now done using a binary search; the source code
715 takes up less space and should be simpler to maintain. Merge the ACL
716 condition decode tables also, with similar effect.
718 JH/26 Fix problem with one_time used on a redirect router which returned the
719 parent address unchanged. A retry would see the parent address marked as
720 delivered, so not attempt the (identical) child. As a result mail would
723 JH/27 Fix a possible security hole, wherein a process operating with the Exim
724 UID can gain a root shell. Credit to http://www.halfdog.net/ for
725 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
728 JH/28 Enable {spool,log} filesystem space and inode checks as default.
729 Main config options check_{log,spool}_{inodes,space} are now
730 100 inodes, 10MB unless set otherwise in the configuration.
732 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
733 Previously it only applied to the main-section connection policy
736 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
738 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
739 by me. Added RFC7919 DH primes as an alternative.
741 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
742 are not in the system include path.
744 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
745 GnuTLS, when a session startup failed (eg because the client disconnected)
746 Exim did stdio operations after fclose. This was exposed by a recent
747 change which nulled out the file handle after the fclose.
749 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
750 signed directly by the cert-signing cert, rather than an intermediate
751 OCSP-signing cert. This is the model used by LetsEncrypt.
753 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
755 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
756 an incoming connection.
758 HS/02 Bug 1802: Do not half-close the connection after sending a request
761 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
762 fallback to "prime256v1".
764 JH/34 SECURITY: Use proper copy of DATA command in error message.
765 Could leak key material. Remotely exploitable. CVE-2016-9963.
771 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
772 and 3.4.4 - once the server is enabled to respond to an OCSP request
773 it does even when not requested, resulting in a stapling non-aware
774 client dropping the TLS connection.
776 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
777 support variable-length bit vectors. No functional change.
779 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
780 The I= interface field on outgoing lines is now after the H= remote
781 host field, same as incoming lines. There is a separate
782 outgoing_interface log selector which allows you to disable the
785 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
786 If not running log_selector +smtp_connection the mainlog would be held
787 open indefinitely after a "too many connections" event, including to a
788 deleted file after a log rotate. Leave the per net connection logging
789 leaving it open for efficiency as that will be quickly detected by the
790 check on the next write.
792 HS/01 Bug 1671: Fix post transport crash.
793 Processing the wait-<transport> messages could crash the delivery
794 process if the message IDs didn't exist for some reason. When
795 using 'split_spool_directory=yes' the construction of the spool
796 file name failed already, exposing the same netto behaviour.
798 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
799 mime_regex ACL conditions.
801 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
802 to DSN fail messages (bounces): remote IP, remote greeting, remote response
803 to HELO, local diagnostic string.
805 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
806 log line to debug. Even when configured with a tls authenticator many
807 client connections are expected to not authenticate in this way, so
808 an authenticate fail is not an error.
810 HS/02 Add the Exim version string to the process info. This way exiwhat
811 gives some more detail about the running daemon.
813 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
814 matter for fast-change records such as DNSBLs.
816 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
817 retry record, even if constant. There may be multiple transports with
818 different interface settings and the retry behaviour needs to be kept
821 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
823 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
825 JH/10 Bug 840: fix log_defer_output option of pipe transport
827 JH/11 Bug 830: use same host for all RCPTS of a message, even under
828 hosts_randomize. This matters a lot when combined with mua_wrapper.
830 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
831 ${quote_pgsql:<string>} operator.
833 JH/13 Bug 1708: avoid misaligned access in cached lookup.
835 JH/14 Change header file name for freeradius-client. Relevant if compiling
836 with Radius support; from the Gentoo tree and checked under Fedora.
838 JH/15 Bug 1712: Introduce $prdr_requested flag variable
840 JH/16 Bug 1714: Permit an empty string as expansion result for transport
841 option transport_filter, meaning no filtering.
843 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
845 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
846 defaults to "*" (all hosts). The variable is now available when not built
847 with TLS, default unset, mainly to enable keeping the testsuite sane.
848 If a server certificate is not supplied (via tls_certificate) an error is
849 logged, and clients will find TLS connections fail on startup. Presumably
850 they will retry in-clear.
851 Packagers of Exim are strongly encouraged to create a server certificate
852 at installation time.
854 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
855 with the $config_file variable.
857 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
858 in transport context, after the attempt, and per-recipient. The latter type
859 is per host attempted. The event data is the error message, and the errno
860 information encodes the lookup type (A vs. MX) used for the (first) host,
861 and the trailing two digits of the smtp 4xx response.
863 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
864 to write to mainlog (or rejectlog, paniclog) in the window between file
865 creation and permissions/ownership being changed. Particularly affects
866 installations where exicyclog is run as root, rather than exim user;
867 result is that the running daemon panics and dies.
869 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
871 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
872 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
873 "pri" and "weight". Note that the previous implicit priority given by the
874 list order is no longer honoured.
876 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
879 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
880 by defining SUPPORT_SOCKS.
882 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
883 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
884 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
885 variables are renamed to proxy_{local,external}_{address,port}.
887 JH/27 Move Internationalisation support from Experimental to mainline, enabled
888 for a build by defining SUPPORT_I18N
890 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
891 of the query string, and make ${quote_redis:} do that quoting.
893 JH/29 Move Events support from Experimental to mainline, enabled by default
894 and removable for a build by defining DISABLE_EVENT.
896 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
898 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
899 cached by the daemon.
901 JH/32 Move Redis support from Experimental to mainline, enabled for a build
902 by defining LOOKUP_REDIS. The libhiredis library is required.
904 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
905 keys are given for lookup.
907 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
908 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
909 only supported when built with TLS support. The PolarSSL SHA routines
910 are still used when the TLS library is too old for convenient support.
912 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
913 openssl_options), for security. OpenSSL forces this from version 1.1.0
914 server-side so match that on older versions.
916 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
917 allocation for $value could be released as the expansion processing
918 concluded, but leaving the global pointer active for it.
920 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
921 and to use the domains and local_parts ACL conditions.
923 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
924 incorrectly not doubled on cutthrough transmission, hence seen as a
925 body-termination at the receiving system - resulting in truncated mails.
926 Commonly the sender saw a TCP-level error, and retransmitted the message
927 via the normal store-and-forward channel. This could result in duplicates
928 received - but deduplicating mailstores were liable to retain only the
929 initial truncated version.
931 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
933 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
935 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
936 we're in there, support oversigning also; bug 1309.
938 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
940 HS/04 Add support for keep_environment and add_environment options.
942 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
943 either intentional arithmetic overflow during PRNG, or testing config-
946 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
947 delivery resulted in actual delivery. Cancel cutthrough before DATA
950 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
951 rejects a recipient: pass the reject to the originator.
953 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
954 Many were false-positives and ignorable, but it's worth fixing the
957 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
958 for the new environment-manipulation done at startup. Move the routines
959 from being local to tls.c to being global via the os.c file.
961 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
962 an extract embedded as result-arg for a map, the first arg for extract
963 is unavailable so we cannot tell if this is a numbered or keyed
964 extraction. Accept either.
970 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
973 JH/02 The smtp transport option "multi_domain" is now expanded.
975 JH/03 The smtp transport now requests PRDR by default, if the server offers
978 JH/04 Certificate name checking on server certificates, when exim is a client,
979 is now done by default. The transport option tls_verify_cert_hostnames
980 can be used to disable this per-host. The build option
981 EXPERIMENTAL_CERTNAMES is withdrawn.
983 JH/05 The value of the tls_verify_certificates smtp transport and main options
984 default to the word "system" to access the system default CA bundle.
985 For GnuTLS, only version 3.0.20 or later.
987 JH/06 Verification of the server certificate for a TLS connection is now tried
988 (but not required) by default. The verification status is now logged by
989 default, for both outbound TLS and client-certificate supplying inbound
992 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
995 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
996 Status Notification (bounce) messages are now MIME format per RFC 3464.
997 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
998 under the control of the dsn_advertise_hosts option, and routers may
999 have a dsn_lasthop option.
1001 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1002 default, modifiable by a malware= option. The list separator for
1003 the options can now be changed in the usual way. Bug 68.
1005 JH/10 The smtp_receive_timeout main option is now expanded before use.
1007 JH/11 The incoming_interface log option now also enables logging of the
1008 local interface on delivery outgoing connections.
1010 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1011 if the interface and destination host and port all match.
1013 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1016 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1017 Patch from Andrew Lewis.
1019 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1020 now supports optional time-restrictions, weighting, and priority
1021 modifiers per server. Patch originally by <rommer@active.by>.
1023 JH/16 The spamd_address main option now supports a mixed list of local
1024 and remote servers. Remote servers can be IPv6 addresses, and
1025 specify a port-range.
1027 JH/17 Bug 68: The spamd_address main option now supports an optional
1028 timeout value per server.
1030 JH/18 Bug 1581: Router and transport options headers_add/remove can
1031 now have the list separator specified.
1033 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1036 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1039 JH/21 Support for the A6 type of dns record is withdrawn.
1041 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1042 rather than the verbs used.
1044 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1045 from 255 to 1024 chars.
1047 JH/24 Verification callouts now attempt to use TLS by default.
1049 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1050 are generic router options now. The defaults didn't change.
1052 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1053 Original patch from Alexander Shikoff, worked over by JH.
1055 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1056 files if ZCAT_COMMAND is not executable.
1058 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1060 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1062 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1063 Normally benign, it bites when the pair was led to by a CNAME;
1064 modern usage is to not canonicalize the domain to a CNAME target
1065 (and we were inconsistent anyway for A-only vs AAAA+A).
1067 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1069 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1070 when evaluating $sender_host_dnssec.
1072 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1073 $sender_helo_dnssec variable.
1075 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1077 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1079 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1081 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1082 documented as working, but never had. Support all but $spam_report.
1084 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1085 added for tls authenticator.
1087 HS/03 Add perl_taintmode main config option
1093 TL/01 When running the test suite, the README says that variables such as
1094 no_msglog_check are global and can be placed anywhere in a specific
1095 test's script, however it was observed that placement needed to be near
1096 the beginning for it to behave that way. Changed the runtest perl
1097 script to read through the entire script once to detect and set these
1098 variables, reset to the beginning of the script, and then run through
1099 the script parsing/test process like normal.
1101 TL/02 The BSD's have an arc4random API. One of the functions to induce
1102 adding randomness was arc4random_stir(), but it has been removed in
1103 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1104 function when detected.
1106 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1107 cause callback expansion.
1109 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1110 syntax errors in an expansion can be treated as a string instead of
1111 logging or causing an error, due to the internal use of bool_lax
1112 instead of bool when processing it.
1114 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1115 server certificates when making smtp deliveries.
1117 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1119 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1121 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1123 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1124 Merged patch from Sebastian Wiedenroth.
1126 JH/05 Fix results-pipe from transport process. Several recipients, combined
1127 with certificate use, exposed issues where response data items split
1128 over buffer boundaries were not parsed properly. This eventually
1129 resulted in duplicates being sent. This issue only became common enough
1130 to notice due to the introduction of connection certificate information,
1131 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1133 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1134 size buffer was used, resulting in syntax errors when an expansion
1137 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1138 version 3.3.6 or later.
1140 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1141 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1142 both become "event_action", the variables become $event_name, $event_data
1143 and $event_defer_errno. There is a new variable $verify_mode, usable in
1144 routers, transports and related events. The tls:cert event is now also
1145 raised for inbound connections, if the main configuration event_action
1148 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1149 early OCSP support, but no stapling (appears to be less than 1.0.0).
1151 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1152 server certificate names available under the smtp transport option
1153 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1156 JH/10 Time-related extraction expansions from certificates now use the main
1157 option "timezone" setting for output formatting, and are consistent
1158 between OpenSSL and GnuTLS compilations. Bug 1541.
1160 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1161 encoded parameter in the incoming message. Bug 1558.
1163 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1164 include certificate info, eximon was claiming there were spoolfile
1167 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1169 JH/14 Log delivery-related information more consistently, using the sequence
1170 "H=<name> [<ip>]" wherever possible.
1172 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1173 are problematic for Debian distribution, omit them from the release
1176 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1178 JH/16 Fix string representation of time values on 64bit time_t architectures.
1181 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1182 output list separator was used.
1187 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1188 checkers that were complaining about end of non-void function with no
1191 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1192 This was a regression introduced in 4.83 by another bugfix.
1194 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1196 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1197 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1203 TF/01 Correctly close the server side of TLS when forking for delivery.
1205 When a message was received over SMTP with TLS, Exim failed to clear up
1206 the incoming connection properly after forking off the child process to
1207 deliver the message. In some situations the subsequent outgoing
1208 delivery connection happened to have the same fd number as the incoming
1209 connection previously had. Exim would try to use TLS and fail, logging
1210 a "Bad file descriptor" error.
1212 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1213 utilities have not been installed.
1215 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1216 temporary space as the ACL may create new global variables.
1218 TL/01 LDAP support uses per connection or global context settings, depending
1219 upon the detected version of the libraries at build time.
1221 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1222 to extract and use the src ip:port in logging and expansions as if it
1223 were a direct connection from the outside internet. PPv2 support was
1224 updated based on HAProxy spec change in May 2014.
1226 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1228 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1229 Properly escape header and check for NULL return.
1231 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1234 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1236 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1237 characters in header names, implemented as a verify condition.
1238 Contributed by Michael Fischer v. Mollard.
1240 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1241 results permerror and temperror. Previous values are deprecated but
1242 still accepted. In a future release, err_perm and err_temp will be
1243 completely removed, which will be a backward incompatibility if the
1244 ACL tests for either of these two old results. Patch contributed by
1245 user bes-internal on the mailing list.
1247 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1249 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1250 selectors, in both main and reject logs.
1252 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1255 JH/07 Add malware type "sock" for talking to simple daemon.
1257 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1259 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1260 routers/transports under cutthrough routing.
1262 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1263 numbers. Touch up "bool" conditional to keep the same definition.
1265 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1267 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1269 JH/12 Expand items in router/transport headers_add or headers_remove lists
1270 individually rather than the list as a whole. Bug 1452.
1272 Required for reasonable handling of multiple headers_ options when
1273 they may be empty; requires that headers_remove items with embedded
1274 colons must have them doubled (or the list-separator changed).
1276 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1277 view the policy declared in the DMARC record. Currently, $dmarc_status
1278 is a combined value of both the record presence and the result of the
1281 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1283 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1284 dnslookup router and the smtp transport (applying to the forward
1287 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1288 of ldap servers used for a specific lookup. Patch provided by Heiko
1291 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1292 New variable $lookup_dnssec_authenticated for observability.
1294 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1295 Patch submitted by Lars Timman.
1297 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1299 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1300 Requires trusted mode and valid format message id, aborts otherwise.
1301 Patch contributed by Heiko Schlichting.
1303 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1304 certextract with support for various fields. Bug 1358.
1306 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1307 is requested by default, modifiable by smtp transport option
1310 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1311 operate on certificate variables to give certificate fingerprints
1312 Also new ${sha256:cert_variable}.
1314 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1316 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1317 Christian Aistleitner.
1319 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1321 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1322 file. Patch from Wolfgang Breyha.
1324 JH/25 Expand the coverage of the delivery $host and $host_address to
1325 client authenticators run in verify callout. Bug 1476.
1327 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1328 align with daemon_smtp_ports. Bug 72.
1330 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1331 support and error reporting did not work properly.
1333 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1334 and is readable. Patch from Andrew Colin Kissa.
1336 TL/14 Enhance documentation of ${run expansion and how it parses the
1337 commandline after expansion, particularly in the case when an
1338 unquoted variable expansion results in an empty value.
1340 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1342 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1343 by renaming away from it. Addresses GH issue 15, reported by
1346 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1348 TL/15 SECURITY: prevent double expansion in math comparison functions
1349 (can expand unsanitized data). Not remotely exploitable.
1356 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1358 PP/02 Make -n do something, by making it not do something.
1359 When combined with -bP, the name of an option is not output.
1361 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1364 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1365 $sender_host_name and config options to manage this, and basic check
1368 PP/05 DSCP support for outbound connections and control modifier for inbound.
1370 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1371 (Only plugin which currently uses this is kerberos4, which nobody should
1372 be using, but we should make it available and other future plugins might
1373 conceivably use it, even though it would break NAT; stuff *should* be
1374 using channel bindings instead).
1376 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1377 name; added for Sendmail compatibility; requires admin caller.
1378 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1379 just ignore it); requires trusted caller.
1380 Also parse but ignore: -Ac -Am -X<logfile>
1383 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1385 TL/02 Add +smtp_confirmation as a default logging option.
1387 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1388 Patch by Magnus Holmgren from 2007-02-20.
1390 TL/04 Bugzilla 1281 - Spec typo.
1391 Bugzilla 1283 - Spec typo.
1392 Bugzilla 1290 - Spec grammar fixes.
1394 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1396 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1398 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1399 mailing list by by Dmitry Isaikin.
1401 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1403 JH/02 Support "G" suffix to numbers in ${if comparisons.
1405 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1407 NM/01 Bugzilla 1197 - Spec typo
1408 Bugzilla 1196 - Spec examples corrections
1410 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1412 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1413 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1416 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1417 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1419 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1420 "acl {{name}{arg}...}", and optional args on acl condition
1423 JH/05 Permit multiple router/transport headers_add/remove lines.
1425 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1427 JH/07 Avoid using a waiting database for a single-message-only transport.
1428 Performance patch from Paul Fisher. Bugzilla 1262.
1430 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1433 JH/09 Add $headers_added variable, with content from use of ACL modifier
1434 add_header (but not yet added to the message). Bugzilla 199.
1436 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1437 Pulled from Bugzilla 817 by Wolfgang Breyha.
1439 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1441 (nb: this is the same fix as in Exim 4.80.1)
1443 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1446 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1449 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1451 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1452 advertise SMTP AUTH mechanism to us, instead of a generic
1453 protocol violation error. Also, make Exim more robust to bad
1454 data from the Dovecot auth socket.
1456 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1458 When a queue runner is handling a message, Exim first routes the
1459 recipient addresses, during which it prunes them based on the retry
1460 hints database. After that it attempts to deliver the message to
1461 any remaining recipients. It then updates the hints database using
1464 So if a recipient address works intermittently, it can get repeatedly
1465 deferred at routing time. The retry hints record remains fresh so the
1466 address never reaches the final cutoff time.
1468 This is a fairly common occurrence when a user is bumping up against
1469 their storage quota. Exim had some logic in its local delivery code
1470 to deal with this. However it did not apply to per-recipient defers
1471 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1473 This change adds a proper retry rule check during routing so that the
1474 final cutoff time is checked against the message's age. We only do
1475 this check if there is an address retry record and there is not a
1476 domain retry record; this implies that previous attempts to handle
1477 the address had the retry_use_local_parts option turned on. We use
1478 this as an approximation for the destination being like a local
1479 delivery, as in LMTP.
1481 I suspect this new check makes the old local delivery cutoff check
1482 redundant, but I have not verified this so I left the code in place.
1484 TF/02 Correct gecos expansion when From: is a prefix of the username.
1486 Test 0254 submits a message to Exim with the header
1490 When I ran the test suite under the user fanf2, Exim expanded
1491 the header to contain my full name, whereas it should have added
1492 a Resent-Sender: header. It erroneously treats any prefix of the
1493 username as equal to the username.
1495 This change corrects that bug.
1497 GF/01 DCC debug and logging tidyup
1498 Error conditions log to paniclog rather than rejectlog.
1499 Debug lines prefixed by "DCC: " to remove any ambiguity.
1501 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1503 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1504 Bug spotted by Jeremy Harris; was flawed since initial commit.
1505 Would have resulted in OCSP responses post-SNI triggering an Exim
1506 NULL dereference and crash.
1508 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1510 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1511 Bug detection, analysis and fix by Samuel Thibault.
1512 Bugzilla 1331, Debian bug #698092.
1514 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1516 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1517 Server implementation by Todd Lyons, client by JH.
1518 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1519 config variable "prdr_enable" controls whether the server
1520 advertises the facility. If the client requests PRDR a new
1521 acl_data_smtp_prdr ACL is called once for each recipient, after
1522 the body content is received and before the acl_smtp_data ACL.
1523 The client is controlled by both of: a hosts_try_prdr option
1524 on the smtp transport, and the server advertisement.
1525 Default client logging of deliveries and rejections involving
1526 PRDR are flagged with the string "PRDR".
1528 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1529 fclose(). Diagnosis by Todd Lyons.
1531 PP/17 Update configure.default to handle IPv6 localhost better.
1532 Patch by Alain Williams (plus minor tweaks).
1535 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1536 This is now consistent with GnuTLS, and is now documented: the
1537 previous undocumented portable approach to treating the option as
1538 unset was to force an expansion failure. That still works, and
1539 an empty string is now equivalent.
1541 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1542 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1543 not performing validation itself.
1545 PP/20 Added force_command boolean option to pipe transport.
1546 Patch from Nick Koston, of cPanel Inc.
1548 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1551 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1553 PP/21 Fix eximon continuous updating with timestamped log-files.
1554 Broken in a format-string cleanup in 4.80, missed when I repaired the
1555 other false fix of the same issue.
1556 Report and fix from Heiko Schlichting.
1559 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1560 Report from Prashanth Katuri.
1562 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1563 It's SecureTransport, so affects any MacOS clients which use the
1564 system-integrated TLS libraries, including email clients.
1566 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1567 using a MIME ACL for non-SMTP local injection.
1568 Report and assistance in diagnosis by Warren Baker.
1570 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1572 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1574 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1575 last id that failed so it may be referenced in subsequent ACL's.
1577 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1580 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1581 ldap library initialization, allowing self-signed CA's to be
1582 used. Also properly sets require_cert option later in code by
1583 using NULL (global ldap config) instead of ldap handle (per
1584 session). Bug diagnosis and testing by alxgomz.
1586 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1587 the src/util/ subdirectory.
1589 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1590 renamed to Transport Post Delivery Action by Jeremy Harris, as
1593 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1594 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1595 redis_servers = needs to be configured which will be used by the redis
1596 lookup. Patch from Warren Baker, of The Packet Hub.
1598 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1600 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1601 hostname or reverse DNS when processing a host list. Used suggestions
1602 from multiple comments on this bug.
1604 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1606 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1607 Missed a few lines, added it to make the runtest require no keyboard
1610 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1611 contains upper case chars. Make router use caseful_local_part.
1613 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1614 support when GnuTLS has been built with p11-kit.
1620 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1622 This, or similar/improved, will also be change PP/11 of 4.82.
1628 PP/01 Handle short writes when writing local log-files.
1629 In practice, only affects FreeBSD (8 onwards).
1630 Bugzilla 1053, with thanks to Dmitry Isaikin.
1632 NM/01 Bugzilla 949 - Documentation tweak
1634 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1637 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1639 PP/02 Implemented gsasl authenticator.
1641 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1643 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1644 `pkg-config foo` for cflags/libs.
1646 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1647 with rest of GSASL and with heimdal_gssapi.
1649 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1650 `pkg-config foo` for cflags/libs for the TLS implementation.
1652 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1653 properties get this fed in as external SSF. A number of robustness
1654 and debugging improvements to the cyrus_sasl authenticator.
1656 PP/08 cyrus_sasl server now expands the server_realm option.
1658 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1659 Patch by Jeremy Harris.
1661 PP/10 Added dbmjz lookup type.
1663 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1665 PP/12 MAIL args handles TAB as well as SP, for better interop with
1666 non-compliant senders.
1667 Analysis and variant patch by Todd Lyons.
1669 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1670 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1671 Patch from Dirk Mueller <dmueller@suse.com>
1673 PP/13 tls_peerdn now print-escaped for spool files.
1674 Observed some $tls_peerdn in wild which contained \n, which resulted
1675 in spool file corruption.
1677 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1678 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1679 or write after TLS renegotiation, which otherwise led to messages
1682 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1683 as a tracking header (ie: a signed header comes before the signature).
1684 Patch from Wolfgang Breyha.
1686 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1687 comma-sep list; embedded commas doubled.
1689 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1691 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1693 Report and patch from Dmitry Banschikov.
1695 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1696 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1697 failures. We appear to now support TLS1.1+ with Exim.
1699 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1700 lets Exim select keys and certificates based upon TLS SNI from client.
1701 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1702 before an outbound SMTP session. New log_selector, +tls_sni.
1704 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1705 NULL dereference. Report and patch from Alun Jones.
1707 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1708 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1709 Not seeing resolver debug output on NetBSD, but suspect this is a
1710 resolver implementation change.
1712 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1713 Left warnings. Added "eximon gdb" invocation mode.
1715 PP/21 Defaulting "accept_8bitmime" to true, not false.
1717 PP/22 Added -bw for inetd wait mode support.
1719 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1720 locate the relevant includes and libraries. Made this the default.
1722 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1723 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1725 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1726 This may cause build issues on older platforms.
1728 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1729 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1730 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1731 Added SNI support via GnuTLS too.
1732 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1734 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1736 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1737 Applied second patch from Janne, implementing suggestion to default
1738 multiple-strings-in-record handling to match SPF spec.
1740 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1742 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1743 read-only, out of scope).
1744 Patch from Wolfgang Breyha, report from Stuart Northfield.
1746 PP/29 Fix three issues highlighted by clang analyser static analysis.
1747 Only crash-plausible issue would require the Cambridge-specific
1748 iplookup router and a misconfiguration.
1749 Report from Marcin Mirosław.
1751 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1753 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1754 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1755 As part of this, removing so much warning spew let me fix some minor
1756 real issues in debug logging.
1758 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1759 assignment on my part. Fixed.
1761 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1762 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1763 Janne Snabb (who went above and beyond: thank you).
1765 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1766 string otherwise requires a connection and a bunch more work and it's
1767 relatively easy to get wrong. Should also expose TLS library linkage
1770 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1771 64-bit ${eval} (JH/03).
1773 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1774 GNU libc to support some of the 64-bit stuff, should not lead to
1775 conflicts. Defined before os.h is pulled in, so if a given platform
1776 needs to override this, it can.
1778 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1779 protection layer was required, which is not implemented.
1780 Bugzilla 1254, patch from Wolfgang Breyha.
1782 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1783 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1784 tls_dhparam take prime identifiers. Also unbreak combination of
1785 OpenSSL+DH_params+TLSSNI.
1787 PP/39 Disable SSLv2 by default in OpenSSL support.
1793 PP/01 Solaris build fix for Oracle's LDAP libraries.
1794 Bugzilla 1109, patch from Stephen Usher.
1796 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1798 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1801 TF/02 Fix a couple more cases where we did not log the error message
1802 when unlink() failed. See also change 4.74-TF/03.
1804 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1805 lock up or crash if it happened to be inside a call to libc when it
1806 got a SIGUSR1 from exiwhat.
1808 The SIGUSR1 handler appends the current process status to the process
1809 log which is later printed by exiwhat. It used to use the general
1810 purpose logging code to do this, but several functions it calls are
1811 not safe for signals.
1813 The new output code in the SIGUSR1 handler is specific to the process
1814 log, and simple enough that it's easy to inspect for signal safety.
1815 Removing some special cases also simplifies the general logging code.
1816 Removing the spurious timestamps from the process log simplifies
1819 TF/04 Improved ratelimit ACL condition.
1821 The /noupdate option has been deprecated in favour of /readonly which
1822 has clearer semantics. The /leaky, /strict, and /readonly update modes
1823 are mutually exclusive. The update mode is no longer included in the
1824 database key; it just determines when the database is updated. (This
1825 means that when you upgrade Exim will forget old rate measurements.)
1827 Exim now checks that the per_* options are used with an update mode that
1828 makes sense for the current ACL. For example, when Exim is processing a
1829 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1830 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1831 must specify per_mail/readonly. If you omit the update mode it defaults to
1832 /leaky where that makes sense (as before) or /readonly where required.
1834 The /noupdate option is now undocumented but still supported for
1835 backwards compatibility. It is equivalent to /readonly except that in
1836 ACLs where /readonly is required you may specify /leaky/noupdate or
1837 /strict/noupdate which are treated the same as /readonly.
1839 A useful new feature is the /count= option. This is a generalization
1840 of the per_byte option, so that you can measure the throughput of other
1841 aggregate values. For example, the per_byte option is now equivalent
1842 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1844 The per_rcpt option has been generalized using the /count= mechanism
1845 (though it's more complicated than the per_byte equivalence). When it is
1846 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1847 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1848 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1849 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1850 Note that using per_rcpt with a non-readonly update mode in more than
1851 one ACL will cause the recipients to be double-counted. (The per_mail
1852 and per_byte options don't have this problem.)
1854 The handling of very low rates has changed slightly. If the computed rate
1855 is less than the event's count (usually one) then this event is the first
1856 after a long gap. In this case the rate is set to the same as this event's
1857 count, so that the first message of a spam run is counted properly.
1859 The major new feature is a mechanism for counting the rate of unique
1860 events. The new per_addr option counts the number of different
1861 recipients that someone has sent messages to in the last time period. It
1862 behaves like per_rcpt if all the recipient addresses are different, but
1863 duplicate recipient addresses do not increase the measured rate. Like
1864 the /count= option this is a general mechanism, so the per_addr option
1865 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1866 example, measure the rate that a client uses different sender addresses
1867 with the options per_mail/unique=$sender_address. There are further
1868 details in the main documentation.
1870 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1872 TF/06 Removed a few PCRE remnants.
1874 TF/07 Automatically extract Exim's version number from tags in the git
1875 repository when doing development or release builds.
1877 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1878 Bugzilla 879. Patch from Paul Fisher.
1880 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1881 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1884 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1886 PP/05 Variable $av_failed, true if the AV scanner deferred.
1887 Bugzilla 1078. Patch from John Horne.
1889 PP/06 Stop make process more reliably on build failure.
1890 Bugzilla 1087. Patch from Heiko Schlittermann.
1892 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1893 Bugzilla 1089. Patch from Heiko Schlittermann.
1895 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1896 Bugzilla 1131. Patch from Holger Weiß.
1898 PP/09 Handle IPv6 addresses with SPF.
1899 Bugzilla 860. Patch from Wolfgang Breyha.
1901 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1903 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1906 PP/11 match_* no longer expand right-hand-side by default.
1907 New compile-time build option, EXPAND_LISTMATCH_RHS.
1908 New expansion conditions, "inlist", "inlisti".
1910 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1912 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1914 PP/14 fix log_write() format string regression from TF/03.
1915 Bugzilla 1152. Patch from Dmitry Isaikin.
1921 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1923 PP/02 Harmonised TLS library version reporting; only show if debugging.
1924 Layout now matches that introduced for other libraries in 4.74 PP/03.
1926 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1928 PP/04 New "dns_use_edns0" global option.
1930 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1933 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1934 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1936 TK/01 Updated PolarSSL code to 0.14.2.
1937 Bugzilla 1097. Patch from Andreas Metzler.
1939 PP/07 Catch divide-by-zero in ${eval:...}.
1940 Fixes bugzilla 1102.
1942 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1945 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1946 format-string attack -- SECURITY: remote arbitrary code execution.
1948 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1949 time unintentionally subject to list matching rules, letting the header
1950 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1951 arbitrary string expansion). This allowed for information disclosure.
1953 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1954 INT_MIN/-1 -- value coerced to INT_MAX.
1960 NM/01 Workaround for PCRE version dependency in version reporting
1963 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1964 This fixes portability to compilers other than gcc, notably
1965 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1967 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1968 makefiles for portability to HP-UX and POSIX correctness.
1970 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1971 Also via indented variable definition in the Makefile.
1972 (Debugging by Oliver Heesakkers).
1974 PP/02 Restore caching of spamd results with expanded spamd_address.
1975 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1977 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1978 Improves build reliability. Fix from: Frank Elsner
1980 NM/02 Fix wide character breakage in the rfc2047 coding
1981 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1983 NM/03 Allow underscore in dnslist lookups
1984 Fixes bug 1026. Patch from Graeme Fowler
1986 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1987 Code patches from Adam Ciarcinski of NetBSD.
1989 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1992 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1993 is logged, to avoid truncation. Patch from John Horne.
1995 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1996 Patch from Jakob Hirsch.
1998 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1999 SQL string expansion failure details.
2000 Patch from Andrey Oktyabrski.
2002 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2003 Patch from Simon Arlott.
2005 PP/09 New lookups functionality failed to compile on old gcc which rejects
2006 extern declarations in function scope.
2007 Patch from Oliver Fleischmann
2009 PP/10 Use sig_atomic_t for flags set from signal handlers.
2010 Check getgroups() return and improve debugging.
2011 Fixed developed for diagnosis in bug 927 (which turned out to be
2014 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2015 Patch from Mark Zealey.
2017 PP/12 Bugzilla 1056: Improved spamd server selection.
2018 Patch from Mark Zealey.
2020 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2021 Based on patch from Heiko Schlittermann.
2023 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2024 Patch from Uwe Doering, sign-off by Michael Haardt.
2026 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2027 variable declaration deep within a block. Bug and patch from
2030 PP/15 lookups-Makefile IRIX compatibility coercion.
2032 PP/16 Make DISABLE_DKIM build knob functional.
2034 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2035 Patch by Simon Arlott
2037 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2038 variable argument macros. Our copy now differs from upstream.
2044 TF/01 Failure to get a lock on a hints database can have serious
2045 consequences so log it to the panic log.
2047 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2048 controlled using the smtp_confirmation log selector.
2050 TF/03 Include the error message when we fail to unlink a spool file.
2052 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2053 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2054 for maintaining out-of-tree patches for some time.
2056 PP/01 Bugzilla 139: Documentation and portability issues.
2057 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2058 Handle per-OS dynamic-module compilation flags.
2060 PP/02 Let /dev/null have normal permissions.
2061 The 4.73 fixes were a little too stringent and complained about the
2062 permissions on /dev/null. Exempt it from some checks.
2063 Reported by Andreas M. Kirchwitz.
2065 PP/03 Report version information for many libraries, including
2066 Exim version information for dynamically loaded libraries. Created
2067 version.h, now support a version extension string for distributors
2068 who patch heavily. Dynamic module ABI change.
2070 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2071 privilege escalation vulnerability whereby the Exim run-time user
2072 can cause root to append content of the attacker's choosing to
2075 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2078 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2079 If dropping privileges for untrusted macros, we disabled normal logging
2080 on the basis that it would fail; for the Exim run-time user, this is not
2081 the case, and it resulted in successful deliveries going unlogged.
2082 Fixed. Reported by Andreas Metzler.
2088 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2089 only prepend for the Resent-* case. Fixes regression introduced in
2090 Exim 4.70 by NM/22 for Bugzilla 607.
2092 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2093 increasing numbers of administrators be bitten by this.
2095 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2097 PP/03 Bugzilla 994: added openssl_options main configuration option.
2099 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2101 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2103 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2105 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2106 without a peer certificate, leading to a segfault because of an
2107 assumption that peers always have certificates. Be a little more
2108 paranoid. Problem reported by Martin Tscholak.
2110 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2111 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2112 NB: ClamAV planning to remove STREAM in "middle of 2010".
2113 CL also introduces -bmalware, various -d+acl logging additions and
2114 more caution in buffer sizes.
2116 PP/09 Implemented reverse_ip expansion operator.
2118 PP/10 Bugzilla 937: provide a "debug" ACL control.
2120 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2122 PP/12 Bugzilla 973: Implement --version.
2124 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2126 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2128 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2130 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2131 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2132 ignore trailing whitespace.
2134 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2136 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2139 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2140 Notification from Dr Andrew Aitchison.
2142 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2143 ExtendedDetectionInfo response format.
2144 Notification from John Horne.
2146 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2149 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2150 XSL and documented dependency on system catalogs, with examples of how
2153 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2156 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2157 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2158 configuration file which is writeable by the Exim user or group.
2160 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2161 of configuration files to cover files specified with the -C option if
2162 they are going to be used with root privileges, not just the default
2165 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2166 option (effectively making it always true).
2168 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2169 files to be used while preserving root privileges.
2171 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2172 that rogue child processes cannot use them.
2174 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2175 run-time user, instead of root.
2177 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2178 Exim run-time user without dropping privileges.
2180 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2181 result string, instead of calling string_vformat() twice with the same
2184 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2185 for other users. Others should always drop root privileges if they use
2186 -C on the command line, even for a whitelisted configure file.
2188 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2190 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2196 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2197 $data_path, and $header_path variables; fixed documentation bugs and
2200 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2201 exipick to access non-standard spools, including the "frozen" queue
2204 NM/01 Bugzilla 965: Support mysql stored procedures.
2205 Patch from Alain Williams
2207 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2209 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2210 Patch from Andreas Metzler
2212 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2213 Patch from Kirill Miazine
2215 NM/05 Bugzilla 671: Added umask to procmail example.
2217 JJ/03 installed exipick 20100323.0, fixing doc bug
2219 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2220 directory. Notification and patch from Dan Rosenberg.
2222 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2224 TK/02 Improve log output when DKIM signing operation fails.
2226 MH/01 Treat the transport option dkim_domain as a colon separated
2227 list, not as a single string, and sign the message with each element,
2228 omitting multiple occurences of the same signer.
2230 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2231 Bugzilla 985, 986. Patch by Simon Arlott
2233 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2234 Patch by Simon Arlott
2236 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2237 MBX locking. Notification from Dan Rosenberg.
2243 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2245 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2247 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2249 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2251 NM/04 Bugzilla 915: Flag broken perl installation during build.
2257 TK/01 Added patch by Johannes Berg that expands the main option
2258 "spamd_address" if it starts with a dollar sign.
2260 TK/02 Write list of recipients to X-Envelope-Sender header when building
2261 the mbox-format spool file for content scanning (suggested by Jakob
2264 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2265 (http://www.dcc-servers.net/) support via dccifd. Activated by
2266 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2268 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2269 by Mark Daniel Reidel <mr@df.eu>.
2271 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2272 When building exim an external PCRE library is now needed -
2273 PCRE is a system library on the majority of modern systems.
2274 See entry on PCRE_LIBS in EDITME file.
2276 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2277 conversation. Added nologin parameter to request.
2278 Patch contributed by Kirill Miazine.
2280 TF/01 Do not log submission mode rewrites if they do not change the address.
2282 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2284 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2285 log files in place. Contributed by Roberto Lima.
2287 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2289 TF/03 Bugzilla 615: When checking the local_parts router precondition
2290 after a local_part_suffix or local_part_prefix option, Exim now
2291 does not use the address's named list lookup cache, since this
2292 contains cached lookups for the whole local part.
2294 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2295 Robert Millan. Documentation is in experimental-spec.txt.
2297 TF/04 Bugzilla 668: Fix parallel build (make -j).
2299 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2301 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2302 Patch provided by Jan Srzednicki.
2304 TF/05 Leading white space used to be stripped from $spam_report which
2305 wrecked the formatting. Now it is preserved.
2307 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2308 that they are available at delivery time.
2310 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2312 TF/08 TLS error reporting now respects the incoming_interface and
2313 incoming_port log selectors.
2315 TF/09 Produce a more useful error message if an SMTP transport's hosts
2316 setting expands to an empty string.
2318 NM/06 Bugzilla 744: EXPN did not work under TLS.
2319 Patch provided by Phil Pennock.
2321 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2322 Patch provided by Richard Godbee.
2324 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2325 acl_smtp_notquit, added index entry.
2327 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2328 Patch provided by Eugene Bujak.
2330 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2331 accept(). Patch provided by Maxim Dounin.
2333 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2334 Patch provided by Phil Pennock.
2336 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2338 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2339 Patch provided by Brad "anomie" Jorsch.
2341 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2342 Patch provided by Dean Brooks.
2344 TK/05 Add native DKIM support (does not depend on external libraries).
2346 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2347 Patch provided by Graeme Fowler.
2349 NM/16 Bugzilla 851: Documentation example syntax fix.
2351 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2353 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2356 NM/19 Bugzilla 745: TLS version reporting.
2357 Patch provided by Phil Pennock.
2359 NM/20 Bugzilla 167: bool: condition support.
2360 Patch provided by Phil Pennock.
2362 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2363 clients. Patch provided by Phil Pennock.
2365 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2366 Patch provided by Brad "anomie" Jorsch.
2368 NM/23 Bugzilla 687: Fix misparses in eximstats.
2369 Patch provided by Heiko Schlittermann.
2371 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2372 Patch provided by Heiko Schlittermann.
2374 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2375 plus update to original patch.
2377 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2379 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2380 Patch provided by David Brownlee.
2382 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2384 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2386 NM/30 Bugzilla 888: TLS documentation bugfixes.
2388 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2390 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2391 Unlike the original bugzilla I have changed all shell scripts in src tree.
2393 NM/33 Bugzilla 898: Transport filter timeout fix.
2394 Patch by Todd Rinaldo.
2396 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2397 Patch by Serge Demonchaux.
2399 NM/35 Bugzilla 39: Base64 decode bug fixes.
2400 Patch by Jakob Hirsch.
2402 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2404 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2406 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2408 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2414 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2415 ALT-N's libdkim that I have put here:
2416 http://duncanthrax.net/exim-experimental/
2418 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2419 were called 'true' and it seems that C99 defines that as a reserved
2420 keyword to be used with 'bool' variable types. That means you could
2421 not include C99-style headers which use bools without triggering
2422 build errors in sieve.c.
2424 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2425 as mailq or other aliases. Changed the --help handling significantly
2426 to do whats expected. exim_usage() emits usage/help information.
2428 SC/01 Added the -bylocaldomain option to eximstats.
2430 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2432 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2434 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2440 PH/01 Another patch from the Sieve maintainer.
2442 PH/02 When an IPv6 address is converted to a string for single-key lookup
2443 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2444 dots are used instead of colons so that keys in lsearch files need not
2445 contain colons. This was done some time before quoting was made available
2446 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2447 (notated using the quote facility) so as to distinguish them from IPv4
2448 keys. This meant that lookups for IP addresses in host lists did not work
2449 for iplsearch lookups.
2451 This has been fixed by arranging for IPv6 addresses to be expressed with
2452 colons if the lookup type is iplsearch. This is not incompatible, because
2453 previously such lookups could never work.
2455 The situation is now rather anomalous, since one *can* have colons in
2456 ordinary lsearch keys. However, making the change in all cases is
2457 incompatible and would probably break a number of configurations.
2459 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2462 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2463 conversion specification without a maximum field width, thereby enabling
2464 a rogue spamd server to cause a buffer overflow. While nobody in their
2465 right mind would setup Exim to query an untrusted spamd server, an
2466 attacker that gains access to a server running spamd could potentially
2467 exploit this vulnerability to run arbitrary code as the Exim user.
2469 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2470 $primary_hostname instead of what libspf2 thinks the hosts name is.
2472 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2473 a directory entry by the name of the lookup key. Previously, if a
2474 symlink pointed to a non-existing file or a file in a directory that
2475 Exim lacked permissions to read, a lookup for a key matching that
2476 symlink would fail. Now it is enough that a matching directory entry
2477 exists, symlink or not. (Bugzilla 503.)
2479 PH/03 The body_linecount and body_zerocount variables are now exported in the
2482 PH/04 Added the $dnslist_matched variable.
2484 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2485 This means they are set thereafter only if the connection becomes
2488 PH/06 Added the client_condition to authenticators so that some can be skipped
2489 by clients under certain conditions.
2491 PH/07 The error message for a badly-placed control=no_multiline_responses left
2492 "_responses" off the end of the name.
2494 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2496 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2497 (without spaces) instead of just copying the configuration text.
2499 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2501 PH/11 Added $max_received_linelength.
2503 PH/12 Added +ignore_defer and +include_defer to host lists.
2505 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2506 way in which PCRE > 7.0 is built.
2508 PH/14 Implemented queue_only_load_latch.
2510 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2511 MAIL command. The effect was to mangle the value on 64-bit systems.
2513 PH/16 Another patch from the Sieve maintainer.
2515 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2517 PH/18 If a system quota error occurred while trying to create the file for
2518 a maildir delivery, the message "Mailbox is full" was not appended to the
2519 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2520 only to a quota excession during the actual writing of the file.
2522 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2523 characters?) which causes problems in log lines. The DN values are now
2524 passed through string_printing() before being added to log lines.
2526 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2527 and InterBase are left for another time.)
2529 PH/21 Added message_body_newlines option.
2531 PH/22 Guard against possible overflow in moan_check_errorcopy().
2533 PH/23 POSIX allows open() to be a macro; guard against that.
2535 PH/24 If the recipient of an error message contained an @ in the local part
2536 (suitably quoted, of course), incorrect values were put in $domain and
2537 $local_part during the evaluation of errors_copy.
2543 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2544 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2547 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2548 issue a MAIL command.
2550 PH/02 In an ACL statement such as
2552 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2554 if a client was not listed at all, or was listed with a value other than
2555 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2556 the condition was not true (as it should be), so access was not denied.
2557 The bug was that the ! inversion was incorrectly passed on to the second
2558 item. This has been fixed.
2560 PH/03 Added additional dnslists conditions == and =& which are different from
2561 = and & when the dns lookup returns more than one IP address.
2563 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2564 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2566 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2567 FSYNC, which compiles an option called disable_fsync that allows for
2568 bypassing fsync(). The documentation is heavily laced with warnings.
2570 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2572 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2573 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2574 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2575 including adding "make clean"; (3) Added -fPIC when compiling the test
2576 dynamically loaded module, to get rid of a warning.
2578 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2579 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2580 = 0s. The bug is otherwise harmless.
2582 PH/07 There was a bug in the dovecot authenticator such that the value of
2583 $auth1 could be overwritten, and so not correctly preserved, after a
2584 successful authentication. This usually meant that the value preserved by
2585 the server_setid option was incorrect.
2587 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2589 PH/09 Installed PCRE release 7.0.
2591 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2592 run for batched SMTP input. It is now run at the start of every message
2593 in the batch. While fixing this I discovered that the process information
2594 (output by running exiwhat) was not always getting set for -bs and -bS
2595 input. This is fixed, and it now also says "batched" for BSMTP.
2597 PH/11 Added control=no_pipelining.
2599 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2600 patch, slightly modified), and move the expansion of helo_data till after
2601 the connection is made in the smtp transport (so it can use these
2604 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2606 PH/14 Added log_selector = +pid.
2608 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2610 PH/16 Add ${if forany and ${if forall.
2612 PH/17 Added dsn_from option to vary the From: line in DSNs.
2614 PH/18 Flush SMTP output before performing a callout, unless control =
2615 no_callout_flush is set.
2617 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2618 was true (the default) a successful delivery failed to delete the retry
2619 item, thus causing premature timeout of the address. The bug is now
2622 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2624 PH/21 Long custom messages for fakedefer and fakereject are now split up
2625 into multiline responses in the same way that messages for "deny" and
2626 other ACL rejections are.
2628 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2629 with slight modification.
2631 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2632 draft, changing the syntax and factoring some duplicate code".
2634 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2635 for deliveries of the second and subsequent messages over the same SMTP
2638 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2639 ${reduce, with only minor "tidies".
2641 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2643 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2644 expansion side effects.
2646 PH/27 When a message times out after an over-quota error from an Exim-imposed
2647 quota, the bounce message says "mailbox is full". This message was not
2648 being given when it was a system quota that was exceeded. It now should
2651 MH/03 Made $recipients available in local_scan(). local_scan() already has
2652 better access to the recipient list through recipients_list[], but
2653 $recipients can be useful in postmaster-provided expansion strings.
2655 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2656 in the case of a MAIL command with additional options following the
2657 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2658 were accidentally chopped off.
2660 PH/29 SMTP synchronization checks are implemented when a command is read -
2661 there is a check that no more input is waiting when there shouldn't be
2662 any. However, for some commands, a delay in an ACL can mean that it is
2663 some time before the response is written. In this time, more input might
2664 arrive, invalidly. So now there are extra checks after an ACL has run for
2665 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2666 pipelining has not been advertised.
2668 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2670 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2671 returned in the SMTP response when smtp_return_error_details was false.
2672 This has been fixed.
2674 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2675 instead of the C I/O that was originally supplied, because problems were
2676 reported on Solaris.
2678 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2679 Exim which did not show up earlier: it was assuming that a call to
2680 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2681 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2682 was a macro that became an assignment, so it seemed to work. This has
2683 changed to a proper function call with a void return, hence the compile
2684 error. Exim's code has been fixed.
2686 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2689 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2691 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2693 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2694 because it turns out to be too big or there is a timeout) there is no
2695 "Completed" line in the log. When some messages of this type were
2696 selected by exigrep, they were listed as "not completed". Others were
2697 picked up by some special patterns. I have improved the selection
2698 criteria to be more general.
2700 PH/38 The host_find_failed option in the manualroute router can now be set
2701 to "ignore", to completely ignore a host whose IP address cannot be
2702 found. If all hosts are ignored, the behaviour is controlled by the new
2703 host_all_ignored option.
2705 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2706 homing or because of multiple MX records with /mx) generated more than
2707 one IP address, and the following item turned out to be the local host,
2708 all the secondary addresses of the first item were incorrectly removed
2709 from the list, along with the local host and any following hosts (which
2710 is what is supposed to happen).
2712 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2713 whoever called Exim into the -H file. In the case of the daemon it was
2714 behaving confusingly. When first started, it used values for whoever
2715 started the daemon, but after a SIGHUP it used the Exim user (because it
2716 calls itself on a restart). I have changed the code so that it now always
2719 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2720 message are rejected with the same error (e.g. no authentication or bad
2721 sender address), and a DATA command is nevertheless sent (as can happen
2722 with PIPELINING or a stupid MUA), the error message that was given to the
2723 RCPT commands is included in the rejection of the DATA command. This is
2724 intended to be helpful for MUAs that show only the final error to their
2727 PH/42 Another patch from the Sieve maintainer.
2729 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2730 Eximstats - Fixed some broken HTML links and added missing column headers
2732 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2733 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2735 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2737 PH/43 Yet another patch from the Sieve maintainer.
2739 PH/44 I found a way to check for a TCP/IP connection going away before sending
2740 the response to the final '.' that terminates a message, but only in the
2741 case where the client has not sent further data following the '.'
2742 (unfortunately, this is allowed). However, in many cases there won't be
2743 any further data because there won't be any more messages to send. A call
2744 to select() can be used: if it shows that the input is "ready", there is
2745 either input waiting, or the socket has been closed. An attempt to read
2746 the next input character can distinguish the two cases. Previously, Exim
2747 would have sent an OK response which the client would never have see.
2748 This could lead to message repetition. This fix should cure that, at
2749 least in a lot of common cases.
2751 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2752 advertised in response to EHLO.
2758 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2759 fixed by 4.65/MH/01 (is this a record?) are fixed:
2761 (i) An empty string was always treated as zero by the numeric comparison
2762 operators. This behaviour has been restored.
2764 (ii) It is documented that the numeric comparison operators always treat
2765 their arguments as decimal numbers. This was broken in that numbers
2766 starting with 0 were being interpreted as octal.
2768 While fixing these problems I realized that there was another issue that
2769 hadn't been noticed. Values of message_size_limit (both the global option
2770 and the transport option) were treated as octal if they started with 0.
2771 The documentation was vague. These values are now always treated as
2772 decimal, and I will make that clear in the documentation.
2778 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2779 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2782 MH/01 Don't check that the operands of numeric comparison operators are
2783 integers when their expansion is in "skipping" mode (fixes bug
2784 introduced by 4.64-PH/07).
2786 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2787 child addresses, Exim now panics and dies. Previously, because the count
2788 is held in a short int, deliveries were likely to be lost. As such a
2789 large number of recipients for a single message is ridiculous
2790 (performance will be very, very poor), I have chosen to impose a limit
2791 rather than extend the field.
2797 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2798 leftover -K file (the existence of which was triggered by #402).
2799 While we were at it, introduced process PID as part of the -K
2800 filename. This should rule out race conditions when creating
2803 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2804 processing considerably. Previous code took too long for large mails,
2805 triggering a timeout which in turn triggers #401.
2807 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2808 in the DK code in transports.c. sendfile() is not really portable,
2809 hence the _LINUX specificness.
2811 TF/01 In the add_headers option to the mail command in an Exim filter,
2812 there was a bug that Exim would claim a syntax error in any
2813 header after the first one which had an odd number of characters
2816 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2817 callout verification, Exim cached a "reject" for the entire domain. This
2818 is correct for most verifications, but it is not correct for a recipient
2819 verification with use_sender or use_postmaster set, because in that case
2820 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2821 case of MAIL FROM:<> rejection from other early rejections (e.g.
2822 rejection of HELO). When verifying a recipient using a non-null MAIL
2823 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2824 Whatever the result of the callout, the value of the domain cache is
2825 left unchanged (for any other kind of callout, getting as far as trying
2826 RCPT means that the domain itself is ok).
2828 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2831 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2832 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2833 session, the daemon ignores ECONNECT errors and logs others; it now
2834 ignores EPIPE as well.
2836 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2837 (quoted-printable decoding).
2839 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2840 later the small subsequent patch to fix an introduced bug.
2842 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2844 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2846 PH/08 An error is now given if message_size_limit is specified negative.
2848 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2849 to be given (somewhat) arbitrary names.
2851 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2854 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2855 miscellaneous code fixes
2857 PH/10 Added the log_reject_target ACL modifier to specify where to log
2860 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2861 hostname. This is wrong, because it relates to the incoming message (and
2862 probably the interface on which it is arriving) and not to the outgoing
2863 callout (which could be using a different interface). This has been
2864 changed to use the value of the helo_data option from the smtp transport
2865 instead - this is what is used when a message is actually being sent. If
2866 there is no remote transport (possible with a router that sets up host
2867 addresses), $smtp_active_hostname is used.
2869 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2870 tweaks were necessary in order to get it to work (see also 21 below):
2871 (a) The code assumed that strncpy() returns a negative number on buffer
2872 overflow, which isn't the case. Replaced with Exim's string_format()
2874 (b) There were several signed/unsigned issues. I just did the minimum
2875 hacking in of casts. There is scope for a larger refactoring.
2876 (c) The code used strcasecmp() which is not a standard C function.
2877 Replaced with Exim's strcmpic() function.
2878 (d) The code set only $1; it now sets $auth1 as well.
2879 (e) A simple test gave the error "authentication client didn't specify
2880 service in request". It would seem that Dovecot has changed its
2881 interface. Fortunately there's a specification; I followed it and
2882 changed what the client sends and it appears to be working now.
2884 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2887 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2888 address A is aliased to B and C, where B exists and C does not. Without
2889 -v the output is "A verified" because verification stops after a
2890 successful redirection if more than one address is generated. However,
2891 with -v the child addresses are also verified. Exim was outputting "A
2892 failed to verify" and then showing the successful verification for C,
2893 with its parentage. It now outputs "B failed to verify", showing B's
2894 parentage before showing the successful verification of C.
2896 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2897 look up a TXT record in a specific list after matching in a combined
2900 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2901 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2902 they consult the DNS. I had assumed they would set it the way they
2903 wanted; and indeed my experiments on Linux seem to show that in some
2904 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2905 To be on the safe side, however, I have now made the interface to
2906 host_find_byname() similar to host_find_bydns(), with an argument
2907 containing the DNS resolver options. The host_find_byname() function now
2908 sets these options at its start, just as host_find_bydns() does. The smtp
2909 transport options dns_qualify_single and dns_search_parents are passed to
2910 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2911 of host_find_byname() use the default settings of RES_DEFNAMES
2912 (qualify_single) but not RES_DNSRCH (search_parents).
2914 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2915 spool_read_header() do less string testing, by means of a preliminary
2916 switch on the second character of optional "-foo" lines. (This is
2917 overdue, caused by the large number of possibilities that now exist.
2918 Originally there were few.) While I was there, I also converted the
2919 str(n)cmp tests so they don't re-test the leading "-" and the first
2920 character, in the hope this might squeeze out yet more improvement.
2922 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2923 flag allowing group syntax was set by the header_syntax check but not
2924 turned off, possible causing trouble later; (2) The flag was not being
2925 set at all for the header_verify test, causing "group"-style headers to
2926 be rejected. I have now set it in this case, and also caused header_
2927 verify to ignore an empty address taken from a group. While doing this, I
2928 came across some other cases where the code for allowing group syntax
2929 while scanning a header line wasn't quite right (mostly, not resetting
2930 the flag correctly in the right place). These bugs could have caused
2931 trouble for malformed header lines. I hope it is now all correct.
2933 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2934 with the "reply" argument non-NULL. The code, however (which originally
2935 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2936 but it didn't always do it. This confused somebody who was copying the
2937 code for some other use. I have removed all the tests.
2939 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2940 feature that was used to support insecure browsers during the U.S. crypto
2941 embargo. It requires special client support, and Exim is probably the
2942 only MTA that supported it -- and would never use it because real RSA is
2943 always available. This code has been removed, because it had the bad
2944 effect of slowing Exim down by computing (never used) parameters for the
2945 RSA_EXPORT functionality.
2947 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2948 authenticator to fail if there's a tab character in the incoming data
2949 (there should never be unless someone is messing about, as it's supposed
2950 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2951 if the connection is using TLS or if the remote IP is the same as the
2952 local IP, and the "valid-client-cert option" if a client certificate has
2955 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2956 authenticators. This can be used for authorization after authentication
2957 succeeds. (In the case of plaintext, it servers for both authentication
2960 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2961 if any retry times were supplied.
2963 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2964 connection, where there is no client IP address to check. In this
2965 situation, the verify now always succeeds.
2967 PH/25 Applied John Jetmore's -Mset patch.
2969 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2971 PH/27 In a string expansion for a processed (not raw) header when multiple
2972 headers of the same name were present, leading whitespace was being
2973 removed from all of them, but trailing whitespace was being removed only
2974 from the last one. Now trailing whitespace is removed from each header
2975 before concatenation. Completely empty headers in a concatenation (as
2976 before) are ignored.
2978 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2979 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2981 PH/29 [Removed. This was a change that I later backed out, and forgot to
2982 correct the ChangeLog entry (that I had efficiently created) before
2983 committing the later change.]
2985 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2986 address errors (4xx response to RCPT) over the same connection as other
2987 messages routed to the same hosts. Such deliveries are always "forced",
2988 so retry times are not inspected. This resulted in far too many retries
2989 for the affected addresses. The effect occurred only when there were more
2990 hosts than the hosts_max_try setting in the smtp transport when it had
2991 the 4xx errors. Those hosts that it had tried were not added to the list
2992 of hosts for which the message was waiting, so if all were tried, there
2993 was no problem. Two fixes have been applied:
2995 (i) If there are any address or message errors in an SMTP delivery, none
2996 of the hosts (tried or untried) are now added to the list of hosts
2997 for which the message is waiting, so the message should not be a
2998 candidate for sending over the same connection that was used for a
2999 successful delivery of some other message. This seems entirely
3000 reasonable: after all the message is NOT "waiting for some host".
3001 This is so "obvious" that I'm not sure why it wasn't done
3002 previously. Hope I haven't missed anything, but it can't do any
3003 harm, as the worst effect is to miss an optimization.
3005 (ii) If, despite (i), such a delivery is accidentally attempted, the
3006 routing retry time is respected, so at least it doesn't keep
3007 hammering the server.
3009 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3010 in ${readsocket because some servers need this prod.
3012 PH/32 Added some extra debug output when updating a wait-xxx database.
3014 PH/33 The hint "could be header name not terminated by colon", which has been
3015 given for certain expansion errors for a long time, was not being given
3016 for the ${if def:h_colon_omitted{... case.
3018 PH/34 The spec says: "With one important exception, whenever a domain list is
3019 being scanned, $domain contains the subject domain." There was at least
3020 one case where this was not true.
3022 PH/35 The error "getsockname() failed: connection reset by peer" was being
3023 written to the panic log as well as the main log, but it isn't really
3024 panic-worthy as it just means the connection died rather early on. I have
3025 removed the panic log writing for the ECONNRESET error when getsockname()
3028 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3029 runs only) independently of the message's sender address. This meant
3030 that, if the 4xx error was in fact related to the sender, a different
3031 message to the same recipient with a different sender could confuse
3032 things. In particular, this can happen when sending to a greylisting
3033 server, but other circumstances could also provoke similar problems.
3034 I have changed the default so that the retry time for these errors is now
3035 based a combination of the sender and recipient addresses. This change
3036 can be overridden by setting address_retry_include_sender=false in the
3039 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3040 remote server are returned as part of bounce messages. This was not
3041 happening for LMTP over a pipe (the lmtp transport), but now it is the
3042 same for both kinds of LMTP.
3044 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3045 in header lines that were in fact CNAMEs. This is no longer the case.
3047 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3048 and queue runs started by the daemon processed all messages. This has
3049 been fixed so that -R and -S can now usefully be given with -q<time>.
3051 PH/40 Import PCRE release 6.7 (fixes some bugs).
3053 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3055 PH/42 Give an error if -q is specified more than once.
3057 PH/43 Renamed the variables $interface_address and $interface_port as
3058 $received_ip_address and $received_port, to make it clear that these
3059 values apply to message reception, and not to the outgoing interface when
3060 a message is delivered. (The old names remain recognized, of course.)
3062 PH/44 There was no timeout on the connect() call when using a Unix domain
3063 socket in the ${readsocket expansion. There now is.
3065 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3066 be meaningful with "accept".
3068 SC/01 Eximstats V1.43
3069 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3071 SC/02 Eximstats V1.44
3072 Use a glob alias rather than an array ref in the generated
3073 parser. This improves both readability and performance.
3075 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3076 Collect SpamAssassin and rejection statistics.
3077 Don't display local sender or destination tables unless
3078 there is data to show.
3079 Added average volumes into the top table text output.
3081 SC/04 Eximstats V1.46
3082 Collect data on the number of addresses (recipients)
3083 as well as the number of messages.
3085 SC/05 Eximstats V1.47
3086 Added 'Message too big' to the list of mail rejection
3087 reasons (thanks to Marco Gaiarin).
3089 SC/06 Eximstats V1.48
3090 Mainlog lines which have GMT offsets and are too short to
3091 have a flag are now skipped.
3093 SC/07 Eximstats V1.49 (Alain Williams)
3094 Added the -emptyok flag.
3096 SC/08 Eximstats V1.50
3097 Fixes for obtaining the IP address from reject messages.
3099 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3100 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3101 whitespace changes from 4.64-PH/27
3103 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3106 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3107 are found, allow negative numbers in numeric criteria)
3109 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3111 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3112 to match changes made in 4.64-PH/43
3114 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3116 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3117 "domain" is actually an IP address, and force a failure. This locks out
3118 those revolvers/nameservers that support "A-for-A" lookups, in
3119 contravention of the specifications.
3121 PH/48 When a host name was looked up from an IP address, and the subsequent
3122 forward lookup of the name timed out, the host name was left in
3123 $sender_host_name, contrary to the specification.
3125 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3126 restricted to single-key lookups, Exim was not diagnosing an error if
3127 * or *@ was used with a query-style lookup.
3129 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3131 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3132 long ago, but noone interested enough thought of it. Let's just say that
3133 the "1.1" means that there are some new functions that weren't there at
3134 some point in the past.
3136 PH/51 Error processing for expansion failure of helo_data from an smtp
3137 transport during callout processing was broken.
3139 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3140 tested/used via the -bh/-bhc/-bs options.
3142 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3143 bug, fixed in subsequent PCRE releases).
3145 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3146 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3148 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3154 SC/01 Use a glob alias rather than an array ref in eximstats generated
3155 parser. This improves both readability and performance.
3157 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3158 Don't display local sender or destination tables in eximstats unless
3159 there is data to show.
3160 Added average volumes into the eximstats top table text output.
3162 SC/03 Collect data on the number of addresses (recipients) as well
3163 as the number of messages in eximstats.
3165 TF/01 Correct an error in the documentation for the redirect router. Exim
3166 does (usually) call initgroups() when daemonizing.
3168 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3169 with consistent privilege compared to when running as a daemon.
3171 TF/03 Note in the spec that $authenticated_id is not set for local
3172 submissions from trusted users.
3174 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3175 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3177 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3178 by adding some example configuration directives to the default
3179 configuration file. A little bit of work is required to uncomment the
3180 directives and define how usernames and passwords are checked, but
3181 there is now a framework to start from.
3183 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3184 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3185 without this. I don't know how relevant this is to other LDAP libraries.
3187 PH/02 Add the verb name to the "unknown ACL verb" error.
3189 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3191 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3193 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3194 directory not expanded when it should be if an expanded home directory
3195 was set for the address (which is overridden by the transport).
3197 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3200 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3201 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3202 because it is too late at that time, and has no effect.
3204 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3205 security issue with \' (bugzilla #107). I could not use the
3206 PQescapeStringConn() function, because it needs a PGconn value as one of
3209 PH/08 When testing addresses using -bt, indicate those final addresses that
3210 are duplicates that would not cause an additional delivery. At least one
3211 person was confused, thinking that -bt output corresponded to deliveries.
3212 (Suppressing duplicates isn't a good idea as you lose the information
3213 about possibly different redirections that led to the duplicates.)
3215 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3216 systems where poll() doesn't work, in particular OS X.
3218 PH/10 Added more information to debugging output for retry time not reached.
3220 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3221 operations in malware.c.
3223 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3226 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3227 syslog_duplication set false, log lines that would normally be written
3228 both the the main log and to the reject log were not written to syslog at
3231 PH/14 In the default configuration, change the use of "message" in ACL warn
3232 statements to "add_header".
3234 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3235 not followed by a command (e.g. "seen endif").
3237 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3238 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3241 PH/17 Added extra conditions to the default value of delay_warning_condition
3245 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3246 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3247 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3250 The Auto-Submitted: and various List- headers are standardised, whereas I
3251 don't think Precedence: ever was.
3253 PH/18 Refactored debugging code in route_finduser() to show more information,
3254 in particular, the error code if getpwnam() issues one.
3256 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3257 This is apparently needed in addition to the PH/07 change above to avoid
3258 any possible encoding problems.
3260 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3261 but not after initializing Perl.
3263 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3264 output them only if debugging. By default they are written stderr,
3265 apparently, which is not desirable.
3267 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3270 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3273 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3275 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3276 authenticated or an ident call has been made. Suppress the default
3277 values for $authenticated_id and $authenticated_sender (but permit -oMai
3278 and -oMas) when testing with -bh.
3280 PH/24 Re-jigged the order of the tests in the default configuration so that the
3281 tests for valid domains and recipients precede the DNS black list and CSA
3282 tests, on the grounds that those ones are more expensive.
3284 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3285 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3286 This bug exists in every version of Exim that I still have, right back to
3289 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3290 However, an attempt to turn on case-sensitivity in a regex key by
3291 including (?-i) didn't work because the subject string was already
3292 lowercased, and the effects were non-intuitive. It turns out that a
3293 one-line patch can be used to allow (?-i) to work as expected.
3299 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3300 other effects) broke the use of negated acl sub-conditions.
3302 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3305 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3306 "Deny" causes Exim to reject the incoming connection with a 554 error.
3307 Unfortunately, if there is a major crisis, such as a disk failure,
3308 tcp-wrappers gives "deny", whereas what one would like would be some
3309 kind of temporary error. A kludge has been added to help with this.
3310 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3311 554 error is used if errno is still zero or contains ENOENT (which occurs
3312 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3315 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3317 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3318 errors. Otherwise a message that provokes a temporary error (when other
3319 messages do not) can cause a whole host to time out.
3321 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3322 addresses were routed directly to files or pipes from a redirect router.
3323 File deliveries just didn't batch; pipe deliveries might have suffered
3326 PH/06 A failure to get a lock for a hints database would erroneously always say
3327 "Failed to get write lock", even when it was really a read lock.
3329 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3330 of 0600. This has been changed to use the value of the lockfile_mode
3331 option (which defaults to 0600).
3333 PH/08 Applied small patch from the Sieve maintainer.
3335 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3336 folder from quota calculations, a direct delivery into this folder messed
3337 up the contents of the maildirsize file. This was because the regex was
3338 used only to exclude .Trash (or whatever) when the size of the mailbox
3339 was calculated. There was no check that a delivery was happening into an
3340 excluded directory. This bug has been fixed by ignoring all quota
3341 processing for deliveries into excluded directories.
3343 PH/10 Added the maildirfolder_create_regex option to appendfile.
3349 PH/01 The code for finding all the local interface addresses on a FreeBSD
3350 system running IPv6 was broken. This may well have applied to all BSD
3351 systems, as well as to others that have similar system calls. The broken
3352 code found IPv4 interfaces correctly, but gave incorrect values for the
3353 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3354 that it would not match correctly against @[] and not recognize the IPv6
3357 PH/02 The ipliteral router was not recognizing addresses of the form user@
3358 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3360 PH/03 Added disable_ipv6.
3362 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3363 decoded form, because it is most often used to construct To: headers
3364 lines in autoreplies, and the decoded form may well be syntactically
3365 invalid. However, $reply_address has leading white space removed, and all
3366 newlines turned into spaces so that the autoreply transport does not
3369 PH/05 If group was specified without a user on a router, and no group or user
3370 was specified on a transport, the group from the router was ignored.
3372 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3373 for visible compile-time settings that can be used to change these
3374 numbers, for those that want even more. Backwards compatibility with old
3375 spool files has been maintained. However, going back to a previous Exim
3376 release will lost any variables that are in spool files.
3378 PH/07 Two small changes when running in the test harness: increase delay when
3379 passing a TCP/IP connection to a new process, in case the original
3380 process has to generate a bounce, and remove special handling of
3381 127.0.0.2 (sic), which is no longer necessary.
3383 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3384 be the same on different OS.
3386 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3389 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3390 whether --show-vars was specified or not
3392 JJ/02 exipick: Added support for new ACL variable spool format introduced
3395 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3396 syntactically invalid From: or Reply-to: line, and a filter used this to
3397 generate an autoreply, and therefore failed to obtain an address for the
3398 autoreply, Exim could try to deliver to a non-existent relative file
3399 name, causing unrelated and misleading errors. What now happens is that
3400 it logs this as a hard delivery error, but does not attempt to create a
3403 PH/11 The exinext utility has a -C option for testing purposes, but although
3404 the given file was scanned by exinext itself; it wasn't being passed on
3405 when Exim was called.
3407 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3408 an end-of-file indication when reading a command response.
3410 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3411 compiled. In many other places in Exim, IPv6 addresses are always
3412 recognized, so I have changed this. It also means that IPv4 domain
3413 literals of the form [IPV4:n.n.n.n] are now always recognized.
3415 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3416 used if the router is not running as root, for example, when verifying at
3417 ACL time, or when using -bh. The debugging output from this situation was
3418 non-existent - all you got was a failure to exec. I have made two
3421 (a) Failures to set uid/gid, the current directory, or a process leader
3422 in a subprocess such as that created by queryprogram now generate
3423 suitable debugging output when -d is set.
3425 (b) The queryprogram router detects when it is not running as root,
3426 outputs suitable debugging information if -d is set, and then runs
3427 the subprocess without attempting to change uid/gid.
3429 PH/15 Minor change to Makefile for building test_host (undocumented testing
3432 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3433 additional section of a DNS packet that returns MX or SRV records.
3434 Instead, it always explicitly searches for A/AAAA records. This avoids
3435 major problems that occur when a DNS server includes only records of one
3436 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3437 fixed another bug: if SRV records were looked up and the corresponding
3438 address records were *not* found in the additional section, the port
3439 values from the SRV records were lost.
3441 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3442 using the correct key (the original address) when searching the retry
3443 rules in order to find which one to use for generating the retry hint.
3445 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3446 adding the default one. Similarly, if it contains a Reply-To: header, the
3447 errors_reply_to option, if set, is not used.
3449 PH/19 When calculating a retry time, Exim used to measure the "time since
3450 failure" by looking at the "first failed" field in the retry record. Now
3451 it does not use this if it is later than than the arrival time of the
3452 message. Instead it uses the arrival time. This makes for better
3453 behaviour in cases where some deliveries succeed, thus re-setting the
3454 "first failed" field. An example is a quota failure for a huge message
3455 when small messages continue to be delivered. Without this change, the
3456 "time since failure" will always be short, possible causing more frequent
3457 delivery attempts for the huge message than are intended.
3458 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3460 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3461 $1, $2, $3) because the numerical variables can be reset during some
3462 expansion items (e.g. "match"), thereby losing the authentication data.
3464 PH/21 Make -bV show the size of off_t variables so that the test suite can
3465 decide whether to run tests for quotas > 2G.
3467 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3468 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3469 filecount value is greater than 2G or if a quota value is greater than 2G
3470 on a system where the size of off_t is not greater than 4, a panic error
3473 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3474 never match. The debug and -bh output now contains an explicit error
3475 message indicating a malformed IPv4 address or mask.
3477 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3478 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3479 PH/23 above applies.
3481 PH/25 Do not write to syslog when running in the test harness. The only
3482 occasion when this arises is a failure to open the main or panic logs
3483 (for which there is an explicit test).
3485 PH/26 Added the /no_tell option to "control=freeze".
3487 PH/27 If a host name lookup failed very early in a connection, for example, if
3488 the IP address matched host_lookup and the reverse lookup yielded a name
3489 that did not have a forward lookup, an error message of the form "no IP
3490 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3491 could be logged. Now it outputs the IP address instead of "NULL".
3493 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3494 allows the sender and the authenticated sender to be set when
3495 submitting a message from within Exim. Since child_open_exim() is
3496 documented for local_scan(), the new function should be too.
3498 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3499 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3500 results in an empty string is now treated as unset.
3502 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3504 PH/31 Added sender_verify_fail logging option.
3506 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3507 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3508 tidied the source and removed it altogether.
3510 PH/33 When a queue run was abandoned because the load average was too high, a
3511 log line was always written; now it is written only if the queue_run log
3512 selector is set. In addition, the log line for abandonment now contains
3513 information about the queue run such as the pid. This is always present
3514 in "start" and "stop" lines but was omitted from the "abandon" line.
3516 PH/34 Omit spaces between a header name and the colon in the error message that
3517 is given when verify = headers_syntax fails (if there are lots of them,
3518 the message gets confusing).
3520 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3521 names, as there are now some PTR records that contain slashes. This check
3522 is only to protect against broken name servers that fall over on strange
3523 characters, so the fact that it applies to all lookups doesn't matter.
3525 PH/36 Now that the new test suite is complete, we can remove some of the
3526 special code in Exim that was needed for the old test suite. For example,
3527 sorting DNS records because real resolvers return them in an arbitrary
3528 order. The new test suite's fake resolver always returns records in the
3531 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3532 bounces) except when queue_only is set, to avoid logging races between
3533 the different processes.
3535 PH/38 Panic-die if .include specifies a non-absolute path.
3537 PH/39 A tweak to the "H" retry rule from its user.
3539 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3540 a label. They prevented compilation on older perls.
3542 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3543 a warning to be raised on newish perls.
3545 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3546 on queue. Changes to match documented behaviour of showing count of
3547 messages matching specified criteria.
3549 PH/40 Changed the default ident timeout from 30s to 5s.
3551 PH/41 Added support for the use of login_cap features, on those BSD systems
3552 that have them, for controlling the resources used by pipe deliveries.
3554 PH/42 The content-scanning code uses fopen() to create files in which to put
3555 message data. Previously it was not paying any attention to the mode of
3556 the files. Exim runs with umask(0) because the rest of the code creates
3557 files with open(), and sets the required mode explicitly. Thus, these
3558 files were ending up world-writeable. This was not a big issue, because,
3559 being within the spool directory, they were not world-accessible. I have
3560 created a function called modefopen, which takes an additional mode
3561 argument. It sets umask(777), creates the file, chmods it to the required
3562 mode, then resets the umask. All the relevant calls to fopen() in the
3563 content scanning code have been changed to use this function.
3565 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3566 to 24 hours. This avoids potential overflow problems when processing G
3567 and H retry rules. I suspect nobody ever tinkers with this value.
3569 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3571 PH/45 When the plaintext authenticator is running as a client, the server's
3572 challenges are checked to ensure they are valid base64 strings. By
3573 default, the authentication attempt is cancelled if an invalid string is
3574 received. Setting client_ignore_invalid_base64 true ignores these errors.
3575 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3576 they are received. Thus, the responses can be made to depend on the
3577 challenges. If an invalid string is ignored, an empty string is placed in
3580 PH/46 Messages that are created by the autoreply transport now contains a
3581 References: header, in accordance with RFCs 2822 and 3834.
3583 PH/47 Added authenticated_sender_force to the smtp transport.
3585 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3587 PH/49 Installed latest patch from the Sieve maintainer.
3589 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3590 was also set, the appendfile transport was unnecessarily scanning a
3591 directory of message files (e.g. for maildir delivery) to find the count
3592 of files (along with the size), even though it did not need this
3593 information. It now does the scan only if it needs to find either the
3594 size of the count of files.
3596 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3598 PH/52 Two bugs concerned with error handling when the smtp transport is
3601 (i) Exim was not creating retry information for temporary errors given
3602 for individual recipients after the DATA command when the smtp transport
3603 was used in LMTP mode. This meant that they could be retried too
3604 frequently, and not timed out correctly.
3606 (ii) Exim was setting the flag that allows error details to be returned
3607 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3608 recipients that were returned after the DATA command.
3610 PH/53 This is related to PH/52, but is more general: for any failing address,
3611 when detailed error information was permitted to be returned to the
3612 sender, but the error was temporary, then after the final timeout, only
3613 "retry timeout exceeded" was returned. Now it returns the full error as
3614 well as "retry timeout exceeded".
3616 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3617 do this, and (what is worse) MTAs that accept it.
3619 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3620 will now be deprecated.
3622 PH/56 New os.c-cygwin from the Cygwin maintainer.
3624 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3625 formats (previously only available in exim formats via -bpr, -bpru,
3626 and -bpra. Now also available in native and exiqgrep formats)
3628 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3629 with very large, slow to parse queues
3631 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3633 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3635 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3636 responses to authentication challenges, though it was showing the
3637 challenges; (ii) I've removed the CR characters from the debug output for
3640 PH/58 Allow for the insertion of a newline as well as a space when a string
3641 is turned into more than one encoded-word during RFC 2047 encoding. The
3642 Sieve code now uses this.
3644 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3645 data_4xx, lost_connection, tls_required.
3647 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3648 message was being sent as an SMTP response.
3650 PH/61 Add -l and -k options to exicyclog.
3652 PH/62 When verifying, if an address was redirected to one new address, so that
3653 verification continued, and the new address failed or deferred after
3654 having set something in $address_data, the value of $address_data was not
3655 passed back to the ACL. This was different to the case when no
3656 redirection occurred. The value is now passed back in both cases.
3658 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3659 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3660 use login_cap.h, so on its own it isn't the distinguishing feature. The
3661 new name refers directly to the setclassresources() function.
3663 PH/65 Added configuration files for NetBSD3.
3665 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3667 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3668 is preferred over IPv4.
3670 PH/68 The bounce_return_message and bounce_return_body options were not being
3671 honoured for bounces generated during the reception of non-SMTP messages.
3672 In particular, this applied to messages rejected by the ACL. This bug has
3673 been fixed. However, if bounce_return_message is true and bounce_return_
3674 body is false, the headers that are returned for a non-SMTP message
3675 include only those that have been read before the error was detected.
3676 (In the case of an ACL rejection, they have all been read.)
3678 PH/69 The HTML version of the specification is now built in a directory called
3679 spec_html instead of spec.html, because the latter looks like a path with
3680 a MIME-type, and this confuses some software.
3682 PH/70 Catch two compiler warnings in sieve.c.
3684 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3685 function verify_get_ident() calls ip_connect() to connect a socket, but
3686 if the "connect()" function timed out, ip_connect() used to close the
3687 socket. However, verify_get_ident() also closes the socket later, and in
3688 between Exim writes to the log, which may get opened at this point. When
3689 the socket was closed in ip_connect(), the log could get the same file
3690 descriptor number as the socket. This naturally causes chaos. The fix is
3691 not to close the socket in ip_connect(); the socket should be closed by
3692 the function that creates it. There was only one place in the code where
3693 this was missing, in the iplookup router, which I don't think anybody now
3694 uses, but I've fixed it anyway.
3696 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3697 well as to direct DNS lookups. Otherwise the handling of names in host
3698 lists is inconsistent and therefore confusing.
3704 PH/01 Two changes to the default runtime configuration:
3706 (1) Move the checks for relay_from_hosts and authenticated clients from
3707 after to before the (commented out) DNS black list checks.
3709 (2) Add control=submission to the relay_from_hosts and authenticated
3710 clients checks, on the grounds that messages accepted by these
3711 statements are most likely to be submissions.
3713 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3715 (1) Generate an error if the third argument for the ${prvs expansion is
3718 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3721 (3) Reset the variables that are obtained from the first argument of
3722 ${prvscheck and used in the second argument before leaving the code,
3723 because their memory is reclaimed, so using them afterwards may do
3726 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3727 one (it's much easier than Tom thought :-).
3729 (5) Because of (4), we can now allow for the use of $prvscheck_result
3730 inside the third argument.
3732 PH/03 For some reason, the default setting of PATH when running a command from
3733 a pipe transport was just "/usr/bin". I have changed it to
3736 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3737 anything to be listed in the output from -bV.
3739 PH/05 When a filter generated an autoreply, the entire To: header line was
3740 quoted in the delivery log line, like this:
3742 => >A.N.Other <ano@some.domain> <original@ddress> ...
3744 This has been changed so that it extracts the operative address. There
3745 may be more than one such address. If so, they are comma-separated, like
3748 => >ano@some.domain,ona@other.domain <original@ddress> ...
3750 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3751 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3752 not being looked up in the rDNS to get a host name, Exim was showing the
3753 IP address twice in Received: lines, even though the IP addresses were
3754 identical. For example:
3756 Received: from [1.2.3.4] (helo=[1.2.3.4])
3758 However, if the real host name was known, it was omitting the HELO data
3759 if it matched the actual IP address. This has been tidied up so that it
3760 doesn't show the same IP address twice.
3762 PH/07 When both +timestamp and +memory debugging was on, the value given by
3763 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3764 called by the debug printing, thereby overwriting the timestamp buffer.
3765 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3767 PH/08 When the original message was included in an autoreply transport, it
3768 always said "this is a copy of the message, including all the headers",
3769 even if body_only or headers_only was set. It now gives an appropriate
3772 PH/09 Applied a patch from the Sieve maintainer which:
3774 o fixes some comments
3775 o adds the (disabled) notify extension core
3776 o adds some debug output for the result of if/elsif tests
3777 o points to the current vacation draft in the documentation
3778 and documents the missing references header update
3782 o fixes a bug in processing the envelope test (when testing
3783 multiple envelope elements, the last element determined the
3786 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3787 Electronic Mail") by including:
3789 Auto-submitted: auto-generated
3791 in the messages that it generates (bounce messages and others, such as
3792 warnings). In the case of bounce messages for non-SMTP messages, there was
3793 also a typo: it was using "Auto_submitted" (underscore instead of
3794 hyphen). Since every message generated by Exim is necessarily in response
3795 to another message, thes have all been changed to:
3797 Auto-Submitted: auto-replied
3799 in accordance with these statements in the RFC:
3801 The auto-replied keyword:
3803 - SHOULD be used on messages sent in direct response to another
3804 message by an automatic process,
3806 - MUST NOT be used on manually-generated messages,
3808 - MAY be used on Delivery Status Notifications (DSNs) and Message
3809 Disposition Notifications (MDNs),
3811 - MUST NOT be used on messages generated by automatic or periodic
3812 processes, except for messages which are automatic responses to
3815 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3816 to the default Received: header definition.
3818 PH/12 Added log selector acl_warn_skipped (default on).
3820 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3821 variables because (a) they are in the wrong storage pool and (b) even if
3822 they were copied, it wouldn't work properly because of the caching.
3824 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3825 checking when decoding. Apparently there are clients that generate
3826 overlong encoded strings. Why am I not surprised?
3828 PH/15 If the first argument of "${if match_address" was not empty, but did not
3829 contain an "@" character, Exim crashed. Now it writes a panic log message
3830 and treats the condition as false.
3832 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3834 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3835 extension "envelope-auth". The code is finished and in agreement with
3836 other implementations, but there is no documentation so far and in fact,
3837 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3838 not changing the active code.
3840 Print executed "if" and "elsif" statements when debugging is used. This
3841 helps a great deal to understand what a filter does.
3843 Document more things not specified clearly in RFC3028. I had all this
3844 sorted out, when out of a sudden new issues came to my mind. Oops."
3846 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3849 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3850 canonical form (as documented). However, after a host name lookup from
3851 the IP address, check_host() was doing a simple string comparison with
3852 addresses acquired from the DNS when checking that the found name did
3853 have the original IP as one of its addresses. Since any found IPv6
3854 addresses are likely to be in abbreviated form, the comparison could
3855 fail. Luckily, there already exists a function for doing the comparison
3856 by converting both addresses to binary, so now that is used instead of
3857 the text comparison.
3859 PH/20 There was another similar case to PH/19, when a complete host name was
3860 given in a host list; looking up its IP address could give an abbreviated
3861 form, whereas the current host's name might or might not be abbreviated.
3862 The same fix has been applied.
3868 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3869 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3872 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3873 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3875 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3877 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3878 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3879 Also, for "Auto-Submitted", treat anything other than "no" as
3880 non-personal, in accordance with RFC 3834. (Previously it treated
3881 anything starting "auto-" as non-personal.)
3883 TF/01 The control=submission/name=... option had a problem with syntax
3884 errors if the name included a slash character. The /name= option
3885 now slurps the rest of the string, so it can include any characters
3886 but it must come last in the list of options (after /sender_retain
3889 PH/05 Some modifications to the interface to the fake nameserver for the new
3897 TK/01 Added the "success_on_redirect" address verification option. See
3898 NewStuff for rationale and an example.
3900 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3902 PH/02 Patch to exigrep to allow it to work on syslog lines.
3904 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3905 fread() to skip over the body file's header line, because in Cygwin the
3906 header line is locked and is inaccessible.
3908 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3909 co-exist for some time) to make it clear that it is the Exim ID that is
3910 referenced, not the Message-ID: header line.
3912 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3913 string_format() function, because snprintf() does not exist on all
3916 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3917 ${stat: expansion item.
3919 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3920 protocol synchronization error", to keep the pedants happy.
3922 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3923 well as for IRIX systems, when gcc is being used. See the host.c source
3926 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3928 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3931 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3932 transport and to the smtp transport in LMTP mode.
3934 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3936 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3937 run a filter in a subprocess. This could lead to confusion in subsequent
3938 lookups in the parent process. There should also be a search_tidyup() at
3939 the end of the subprocess.
3941 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3942 only if the host matched helo_try_verify_hosts, which caused the
3943 verification to occur when the EHLO/HELO command was issued. The ACL just
3944 tested the remembered result. Now, if a previous verification attempt has
3945 not happened, "verify = helo" does it there and then.
3947 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3949 TK/03 Fix log output including CR from clamd.
3951 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3952 exist provoked a memory error which could cause a segfault.
3954 PH/15 Installed PCRE 6.2
3956 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3958 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3959 of the problem. Specifically, suggested +O2 rather than +O1 for the
3962 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3964 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3965 record, the port was not correctly logged when the outgoing_port log
3966 selector was set (it logged the transort's default port).
3968 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3969 fallback_hosts, and "hosts" in the smtp transport.
3971 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3972 host errors such as "Connection refused".
3974 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3975 authentication with radiusclient 0.4.9:
3977 - Error returned from rc_read_config was caught wrongly
3978 - Username/password not passed on to radius server due to wrong length.
3980 The presumption is that some radiusclient API changes for 4.51/PH/17
3981 were not taken care of correctly. The code is still untested by me (my
3982 Linux distribution still has 0.3.2 of radiusclient), but it was
3983 contributed by a Radius user.
3985 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3986 expanding the "port" option of the smtp transport.
3988 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3989 while reading a MIME header. Thanks to Tom Hughes for a patch.
3991 PH/24 Include config.h inside local_scan.h so that configuration settings are
3994 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3995 that in an ACL for RCPT (for example), you can examine exactly what was
3998 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3999 commands, but it was not correctly comparing the address with the actual
4000 client host address. Thus, it would show the EHLO address in Received:
4001 header lines when this was not necessary.
4003 PH/27 Added the % operator to ${eval:}.
4005 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4006 it should be ignoring failures (because with -C, for example, it has lost
4007 privilege). It wasn't ignoring creation failures other than "already
4010 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4013 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4014 because an input line was too long, either on its own, or by virtue of
4015 too many continuations, the temporary file was not being removed, and the
4016 return code was incorrect.
4018 PH/31 Missing "BOOL" in function definition in filtertest.c.
4020 PH/32 Applied Sieve patches from the maintainer.
4022 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4024 PH/33 Added "verify = not_blind".
4026 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4027 Local/Makefile (with some defaults set). These are used in built scripts
4028 such as exicyclog, but they have never been used in the exim_install
4029 script (though there are many overriding facilities there). I have
4030 arranged that the exim_install script now takes note of these two
4033 PH/35 Installed configuration files for Dragonfly.
4035 PH/36 When a locally submitted message by a trusted user did not contain a
4036 From: header, and the sender address was obtained from -f or from an SMTP
4037 MAIL command, and the trusted user did not use -F to supply a sender
4038 name, $originator_name was incorrectly used when constructing a From:
4039 header. Furthermore, $originator_name was used for submission mode
4040 messages from external hosts without From: headers in a similar way,
4041 which is clearly wrong.
4043 PH/37 Added control=suppress_local_fixups.
4045 PH/38 When log_selector = +received_sender was set, and the addition of the
4046 sender made the log line's construction buffer exactly full, or one byte
4047 less than full, an overflow happened when the terminating "\n" was
4050 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4051 when the result of a list match is failure because a DNS lookup failed.
4053 PH/40 RM_COMMAND is now used in the building process.
4055 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4056 the "build-* directories that it finds.
4058 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4059 address in a domain literal was a prefix of an interface address.
4061 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4062 when verifying a sender address, unless rewrite_headers is false.
4064 PH/44 Wrote a long comment about why errors_to addresses are verified as
4065 recipients, not senders.
4067 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4068 the ratelimit ACL was added.
4070 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4072 PH/46 Added extra information about PostgreSQL errors to the error string.
4074 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4075 avoiding the need to install special zones in a real server. This is
4076 backwards compatible; if it can't find the fake resolver, it drops back.
4077 Thus, both old and new test suites can be run.
4079 TF/02 Added util/ratelimit.pl
4081 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4082 clock is set back in time.
4084 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4085 Candler <B.Candler@pobox.com>.
4087 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4088 OK for addresses that are the result of redirections.
4090 PH/48 A number of further additions for the benefit of the new test suite,
4091 including a fake gethostbyname() that interfaces to the fake DNS resolver
4094 TF/06 The fix for widen_domains has also been applied to qualify_single and
4095 search_parents which are the other dnslookup options that can cause
4098 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4101 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4103 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4104 service" if it encountered an MX record with an empty target hostname.
4105 The message is now "an MX or SRV record indicated no SMTP service".
4107 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4108 if the DNS of the sending site is misconfigured. This is quite a
4109 common situation. This change restores the behaviour of treating a
4110 helo verification defer as a failure.
4112 PH/51 If self=fail was set on a router, the bounce message did not include the
4113 actual error message.
4119 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4121 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4122 command itself ended in error, the underlying message about the transport
4123 filter timeout was being overwritten with the pipe command error. Now the
4124 underlying error message should be appended to the second error message.
4126 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4128 PH/02 Reduce the amount of output that "make" produces by default. Full output
4129 can still be requested.
4131 PH/03 The warning log line about a condition test deferring for a "warn" verb
4132 was being output only once per connection, rather than after each
4133 occurrence (because it was using the same function as for successful
4134 "warn" verbs). This seems wrong, so I have changed it.
4136 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4137 it should not have, which might have caused a crash in the right
4138 circumstances, but probably never did.
4140 PH/04 Installed a modified version of Tony Finch's patch to make submission
4141 mode fix the return path as well as the Sender: header line, and to
4142 add a /name= option so that you can make the user's friendly name appear
4145 TF/03 Added the control = fakedefer ACL modifier.
4147 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4148 Mark Lowes for thorough testing.
4150 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4152 TK/03 Merged latest SRS patch from Miles Wilton.
4154 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4155 with the definition in sysexits.h (which is #included earlier).
4156 Fortunately, Exim does not actually use EX_OK. The code used to try to
4157 preserve the sysexits.h value, by assuming that macro definitions were
4158 scanned for macro replacements. I have been disabused of this notion,
4159 so now the code just undefines EX_OK before #including unistd.h.
4161 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4162 in the smtp transport. When a block could not be written in a single
4163 write() function, the timeout was being re-applied to each part-write.
4164 This seems wrong - if the receiver was accepting one byte at a time it
4165 would take for ever. The timeout is now adjusted when this happens. It
4166 doesn't have to be particularly precise.
4168 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4169 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4171 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4172 without a domain if the check to <postmaster@domain> fails.
4174 SC/01 Eximstats: added -xls and the ability to specify output files
4175 (patch written by Frank Heydlauf).
4177 SC/02 Eximstats: use FileHandles for outputting results.
4179 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4181 SC/04 Eximstats: fixed display of large numbers with -nvr option
4183 SC/05 Eximstats: fixed merging of reports with empty tables.
4185 SC/06 Eximstats: added the -include_original_destination flag
4187 SC/07 Eximstats: removed tabs and trailing whitespace.
4189 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4191 TK/06 MBOX spool code: Add real "From " MBOX separator line
4192 so the .eml file is really in mbox format (even though
4193 most programs do not really care). Patch from Alex Miller.
4195 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4196 The latter is generated from $received_to and is only set if the
4197 message has one envelope recipient. SA can use these headers,
4198 obviously out-of-the-box. Patch from Alex Miller.
4200 PH/08 The ${def test on a variable was returning false if the variable's
4201 value was "0", contrary to what the specification has always said!
4202 The result should be true unless the variable is empty.
4204 PH/09 The syntax error of a character other than { following "${if
4205 def:variable_name" (after optional whitespace) was not being diagnosed.
4206 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4207 accidental colon was present, for example, could give incorrect results.
4209 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4210 result is used (not including appendfile, where other changes are about
4213 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4214 This involved changing a lot of size variables from int to off_t. It
4215 should work with maildirs and everything.
4217 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4218 spamd dying while we are connected to it.
4220 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4223 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4224 function for BDB 4.3.
4226 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4228 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4229 the sources for PCRE have been re-arranged and more files are now
4232 PH/15 The code I had for printing potentially long long variables in PH/11
4233 above was not the best (it lost precision). The length of off_t variables
4234 is now inspected at build time, and an appropriate printing format (%ld
4235 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4236 to be "long long int" or "long int". This is needed for the internal
4237 formatting function string_vformat().
4239 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4240 the configuration file to be ":syslog", then the script "guesses" where
4241 the logs files are, rather than using the compiled in default. In our
4242 case the guess is not the same as the compiled default, so the script
4243 suddenly stopped working when I started to use syslog. The patch checks
4244 to see if log_file_path is "". If so, it attempts to read it from exim
4245 with no configuration file to get the compiled in version, before it
4246 falls back to the previous guessing code."
4248 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4249 implementing BATV in an Exim configuration. See NewStuff for the gory
4252 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4253 Makefile that are specific to HP-UX.
4255 PH/18 If the "use_postmaster" option was set for a recipient callout together
4256 with the "random" option, the postmaster address was used as the MAIL
4257 FROM address for the random test, but not for the subsequent recipient
4258 test. It is now used for both.
4260 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4261 patch removes a few documentation additions to RFC 3028, because the
4262 latest draft now contains them. It adds the new en;ascii-case comparator
4263 and a new error check for 8bit text in MIME parts. Comparator and
4264 require names are now matched exactly. I enabled the subaddress
4265 extension, but it is not well tested yet (read: it works for me)."
4267 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4268 rework some of the code of TK/09 above to avoid the hardwired use of
4269 "%lld" and "long long". Replaced the call to snprintf() with a call to
4272 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4273 records point to non-existent hosts", "retry timeout exceeded", and
4274 "retry time not reached for any host after a long failure period".
4276 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4277 experimental DomainKeys support:
4279 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4280 (2) On an error such as an illegally used "control", the wrong name for
4281 the control was given.
4283 These problems did NOT occur unless DomainKeys support was compiled.
4285 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4287 PH/24 Added ${if match_ip condition.
4289 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4290 databases so that it will be absolutely obvious if a crash occurs in the
4291 DB library. This is a regular occurrence (often caused by mis-matched
4294 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4295 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4296 picked up on a user's system that detects such things. There doesn't seem
4297 to be a gcc warning option for this - only an attribute that has to be
4298 put on the function's prototype. It seems that in Fedora Core 4 they have
4299 set this on a number of new functions. No doubt there will be more in due
4302 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4303 specify a transport. However, if an address that was verified by such a
4304 router was the subject of a callout, Exim crashed because it tried to
4305 read the rcpt_include_affixes from the non-existent transport. Now it
4306 just assumes that the setting of that option is false. This bug was
4307 introduced by 4.51/PH/31.
4309 PH/28 Changed -d+all to exclude +memory, because that information is very
4310 rarely of interest, but it makes the output a lot bigger. People tend to
4311 do -d+all out of habit.
4313 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4314 code in os-type was giving problems when libc.so lives in lib64, like on
4317 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4318 aren't the modern standard, and it seems that some systems' include files
4319 don't always have them. Exim was already checking for some of the newer
4320 ones like T_AAAA, and defining it itself. I've added checks for all the
4321 record types that Exim uses.
4323 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4324 not automatically generating a new one, as it is supposed to. This
4325 prevented TLS from working. If the file did exist, but contained invalid
4326 data, a new version was generated, as expected. It was only the case of a
4327 non-existent file that was broken.
4329 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4330 with a change in libdomainkeys > 0.64.
4332 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4333 from DNS. If the selector record carries the flag, it now has
4334 precedence over the domain-wide flag.
4336 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4338 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4339 the use of a port name that isn't defined in /etc/services) occurred, the
4340 message was deferred as in a normal delivery, and thus remained on the
4341 spool, instead of being failed because of the mua_wrapper setting. This
4342 is now fixed, and I tidied up some of the mua_wrapper messages at the
4345 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4346 the messages in a hash of arrays rather than using individual hashes.
4347 This is a bit cleaner and results in dramatic memory savings, albeit
4348 at a slight CPU cost.
4350 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4351 as requested by Marc Sherman.
4353 SC/10 Eximstats: added histograms for user specified patterns as requested
4356 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4358 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4359 fopen() in the content-scanning modules that did not already have it.
4365 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4366 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4368 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4370 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4372 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4373 to test Sieve filters that use "vacation".
4375 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4376 that changes the way the GnuTLS parameters are stored in the cache file.
4377 The new format can be generated externally. For backward compatibility,
4378 if the data in the cache doesn't make sense, Exim assumes it has read an
4379 old-format file, and it generates new data and writes a new file. This
4380 means that you can't go back to an older release without removing the
4383 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4384 work if there are any delivery delays because "one_time" forces the
4385 parent to be marked "delivered", so its unseen clone is never tried
4386 again. For this reason, Exim now forbids the simultaneous setting of
4389 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4390 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4391 entry if you want to know the details. The fix, however, neglected to
4392 consider the case when local delivery batching is involved. The test for
4393 "previously delivered" was not happening when checking to see if an
4394 address could be batched with a previous (undelivered) one; under
4395 certain circumstances this could lead to multiple deliveries to the same
4398 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4399 in its include files, and this causes problems building Exim.
4401 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4402 header_syntax) but Exim was just ignoring anything given after a slash.
4403 In particular, this caused confusion with an attempt to use "verify =
4404 reverse_host_lookup/defer_ok". An error is now given when options are
4405 supplied for verify items that do not have them. (Maybe reverse_host_
4406 lookup should have a defer_ok option, but that's a different point.)
4408 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4409 defined by RFC 821) to 2048, because there were problems with some AUTH
4410 commands, and RFC 1869 says the size should be increased for extended
4411 SMTP commands that take arguments.
4413 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4416 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4417 "unknown" error; now it says that the functionality isn't in the binary.
4419 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4420 an address' error message when a string expansion fails (syntax or
4421 whatever). Otherwise the password may appear in the log. Following change
4422 PH/42 below, there is no longer a chance of it appearing in a bounce
4425 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4427 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4428 of its addresses was ever tried. (Bugzilla bug #2.)
4430 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4431 the result incorrectly in the debug output. (It correctly added a newline
4432 to what was transported.)
4434 TF/01 Added $received_time.
4436 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4437 commented out examples of how to interface to a virus scanner and to
4438 SpamAssassin. Also added commented examples of av_scanner and
4439 spamd_address settings.
4441 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4442 and controls are allowed in which ACLs. There were a couple of minor
4443 errors. Some of the entries in the conditions table (which is a table of
4444 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4445 negation of where the condition IS allowed.
4447 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4449 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4450 header file does not have a version number, so I've had to invent a new
4451 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4452 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4453 radiusclient), but it was contributed by a Radius user.
4455 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4456 files or directories, for OpenSSL.
4458 PH/19 When an Exim process that is running as root has to create an Exim log
4459 file, it does so in a subprocess that runs as exim:exim so as to get the
4460 ownership right at creation (otherwise, other Exim processes might see
4461 the file with the wrong ownership). There was no test for failure of this
4462 fork() call, which would lead to the process getting stuck as it waited
4463 for a non-existent subprocess. Forks do occasionally fail when resources
4464 run out. I reviewed all the other calls to fork(); they all seem to check
4467 PH/20 When checking for unexpected SMTP input at connect time (before writing
4468 the banner), Exim was not dealing correctly with a non-positive return
4469 from the read() function. If the client had disconnected by this time,
4470 the result was a log entry for a synchronization error with an empty
4471 string after "input=" when read() returned zero. If read() returned -1
4472 (an event I could not check), uninitialized data bytes were printed.
4473 There were reports of junk text (parts of files, etc) appearing after
4476 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4478 PH/22 Added support for macro redefinition, and (re)definition in between
4479 driver and ACL definitions.
4481 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4482 forgetting to use the resulting value; it was using the unexpanded value.
4484 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4485 hadn't been configured. The fix is from Juergen Kreileder, who
4486 understands it better than I do:
4488 "Here's what I see happening with three configured cyrus_sasl
4489 authenticators configured (plain, login, cram-md5):
4491 On startup auth_cyrus_sasl_init() gets called for each of these.
4492 This means three calls to sasl_listmech() without a specified mech_list.
4493 => SASL tests which mechs of all available mechs actually work
4494 => three warnings about OTP not working
4495 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4497 With the patch, sasl_listmech() also gets called three times. But now
4498 SASL's mech_list option is set to the server_mech specified in the the
4499 authenticator. Or in other words, the answer from sasl_listmech()
4500 gets limited to just the mech you're testing for (which is different
4502 => the return list contains just 'plain' or 'login', 'cram-md5' or
4503 nothing depending on the value of ob->server_mech.
4505 I've just tested the patch: Authentication still works fine,
4506 unavailable mechs specified in the exim configuration are still
4507 caught, and the auth.log warnings about OTP are gone."
4509 PH/25 When debugging is enabled, the contents of the command line are added
4510 to the debugging output, even when log_selector=+arguments is not
4513 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4514 answer is "GNU", and only if the return is "GNU/something" is the answer
4517 PH/27 $acl_verify_message is now set immediately after the failure of a
4518 verification in an ACL, and so is available in subsequent modifiers. In
4519 particular, the message can be preserved by coding like this:
4521 warn !verify = sender
4522 set acl_m0 = $acl_verify_message
4524 Previously, $acl_verify_message was set only while expanding "message"
4525 and "log_message" when a very denied access.
4527 PH/28 Modified OS/os.c-Linux with
4529 -#ifndef OS_LOAD_AVERAGE
4530 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4532 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4533 nomenclature these days.)
4535 PH/29 Installed patch from the Sieve maintainer that adds the options
4536 sieve_useraddress and sieve_subaddress to the redirect router.
4538 PH/30 In these circumstances:
4539 . Two addresses routed to the same list of hosts;
4540 . First host does not offer TLS;
4541 . First host accepts first address;
4542 . First host gives temporary error to second address;
4543 . Second host offers TLS and a TLS session is established;
4544 . Second host accepts second address.
4545 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4546 and peerdn, if requested) that were in fact used only for the second
4549 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4550 attention to any local part prefix or suffix that was matched by the
4551 router that accepted the address. It now behaves in the same way as it
4552 does for delivery: the affixes are removed from the local part unless
4553 rcpt_include_affixes is set on the transport.
4555 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4556 timeout during the DATA phase of an incoming message.
4558 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4559 applied a patch sanctioned by the Sieve maintainer.
4561 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4562 the uid or gid is negative. A case of a negative gid caused this to be
4563 noticed. The fix allows for either to be negative.
4565 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4566 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4569 PH/36 The change PH/12 above was broken. Fixed it.
4571 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4572 the grounds that routing the same address twice would always produce the
4573 same answer. This might have been true once, but it is certainly no
4574 longer true now. Routing a child address may depend on the previous
4575 routing that produced that child. Some complicated redirection strategies
4576 went wrong when messages had multiple recipients, and made Exim's
4577 behaviour dependent on the order in which the addresses were given.
4579 I have moved the duplicate checking until after the routing is complete.
4580 Exim scans the addresses that are assigned to local and remote
4581 transports, and removes any duplicates. This means that more work will be
4582 done, as duplicates will always all be routed, but duplicates are
4583 presumably rare, so I don't expect this is of any significance.
4585 For deliveries to pipes, files, and autoreplies, the duplicate checking
4586 still happens during the routing process, since they are not going to be
4589 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4590 It corrects a timeout issue with spamd. This is Ian's comment: "The
4591 background is that sometimes spamd either never reads data from a
4592 connection it has accepted, or it never writes response data. The exiscan
4593 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4594 blindly assumes that writes won't block so it may never time out."
4596 PH/39 Allow G after quota size as well as K and M.
4598 PH/40 The value set for $authenticated_id in an authenticator may not contain
4599 binary zeroes or newlines because the value is written to log lines and
4600 to spool files. There was no check on this. Now the value is run through
4601 the string_printing() function so that such characters are converted to
4602 printable escape sequences.
4604 PH/41 $message_linecount is a new variable that contains the total number of
4605 lines in the message. Compare $body_linecount, which is the count for the
4608 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4609 bounce and delay warning messages, except in certain special cases, which
4612 (a) An SMTP error message from a remote host;
4613 (b) A message specified in a :fail: redirection;
4614 (c) A message specified in a "fail" command in a system filter;
4615 (d) A message specified in a FAIL return from the queryprogram router;
4616 (e) A message specified by the cannot_route_message router option.
4618 In these cases only, Exim does include the error details in bounce and
4619 warning messages. There are also a few cases where bland messages such
4620 as "unrouteable address" or "local delivery error" are given.
4622 PH/43 $value is now also set for the "else" part of a ${run expansion.
4624 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4625 being worked on, but at least Exim now implements the latest version to
4628 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4629 process to complete was treated as a delivery failure, a timeout while
4630 writing the message to the pipe was logged, but erroneously treated as a
4631 successful delivery. Such timeouts include transport filter timeouts. For
4632 consistency with the overall process timeout, these timeouts are now
4633 treated as errors, giving rise to delivery failures by default. However,
4634 there is now a new Boolean option for the pipe transport called
4635 timeout_defer, which, if set TRUE, converts the failures into defers for
4636 both kinds of timeout. A transport filter timeout is now identified in
4639 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4640 systems where "make" and "gmake" are different, calling "gmake" at top
4641 level broke things. I've arranged for the value of $(MAKE) to be passed
4642 from the Makefile to this script so that it can call the same version of
4646 A note about Exim versions 4.44 and 4.50
4647 ----------------------------------------
4649 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4650 changes of various kinds. As a consequence, a big documentation update was
4651 needed. This delayed the release for rather longer than seemed good, especially
4652 in the light of a couple of (minor) security issues. Therefore, the changes
4653 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4654 release. So 4.44 and 4.50 are in effect two different branches that both start
4657 I have left the 4.50 change log unchanged; it contains all the changes since
4658 4.43. The change log for 4.44 is below; many of its items are identical to
4659 those for 4.50. This seems to be the most sensible way to preserve the
4660 historical information.
4666 1. Minor wording change to the doc/README.SIEVE file.
4668 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4669 computation of the current number of files was incorrect.
4671 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4672 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4675 4. Give more explanation in the error message when the command for a transport
4676 filter fails to execute.
4678 5. There are several places where Exim runs a non-Exim command in a
4679 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4680 was being done only for the command run by the queryprogram router. It is
4681 now done for all such subprocesses. The other cases are: ${run, transport
4682 filters, and the commands run by the lmtp and pipe transports.
4684 6. Added CONFIGURE_GROUP build-time option.
4686 7. Some older OS have a limit of 256 on the maximum number of file
4687 descriptors. Exim was using setrlimit() to set 1000 as a large value
4688 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4689 systems. I've change it so that if it can't get 1000, it tries for 256.
4691 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4692 was an oversight, and furthermore, ever since the addition of extra
4693 controls (e.g. 4.43/32), the checks on when to allow different forms of
4694 "control" were broken. There should now be diagnostics for all cases when a
4695 control that does not make sense is encountered.
4697 9. Added the /retain_sender option to "control=submission".
4699 10. $recipients is now available in the predata ACL (oversight).
4701 11. Tidy the search cache before the fork to do a delivery from a message
4702 received from the command line. Otherwise the child will trigger a lookup
4703 failure and thereby defer the delivery if it tries to use (for example) a
4704 cached ldap connection that the parent has called unbind on.
4706 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4707 of $address_data from the recipient verification was clobbered by the
4708 sender verification.
4710 13. The value of address_data from a sender verification is now available in
4711 $sender_address_data in subsequent conditions in the ACL statement.
4713 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4715 15. Added a new option "connect=<time>" to callout options, to set a different
4718 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4719 was its contents. (It was OK if the option was not defined at all.)
4721 17. A "Completed" log line is now written for messages that are removed from
4722 the spool by the -Mrm option.
4724 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4725 information about exactly what failed.
4727 19. Added -dd to debug only the daemon process.
4729 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4730 handles timeouts, both on the server side and network timeouts. Renamed the
4731 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4733 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4734 It is now set to "smtps".
4736 22. $host_address is now set to the target address during the checking of
4737 ignore_target_hosts.
4739 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4740 being passed; this would have caused $sender_host_name to have been used if
4741 matching the list had actually called for a host name (not very likely,
4742 since this list is usually IP addresses). A host name is now passed as
4745 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4746 code that specifies a non-restarting handler (typically sigaction() in
4747 modern systems) in an attempt to fix a rare and obscure crash bug.
4749 25. Narrowed the window for a race in the daemon that could cause it to ignore
4750 SIGCHLD signals. This is not a major problem, because they are used only to
4751 wake it up if nothing else does.
4753 26. A malformed maildirsize file could cause Exim to calculate negative values
4754 for the mailbox size or file count. Odd effects could occur as a result.
4755 The maildirsize information is now recalculated if the size or filecount
4758 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4759 support for a long time. Removed HAVE_SYS_VFS_H.
4761 28. Installed the latest version of exipick from John Jetmore.
4763 29. In an address list, if the pattern was not a regular expression, an empty
4764 subject address (from a bounce message) matched only if the pattern was an
4765 empty string. Non-empty patterns were not even tested. This was the wrong
4766 because it is perfectly reasonable to use an empty address as part of a
4767 database query. An empty address is now tested by patterns that are
4768 lookups. However, all the other forms of pattern expect the subject to
4769 contain a local part and a domain, and therefore, for them, an empty
4770 address still always fails if the pattern is not itself empty.
4772 30. Exim went into a mad DNS loop when attempting to do a callout where the
4773 host was specified on an smtp transport, and looking it up yielded more
4774 than one IP address.
4776 31. Re-factored the code for checking spool and log partition space into a
4777 function that finds that data and another that does the check. The former
4778 is then used to implement four new variables: $spool_space, $log_space,
4779 $spool_inodes, and $log_inodes.
4781 32. The RFC2047 encoding function was originally intended for short strings
4782 such as real names; it was not keeping to the 75-character limit for
4783 encoded words that the RFC imposes. It now respects the limit, and
4784 generates multiple encoded words if necessary. To be on the safe side, I
4785 have increased the buffer size for the ${rfc2047: expansion operator from
4788 33. It is now permitted to omit both strings after an "if" condition; if the
4789 condition is true, the result is "true". As before, when the second string
4790 is omitted, a false condition yields an empty string. This makes it less
4791 cumbersome to write custom ACL and router conditions.
4793 34. Failure to deliver a bounce message always caused it to be frozen, even if
4794 there was an errors_to setting on the router. The errors_to setting is now
4797 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4798 canonical form (fully expanded) before being placed in
4799 $sender_host_address.
4801 36. The table in the code that translates DNS record types into text (T_A to
4802 "A" for instance) was missing entries for NS and CNAME. It is just possible
4803 that this could have caused confusion if both these types were looked up
4804 for the same domain, because the text type is used as part of Exim's
4805 per-process caching. But the chance of anyone hitting this buglet seems
4808 37. The dnsdb lookup has been extended in a number of ways.
4810 (1) There is a new type, "zns", which walks up the domain tree until it
4811 finds some nameserver records. It should be used with care.
4813 (2) There is a new type, "mxh", which is like "mx" except that it returns
4814 just the host names, not the priorities.
4816 (3) It is now possible to give a list of domains (or IP addresses) to be
4817 looked up. The behaviour when one of the lookups defers can be
4818 controlled by a keyword.
4820 (4) It is now possible to specify the separator character for use when
4821 multiple records are returned.
4823 38. The dnslists ACL condition has been extended: it is now possible to supply
4824 a list of IP addresses and/or domains to be looked up in a particular DNS
4827 39. Added log_selector=+queue_time_overall.
4829 40. When running the queue in the test harness, wait just a tad after forking a
4830 delivery process, to get repeatability of debugging output.
4832 41. Include certificate and key file names in error message when GnuTLS fails
4833 to set them up, because the GnuTLS error message doesn't include the name
4834 of the failing file when there is a problem reading it.
4836 42. Allow both -bf and -bF in the same test run.
4838 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4840 44. The "Exiscan patch" is now merged into the mainline Exim source.
4842 45. Sometimes the final signoff response after QUIT could fail to get
4843 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4844 < 0 before doing a fflush(). This bug looks as though it goes back to the
4845 introduction of TLS in release 3.20, but "sometimes" must have been rare
4846 because the tests only now provoked it.
4848 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4849 (this can affect the format of dates).
4851 47. exim_tidydb, when checking for the continued existence of a message for
4852 which it has found a message-specific retry record, was not finding
4853 messages that were in split spool directories. Consequently, it was
4854 deleting retry records that should have stayed in existence.
4856 48. Steve fixed some bugs in eximstats.
4858 49. The SPA authentication driver was not abandoning authentication and moving
4859 on to the next authenticator when an expansion was forced to fail,
4860 contradicting the general specification for all authenticators. Instead it
4861 was generating a temporary error. It now behaves as specified.
4863 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4864 (the order specifies the preference for clients). The order is now AES256,
4865 AES128, 3DES, ARCFOUR128.
4867 51. Small patch to Sieve code - explicitly set From: when generating an
4870 52. Exim crashed if a remote delivery caused a very long error message to be
4871 recorded - for instance if somebody sent an entire SpamAssassin report back
4872 as a large number of 550 error lines. This bug was coincidentally fixed by
4873 increasing the size of one of Exim's internal buffers (big_buffer) that
4874 happened as part of the Exiscan merge. However, to be on the safe side, I
4875 have made the code more robust (and fixed the comments that describe what
4878 53. Now that there can be additional text after "Completed" in log lines (if
4879 the queue_time_overall log selector is set), a one-byte patch to exigrep
4880 was needed to allow it to recognize "Completed" as not the last thing in
4883 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4884 patch that reportedly fixes this has been added. I am not expert enough to
4885 create a test for it. This is what the patch creator wrote:
4887 "I found a little strange behaviour of ldap code when working with
4888 Windows 2003 AD Domain, where users was placed in more than one
4889 Organization Units. When I tried to give exim partial DN, the exit code
4890 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4891 But simultaneously result of request was absolutely normal ldap result,
4892 so I produce this patch..."
4894 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4895 REFERENCE, so I have modified the code to exclude the patch when that macro
4898 55. Some experimental protocols are using DNS PTR records for new purposes. The
4899 keys for these records are domain names, not reversed IP addresses. The
4900 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4901 leaves it alone. Component reversal etc. now happens only for IP addresses.
4904 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4906 57. Double the size of the debug message buffer (to 2048) so that more of very
4907 long debug lines gets shown.
4909 58. The exicyclog utility now does better if the number of log files to keep
4910 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4912 59. Two changes related to the smtp_active_hostname option:
4914 (1) $smtp_active_hostname is now available as a variable.
4915 (2) The default for smtp_banner uses $smtp_active_hostname instead
4916 of $primary_hostname.
4918 60. The host_aton() function is supposed to be passed a string that is known
4919 to be a valid IP address. However, in the case of IPv6 addresses, it was
4920 not checking this. This is a hostage to fortune. Exim now panics and dies
4921 if the condition is not met. A case was found where this could be provoked
4922 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4923 components; fortuitously, this particular loophole had already been fixed
4924 by change 4.50/55 above.
4926 If there are any other similar loopholes, the new check in host_aton()
4927 itself should stop them being exploited. The report I received stated that
4928 data on the command line could provoke the exploit when Exim was running as
4929 exim, but did not say which command line option was involved. All I could
4930 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4931 running as the user.
4934 61. There was a buffer overflow vulnerability in the SPA authentication code
4935 (which came originally from the Samba project). I have added a test to the
4936 spa_base64_to_bits() function which I hope fixes it.
4939 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4940 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4942 63. The daemon start-up calls getloadavg() while still root for those OS that
4943 need the first call to be done as root, but it missed one case: when
4944 deliver_queue_load_max is set with deliver_drop_privilege. This is
4945 necessary for the benefit of the queue runner, because there is no re-exec
4946 when deliver_drop_privilege is set.
4948 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4949 This has been fixed.
4951 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4952 was in use, was not putting the data itself into the right store pool;
4953 consequently, it could be overwritten for a subsequent message in the same
4954 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4957 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4959 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4960 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4961 the function were treating the return as a boolean value, which happened to
4962 work because 0=false and not-0=true, but is not correct code.
4964 68. The host_aton() function was not handling scoped IPv6 addresses (those
4965 with, for example, "%eth0" on the end) correctly.
4967 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4968 negated items (that is, ~something) in unsigned ints. Some compilers
4969 apparently mutter when there is no cast.
4971 70. If an address verification called from an ACL failed, and did not produce a
4972 user-specific message (i.e. there was only a "system" message), nothing was
4973 put in $acl_verify_message. In this situation, it now puts the system
4976 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4977 session; change 4.31/43 added the unwanted input to the log line - except
4978 that it did not do this in the start of session case. It now does.
4980 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4981 This is wrong and can cause the other end to generate a synchronization
4982 error if it is another Exim or anything else that does the synchronization
4983 check. A QUIT command is no longer sent after a timeout.
4985 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4986 during host lookups.
4988 74. The defer_ok option of callout verification was not working if it was used
4989 when verifying addresses in header lines, that is, for this case:
4991 verify = header_sender/callout=defer_ok
4993 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4994 those file descriptors could be used for SMTP connections. If anything
4995 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4996 could be sent to the SMTP client, causing chaos. The daemon now opens
4997 stdin, stdout, and stderr to /dev/null when it puts itself into the
5000 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5001 log by default. The user can override this with suitable Perl magic.
5003 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5004 the log message when discard triggers, was not working for the DATA ACL or
5005 for the non-SMTP ACL.
5007 78. Error message wording change in sieve.c.
5009 79. If smtp_accept_max_per_host was set, the number of connections could be
5010 restricted to fewer than expected, because the daemon was trying to set up
5011 a new connection before checking whether the processes handling previous
5012 connections had finished. The check for completed processes is now done
5013 earlier. On busy systems, this bug wouldn't be noticed because something
5014 else would have woken the daemon, and it would have reaped the completed
5017 80. If a message was submitted locally by a user whose login name contained one
5018 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5019 It caused a spool format error. I have fixed the spool reading code. A
5020 related problem was that the "from" clause in the Received: line became
5021 illegal because of the space(s). It is now covered by ${quote_local_part.
5023 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5026 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5028 83. Patch from Sieve maintainer to fix unterminated string problem in
5029 "vacation" handling.
5031 84. Some minor changes to the Linux configuration files to help with other
5032 OS variants using glibc.
5034 85. One more patch for Sieve to update vacation handling to latest spec.
5037 ----------------------------------------------------
5038 See the note above about the 4.44 and 4.50 releases.
5039 ----------------------------------------------------
5045 1. Change 4.43/35 introduced a bug that caused file counts to be
5046 incorrectly computed when quota_filecount was set in an appendfile
5049 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5050 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5053 3. Give more explanation in the error message when the command for a transport
5054 filter fails to execute.
5056 4. There are several places where Exim runs a non-Exim command in a
5057 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5058 was being done only for the command run by the queryprogram router. It is
5059 now done for all such subprocesses. The other cases are: ${run, transport
5060 filters, and the commands run by the lmtp and pipe transports.
5062 5. Some older OS have a limit of 256 on the maximum number of file
5063 descriptors. Exim was using setrlimit() to set 1000 as a large value
5064 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5065 systems. I've change it so that if it can't get 1000, it tries for 256.
5067 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5068 was an oversight, and furthermore, ever since the addition of extra
5069 controls (e.g. 4.43/32), the checks on when to allow different forms of
5070 "control" were broken. There should now be diagnostics for all cases when a
5071 control that does not make sense is encountered.
5073 7. $recipients is now available in the predata ACL (oversight).
5075 8. Tidy the search cache before the fork to do a delivery from a message
5076 received from the command line. Otherwise the child will trigger a lookup
5077 failure and thereby defer the delivery if it tries to use (for example) a
5078 cached ldap connection that the parent has called unbind on.
5080 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5081 of $address_data from the recipient verification was clobbered by the
5082 sender verification.
5084 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5085 was its contents. (It was OK if the option was not defined at all.)
5087 11. A "Completed" log line is now written for messages that are removed from
5088 the spool by the -Mrm option.
5090 12. $host_address is now set to the target address during the checking of
5091 ignore_target_hosts.
5093 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5094 being passed; this would have caused $sender_host_name to have been used if
5095 matching the list had actually called for a host name (not very likely,
5096 since this list is usually IP addresses). A host name is now passed as
5099 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5100 code that specifies a non-restarting handler (typically sigaction() in
5101 modern systems) in an attempt to fix a rare and obscure crash bug.
5103 15. Narrowed the window for a race in the daemon that could cause it to ignore
5104 SIGCHLD signals. This is not a major problem, because they are used only to
5105 wake it up if nothing else does.
5107 16. A malformed maildirsize file could cause Exim to calculate negative values
5108 for the mailbox size or file count. Odd effects could occur as a result.
5109 The maildirsize information is now recalculated if the size or filecount
5112 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5113 support for a long time. Removed HAVE_SYS_VFS_H.
5115 18. Updated exipick to current release from John Jetmore.
5117 19. Allow an empty sender to be matched against a lookup in an address list.
5118 Previously the only cases considered were a regular expression, or an
5121 20. Exim went into a mad DNS lookup loop when doing a callout where the
5122 host was specified on the transport, if the DNS lookup yielded more than
5125 21. The RFC2047 encoding function was originally intended for short strings
5126 such as real names; it was not keeping to the 75-character limit for
5127 encoded words that the RFC imposes. It now respects the limit, and
5128 generates multiple encoded words if necessary. To be on the safe side, I
5129 have increased the buffer size for the ${rfc2047: expansion operator from
5132 22. Failure to deliver a bounce message always caused it to be frozen, even if
5133 there was an errors_to setting on the router. The errors_to setting is now
5136 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5137 canonical form (fully expanded) before being placed in
5138 $sender_host_address.
5140 24. Updated eximstats to version 1.33
5142 25. Include certificate and key file names in error message when GnuTLS fails
5143 to set them up, because the GnuTLS error message doesn't include the name
5144 of the failing file when there is a problem reading it.
5146 26. Expand error message when OpenSSL has problems setting up cert/key files.
5149 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5150 (this can affect the format of dates).
5152 28. exim_tidydb, when checking for the continued existence of a message for
5153 which it has found a message-specific retry record, was not finding
5154 messages that were in split spool directories. Consequently, it was
5155 deleting retry records that should have stayed in existence.
5157 29. eximstats updated to version 1.35
5158 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5159 1.35 - bugfix such that pie charts by volume are generated correctly
5161 30. The SPA authentication driver was not abandoning authentication and moving
5162 on to the next authenticator when an expansion was forced to fail,
5163 contradicting the general specification for all authenticators. Instead it
5164 was generating a temporary error. It now behaves as specified.
5166 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5167 (the order specifies the preference for clients). The order is now AES256,
5168 AES128, 3DES, ARCFOUR128.
5170 31. Small patch to Sieve code - explicitly set From: when generating an
5173 32. Exim crashed if a remote delivery caused a very long error message to be
5174 recorded - for instance if somebody sent an entire SpamAssassin report back
5175 as a large number of 550 error lines. This bug was coincidentally fixed by
5176 increasing the size of one of Exim's internal buffers (big_buffer) that
5177 happened as part of the Exiscan merge. However, to be on the safe side, I
5178 have made the code more robust (and fixed the comments that describe what
5181 33. Some experimental protocols are using DNS PTR records for new purposes. The
5182 keys for these records are domain names, not reversed IP addresses. The
5183 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5184 leaves it alone. Component reversal etc. now happens only for IP addresses.
5187 34. The host_aton() function is supposed to be passed a string that is known
5188 to be a valid IP address. However, in the case of IPv6 addresses, it was
5189 not checking this. This is a hostage to fortune. Exim now panics and dies
5190 if the condition is not met. A case was found where this could be provoked
5191 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5192 components; fortuitously, this particular loophole had already been fixed
5193 by change 4.50/55 or 4.44/33 above.
5195 If there are any other similar loopholes, the new check in host_aton()
5196 itself should stop them being exploited. The report I received stated that
5197 data on the command line could provoke the exploit when Exim was running as
5198 exim, but did not say which command line option was involved. All I could
5199 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5200 running as the user.
5203 35. There was a buffer overflow vulnerability in the SPA authentication code
5204 (which came originally from the Samba project). I have added a test to the
5205 spa_base64_to_bits() function which I hope fixes it.
5208 36. The daemon start-up calls getloadavg() while still root for those OS that
5209 need the first call to be done as root, but it missed one case: when
5210 deliver_queue_load_max is set with deliver_drop_privilege. This is
5211 necessary for the benefit of the queue runner, because there is no re-exec
5212 when deliver_drop_privilege is set.
5214 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5215 was in use, was not putting the data itself into the right store pool;
5216 consequently, it could be overwritten for a subsequent message in the same
5217 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5220 38. Sometimes the final signoff response after QUIT could fail to get
5221 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5222 < 0 before doing a fflush(). This bug looks as though it goes back to the
5223 introduction of TLS in release 3.20, but "sometimes" must have been rare
5224 because the tests only now provoked it.
5230 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5231 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5232 Now it returns an integer. A number of calls were still expecting a T/F
5233 return. Fortuitously, in all cases, the tests worked in OK situations,
5234 which is the norm. However, things would have gone wrong on any write
5235 failures on the smtp file descriptor. This function is used when sending
5236 messages over SMTP and also when doing verify callouts.
5238 2. When Exim is called to do synchronous delivery of a locally submitted
5239 message (the -odf or -odi options), it no longer closes stderr before doing
5242 3. Implemented the mua_wrapper option.
5244 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5246 5. Implemented the functions header_remove(), header_testname(),
5247 header_add_at_position(), and receive_remove_recipient(), and exported them
5250 6. If an ACL "warn" statement specified the addition of headers, Exim already
5251 inserted X-ACL-Warn: at the start if there was no header name. However, it
5252 was not making this test for the second and subsequent header lines if
5253 there were newlines in the string. This meant that an invalid header could
5254 be inserted if Exim was badly configured.
5256 7. Allow an ACL "warn" statement to add header lines at the start or after all
5257 the Received: headers, as well as at the end.
5259 8. Added the rcpt_4xx retry error code.
5261 9. Added postmaster_mailfrom=xxx to callout verification option.
5263 10. Added mailfrom=xxxx to the callout verification option, for verify=
5266 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5267 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5269 12. Inserted some casts to stop certain compilers warning when using pointer
5270 differences as field lengths or precisions in printf-type calls (mostly
5271 affecting debugging statements).
5273 13. Added optional readline() support for -be (dynamically loaded).
5275 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5276 same clock tick as a message's arrival, so that its received time was the
5277 same as the "first fail" time on the retry record, and that message
5278 remained on the queue past the ultimate address timeout, every queue runner
5279 would try a delivery (because it was past the ultimate address timeout) but
5280 after another failure, the ultimate address timeout, which should have then
5281 bounced the address, did not kick in. This was a "< instead of <=" error;
5282 in most cases the first failure would have been in the next clock tick
5283 after the received time, and all would be well.
5285 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5286 being recognized when the domain list was tested by the match_domain
5287 condition in an expansion string.
5289 16. Added the ${str2b64: operator.
5291 17. Exim was always calling setrlimit() to set a large limit for the number of
5292 processes, without checking whether the existing limit was already
5293 adequate. (It did check for the limit on file descriptors.) Furthermore,
5294 errors from getrlimit() and setrlimit() were being ignored. Now they are
5295 logged to the main and panic logs, but Exim does carry on, to try to do its
5296 job under whatever limits there are.
5298 18. Imported PCRE 5.0.
5300 19. Trivial typo in log message " temporarily refused connection" (the leading
5303 20. If the log selector return_path_on_delivery was set and an address was
5304 redirected to /dev/null, the delivery process crashed because it assumed
5305 that a return path would always be set for a "successful" delivery. In this
5306 case, the whole delivery is bypassed as an optimization, and therefore no
5309 21. Internal re-arrangement: the function for sending a challenge and reading
5310 a response while authentication was assuming a zero-terminated challenge
5311 string. It's now changed to take a pointer and a length, to allow for
5312 binary data in such strings.
5314 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5316 23. Exim was not respecting finduser_retries when seeking the login of the
5317 uid under which it was called; it was always trying 10 times. (The default
5318 setting of finduser_retries is zero.) Also, it was sleeping after the final
5319 failure, which is pointless.
5321 24. Implemented tls_on_connect_ports.
5323 25. Implemented acl_smtp_predata.
5325 26. If the domain in control=submission is set empty, Exim assumes that the
5326 authenticated id is a complete email address when it generates From: or
5327 Sender: header lines.
5329 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5330 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5331 chown and chgrp in /bin and hostname in /usr/bin.
5333 28. Exim was keeping the "process log" file open after each use, just as it
5334 does for the main log. This opens the possibility of it remaining open for
5335 long periods when the USR1 signal hits a daemon. Occasional processlog
5336 errors were reported, that could have been caused by this. Anyway, it seems
5337 much more sensible not to leave this file open at all, so that is what now
5340 29. The long-running daemon process does not normally write to the log once it
5341 has entered its main loop, and it closes the log before doing so. This is
5342 so that log files can straightforwardly be renamed and moved. However,
5343 there are a couple of unusual error situations where the daemon does write
5344 log entries, and I had neglected to close the log afterwards.
5346 30. The text of an SMTP error response that was received during a remote
5347 delivery was being truncated at 512 bytes. This is too short for some of
5348 the long messages that one sometimes sees. I've increased the limit to
5351 31. It is now possible to make retry rules that apply only when a message has a
5352 specific sender, in particular, an empty sender.
5354 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5355 it possible to be selective about when SMTP synchronization is enforced.
5357 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5359 32. Implemented hosts_connection_nolog.
5361 33. Added an ACL for QUIT.
5363 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5366 35. Added mailbox_size and mailbox_filecount to appendfile.
5368 36. Added control = no_multiline_responses to ACLs.
5370 37. There was a bug in the logic of the code that waits for the clock to tick
5371 in the case where the clock went backwards by a substantial amount such
5372 that the microsecond fraction of "now" was more than the microsecond
5373 fraction of "then" (but the whole seconds number was less).
5375 38. Added support for the libradius Radius client library this is found on
5376 FreeBSD (previously only the radiusclient library was supported).
5382 1. When certain lookups returned multiple values in the form name=value, the
5383 quoting of the values was not always being done properly. Specifically:
5384 (a) If the value started with a double quote, but contained no whitespace,
5386 (b) If the value contained whitespace other than a space character (i.e.
5387 tabs or newlines or carriage returns) it was not quoted.
5388 This fix has been applied to the mysql and pgsql lookups by writing a
5389 separate quoting function and calling it from the lookup code. The fix
5390 should probably also be applied to nisplus, ibase and oracle lookups, but
5391 since I cannot test any of those, I have not disturbed their existing code.
5393 2. A hit in the callout cache for a specific address caused a log line with no
5394 reason for rejecting RCPT. Now it says "Previous (cached) callout
5395 verification failure".
5397 3. There was an off-by-one bug in the queryprogram router. An over-long
5398 return line was truncated at 256 instead of 255 characters, thereby
5399 overflowing its buffer with the terminating zero. As well as fixing this, I
5400 have increased the buffer size to 1024 (and made a note to document this).
5402 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5403 when Exim is waiting for an SMTP response from a remote server, Exim
5404 restarts its select() call on the socket, thereby resetting its timeout.
5405 This is not a problem when such interrupts are rare. Somebody set up a cron
5406 job to run exiwhat every 2 minutes, which is less than the normal select()
5407 timeout (5 or 10 minutes). This meant that the select() timeout never
5408 kicked in because it was always reset. I have fixed this by comparing the
5409 time when an interrupt arrives with the time at the start of the first call
5410 to select(). If more time than the timeout has elapsed, the interrupt is
5411 treated as a timeout.
5413 5. Some internal re-factoring in preparation for the addition of Sieve
5414 extensions (by MH). In particular, the "personal" test is moved to a
5415 separate function, and given an option for scanning Cc: and Bcc: (which is
5416 not set for Exim filters).
5418 6. When Exim created an email address using the login of the caller as the
5419 local part (e.g. when creating a From: or Sender: header line), it was not
5420 quoting the local part when it contained special characters such as @.
5422 7. Installed new OpenBSD configuration files.
5424 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5425 try to make them clearer.
5427 9. Callout options, other than the timeout value, were being ignored when
5428 verifying sender addresses in header lines. For example, when using
5430 verify = header_sender/callout=no_cache
5432 the cache was (incorrectly) being used.
5434 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5435 only the Cygwin environment.
5437 11. When return_path_on_delivery was set as a log selector, if different remote
5438 addresses in the same message used different return paths and parallel
5439 remote delivery occurred, the wrong values would sometimes be logged.
5440 (Whenever a remote delivery process finished, the return path value from
5441 the most recently started remote delivery process was logged.)
5443 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5444 lines when AUTH and/or TLS are in use. This is the "received protocol"
5445 field. Exim used to use "asmtp" for authenticated SMTP, without any
5446 indication (in the protocol name) for TLS use. Now it follows the RFC and
5447 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5448 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5449 names appear in log lines as well as in Received: header lines.
5451 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5452 extensions, and comparison tests, and to fix some bugs.
5454 14. Changes to the "personal" filter test:
5456 (1) The test was buggy in that it was just doing the equivalent of
5457 "contains" tests on header lines. For example, if a user's address was
5458 anne@some.where, the "personal" test would incorrectly be true for
5460 To: susanne@some.where
5462 This test is now done by extracting each address from the header in turn,
5463 and checking the entire address. Other tests that are part of "personal"
5464 are now done using regular expressions (for example, to check local parts
5465 of addresses in From: header lines).
5467 (2) The list of non-personal local parts in From: addresses has been
5468 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5469 taken from the Sieve specification recommendations.
5471 (3) If the message contains any header line starting with "List-" it is
5472 treated as non-personal.
5474 (4) The test for "circular" in the Subject: header line has been removed
5475 because it now seems ill-conceived.
5477 15. Minor typos in src/EDITME comments corrected.
5479 16. Installed latest exipick from John Jetmore.
5481 17. If headers_add on a router specified a text string that was too long for
5482 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5483 of string_sprintf() is now avoided.
5485 18. $message_body_size was not set (it was always zero) when running the DATA
5486 ACL and the local_scan() function.
5488 19. For the "mail" command in an Exim filter, no default was being set for
5489 the once_repeat time, causing a random time value to be used if "once" was
5490 specified. (If the value happened to be <= 0, no repeat happened.) The
5491 default is now 0s, meaning "never repeat". The "vacation" command was OK
5492 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5493 (I found it when inspecting the code).
5495 20. There is now an overall timeout for performing a callout verification. It
5496 defaults to 4 times the callout timeout, which applies to individual SMTP
5497 commands during the callout. The overall timeout applies when there is more
5498 than one host that can be tried. The timeout is checked before trying the
5499 next host. This prevents very long delays if there are a large number of
5500 hosts and all are timing out (e.g. when the network connections are timing
5501 out). The value of the overall timeout can be changed by specifying an
5502 additional sub-option for "callout", called "maxwait". For example:
5504 verify = sender/callout=5s,maxwait=20s
5506 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5507 to the end before writing, but this should make it even safer).
5509 22. Exim was forgetting that it had advertised PIPELINING for the second and
5510 subsequent messages on an SMTP connection. It was also not resetting its
5511 memory on STARTTLS and an internal HELO.
5513 23. When Exim logs an SMTP synchronization error within a session, it now
5514 records whether PIPELINING has been advertised or not.
5516 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5517 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5518 rather than long int.
5520 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5522 26. Added the never_mail option to autoreply.
5528 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5529 crash if the getsockname() call failed; this can happen if a connection is
5530 closed very soon after it is established. The problem was simply in the
5531 order in which certain operations were done, causing Exim to try to write
5532 to the SMTP stream before it had set up the file descriptor. The bug has
5533 been fixed by making things happen in the correct order.
5539 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5540 before the connection was closed, thus losing the rejection response.
5542 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5543 some early Solaris releases, but causes trouble in current releases where
5544 socklen_t is defined.
5546 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5549 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5552 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5553 that releasing the top of it at the end releases what was used for sub-
5554 expansions (unless the block got too big). However, discard this block if
5555 the first thing is a variable or header, so that we can use its block when
5556 it is dynamic (useful for very large $message_headers, for example).
5558 6. Lookups now cache *every* query, not just the most recent. A new, separate
5559 store pool is used for this. It can be recovered when all lookup caches are
5560 flushed. Lookups now release memory at the end of their result strings.
5561 This has involved some general refactoring of the lookup sources.
5563 7. Some code has been added to the store_xxx() functions to reduce the amount
5564 of flapping under certain conditions.
5566 8. log_incoming_interface used to affect only the <= reception log lines. Now
5567 it causes the local interface and port to be added to several more SMTP log
5568 lines, for example "SMTP connection from", and rejection lines.
5570 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5572 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5574 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5575 could be overwritten at the end of the current message (or the start of a
5576 new message if it was set in a HELO ACL). The value is now preserved for
5577 the duration of the SMTP connection.
5579 12. If a transport had a headers_rewrite setting, and a matching header line
5580 contained an unqualified address, that address was qualified, even if it
5581 did not match any rewriting rules. The underlying bug was that the values
5582 of the flags that permit the existence of unqualified sender and recipient
5583 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5584 non-local messages, and by -bnq for local messages) were not being
5585 preserved with the message after it was received.
5587 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5588 "next input=" as part of the text comprising the host identity instead of
5589 the correct text. The code was using the same buffer for two different
5590 strings. However, depending on which order the printing function evaluated
5591 its arguments, the bug did not always show up. Under Linux, for example, my
5592 test suite worked just fine.
5594 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5595 doesn't work with some older versions of Perl. It has been changed to "my",
5596 which in any case is probably the better facility to use.
5598 15. A really picky compiler found some instances of statements for creating
5599 error messages that either had too many or two few arguments for the format
5602 16. The size of the buffer for calls to the DNS resolver has been increased
5603 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5604 for addresses that have a lot of PTR records. This alleviates a problem; it
5605 does not fully solve it.
5607 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5608 buffer now truncates the list and logs the incident, which is the same
5609 action as happens when Exim is looking up a host name and its aliases.
5610 Previously in this situation something unpredictable would happen;
5611 sometimes it was "internal error: store_reset failed".
5613 18. If a server dropped the connection unexpectedly when an Exim client was
5614 using GnuTLS and trying to read a response, the client delivery process
5615 crashed while trying to generate an error log message.
5617 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5618 string, for example:
5620 warn message = H1: something\nH2: something
5622 the text was added as a single header line from Exim's point of view
5623 though it ended up OK in the delivered message. However, searching for the
5624 second and subsequent header lines using $h_h2: did not work. This has been
5625 fixed. Similarly, if a system filter added multiple headers in this way,
5626 the routers could not see them.
5628 20. Expanded the error message when iplsearch is called with an invalid key to
5629 suggest using net-iplsearch in a host list.
5631 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5632 ACLs are no longer actually imposed (and a message to that effect is
5635 22. If a "gecos" field in a passwd entry contained escaped characters, in
5636 particular, if it contained a \" sequence, Exim got it wrong when building
5637 a From: or a Sender: header from that name. A second bug also caused
5638 incorrect handling when an unquoted " was present following a character
5639 that needed quoting.
5641 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5642 was not being matched caselessly.
5644 24. Arranged for all hyphens in the exim.8 source to be escaped with
5647 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5648 properly. Recipient callout cache records were still being keyed to include
5649 the sender, even when use_sender was set false. This led to far more
5650 callouts that were necessary. The sender is no longer included in the key
5651 when use_sender is false.
5653 26. Added "control = submission" modifier to ACLs.
5655 27. Added the ${base62d: operator to decode base 62 numbers.
5657 28. dnsdb lookups can now access SRV records.
5659 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5660 the configuration file.
5662 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5663 (-v) mode. This makes the output for a verbose queue run more intelligible.
5665 31. Added a use_postmaster feature to recipient callouts.
5667 32. Added the $body_zerocount variable, containing the number of binary zero
5668 bytes in the message body.
5670 33. The time of last modification of the "new" subdirectory is now used as the
5671 "mailbox time last read" when there is a quota error for a maildir
5674 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5676 35. Added +ignore_unknown as a special item in host lists.
5678 36. Code for decoding IPv6 addresses in host lists is now included, even if
5679 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5680 address was recognized as an IP address, but was then not correctly decoded
5681 into binary, causing unexpected and incorrect effects when compared with
5688 1. Very minor rewording of debugging text in manualroute to say "list of
5689 hosts" instead of "hostlist".
5691 2. If verify=header_syntax was set, and a header line with an unqualified
5692 address (no domain) and a large number of spaces between the end of the
5693 name and the colon was received, the reception process suffered a buffer
5694 overflow, and (when I tested it) crashed. This was caused by some obsolete
5695 code that should have been removed. The fix is to remove it!
5697 3. When running in the test harness, delay a bit after writing a bounce
5698 message to get a bit more predictability in the log output.
5700 4. Added a call to search_tidyup() just before forking a reception process. In
5701 theory, someone could use a lookup in the expansion of smtp_accept_max_
5702 per_host which, without the tidyup, could leave open a database connection.
5704 5. Added the variables $recipient_data and $sender_data which get set from a
5705 lookup success in an ACL "recipients" or "senders" condition, or a router
5706 "senders" option, similar to $domain_data and $local_part_data.
5708 6. Moved the writing of debug_print from before to after the "senders" test
5711 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5712 problems for message scanning, either using a data ACL, or using
5713 local_scan() because the Received: header was not generated till after they
5714 were called (in order to set the time as the time of reception completion).
5715 I have revised the way this works. The header is now generated after the
5716 body is received, but before the ACL or local_scan() are called. After they
5717 are run, the timestamp in the header is updated.
5723 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5724 before starting a queue runner without re-exec. This happened only when
5725 deliver_drop_privilege was set or when the Exim user was set to root. The
5726 effect of the bug was that timeouts during subsequent deliveries caused
5727 crashes instead of being properly handled. The handler is now left at its
5728 default (and expected) setting.
5730 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5731 message, again when deliver_drop_privilege is set or Exim is run as root.
5732 The bug described in (1) was not present in this case, but the tidying up
5733 of the other signals was missing. I have made the two cases consistent.
5735 3. The ignore_target_hosts setting on a manualroute router was being ignored
5736 for hosts that were looked up using the /MX notation.
5738 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5741 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5742 operated on the sender address. After changing the $sender_address to <>
5743 for the sender address verify, Exim was re-instated it as the original
5744 (before rewriting) address, but remembering that it had rewritten it, so it
5745 wasn't rewriting it again. This bug also had the effect of breaking the
5746 sender address verification caching when the sender address was rewritten.
5748 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5749 This has been changed so that if the ip literal address matches
5750 ignore_target_hosts, the router declines.
5752 7. Added expansion conditions match_domain, match_address, and match_local_
5753 part (NOT match_host).
5755 8. The placeholder for the Received: header didn't have a length field set.
5757 9. Added code to Exim itself and to exim_lock to test for a specific race
5758 condition that could lead to file corruption when using MBX delivery. The
5759 issue is with the lockfile that is created in /tmp. If this file is removed
5760 after a process has opened it but before that process has acquired a lock,
5761 there is the potential for a second process to recreate the file and also
5762 acquire a lock. This could lead to two Exim processes writing to the file
5763 at the same time. The added code performs the same test as UW imapd; it
5764 checks after acquiring the lock that its file descriptor still refers to
5765 the same named file.
5767 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5768 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5771 11. Added the smtp_active_hostname option. If used, this will typically be made
5772 to depend on the incoming interface address. Because $interface_address is
5773 not set up until the daemon has forked a reception process, error responses
5774 that can happen earlier (such as "too many connections") no longer contain
5777 12. If an expansion in a condition on a "warn" statement fails because a lookup
5778 defers, the "warn" statement is abandoned, and the next ACL statement is
5779 processed. Previously this caused the whole ACL to be aborted.
5781 13. Added the iplsearch lookup type.
5783 14. Added ident_timeout as a log selector.
5785 15. Added tls_certificate_verified as a log selector.
5787 16. Added a global option tls_require_ciphers (compare the smtp transport
5788 option of the same name). This controls incoming TLS connections.
5790 17. I finally figured out how to make tls_require_ciphers do a similar thing
5791 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5792 before starting the TLS session.
5794 18. Tabs are now shown as \t in -bP output.
5796 19. If the log selector return_path_on_delivery was set, Exim crashed when
5797 bouncing a message because it had too many Received: header lines.
5799 20. If two routers both had headers_remove settings, and the first one included
5800 a superfluous trailing colon, the final name in the first list and the
5801 first name in the second list were incorrectly joined into one item (with a
5802 colon in the middle).
5808 1. Added -C and -D options to the exinext utility, mainly to make it easier
5809 to include in the automated testing, but these could be helpful when
5810 multiple configurations are in use.
5812 2. The exinext utility was not formatting the output nicely when there was
5813 an alternate port involved in the retry record key, nor when there was a
5814 message id as well (for retries that were specific to a specific message
5815 and a specific host). It was also confused by IPv6 addresses, because of
5816 the additional colons they contain. I have fixed the IPv4 problem, and
5817 patched it up to do a reasonable job for IPv6.
5819 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5820 delivery, the log line now contains "pipelined" if PIPELINING was used.
5822 4. An SMTP transport process used to panic and die if the bind() call to set
5823 an explicit outgoing interface failed. This has been changed; it is now
5824 treated in the same way as a connect() failure.
5826 5. A reference to $sender_host_name in the part of a conditional expansion
5827 that was being skipped was still causing a DNS lookup. This no longer
5830 6. The def: expansion condition was not recognizing references to header lines
5831 that used bh_ and bheader_.
5833 7. Added the _cache feature to named lists.
5835 8. The code for checking quota_filecount in the appendfile transport was
5836 allowing one more file than it should have been.
5838 9. For compatibility with Sendmail, the command line option
5846 and sets the incoming protocol and host name (for trusted callers). The
5847 host name and its colon can be omitted when only the protocol is to be set.
5848 Note the Exim already has two private options, -pd and -ps, that refer to
5849 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5850 "s", but I don't think that's a major issue.
5852 10. A number of refactoring changes to the code, none of which should affect
5855 (a) The number of logging options was getting close to filling up the
5856 32-bit word that was used as a bit map. I have split them into two classes:
5857 those that are passed in the argument to log_write(), and those that are
5858 only ever tested independently outside of that function. These are now in
5859 separate 32-bit words, so there is plenty of room for expansion again.
5860 There is no change in the user interface or the logging behaviour.
5862 (b) When building, for example, log lines, the code previously used a
5863 macro that called string_cat() twice, in order to add two strings. This is
5864 not really sufficiently general. Furthermore, there was one instance where
5865 it was actually wrong because one of the argument was used twice, and in
5866 one call a function was used. (As it happened, calling the function twice
5867 did not affect the overall behaviour.) The macro has been replaced by a
5868 function that can join an arbitrary number of extra strings onto a growing
5871 (c) The code for expansion conditions now uses a table and a binary chop
5872 instead of a serial search (which was left over from when there were very
5873 few conditions). Also, it now recognizes conditions like "pam" even when
5874 the relevant support is not compiled in: a suitably worded error message is
5875 given if an attempt is made to use such a condition.
5877 11. Added ${time_interval:xxxxx}.
5879 12. A bug was causing one of the ddress fields not to be passed back correctly
5880 from remote delivery subprocesses. The field in question was not being
5881 subsequently used, so this caused to problems in practice.
5883 13. Added new log selectors queue_time and deliver_time.
5885 14. Might have fixed a bug in maildirsizefile handling that threw up
5886 "unexpected character" debug warnings, and recalculated the data
5887 unnecessarily. In any case, I expanded the warning message to give more
5890 15. Added the message "Restricted characters in address" to the statements in
5891 the default ACL that block characters like @ and % in local parts.
5893 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5894 Three changes have been made:
5896 (a) There was a serious bug; a negative response to MAIL caused the whole
5897 recipient domain to be cached as invalid, thereby blocking all messages
5898 to all local parts at the same domain, from all senders. This bug has
5899 been fixed. The domain is no longer cached after a negative response to
5900 MAIL if the sender used is not empty.
5902 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5905 (c) A new callout option, "use_sender" has been added for people who want
5906 the modified behaviour.
5912 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5915 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5916 indeed breaks things for older releases.
5918 3. Added additional logging to the case where there is a problem reading data
5919 from a filter that is running in a subprocess using a pipe, in order to
5920 try to track down a specific problem.
5922 4. Testing facility fudge: when running in the test harness and attempting
5923 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5924 getting "No route to host". Convert this to a timeout.
5926 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5929 6. Some OS don't have socklen_t but use size_t instead. This affects the
5930 fifth argument of getsockopt() amongst other things. This is now
5931 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5932 can be set for individual OS. I have set it for SunOS5, OSF1, and
5933 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5934 some earlier ones do not.
5936 7. Change 4.30/15 was not doing the test caselessly.
5938 8. The standard form for an IPv6 address literal was being rejected by address
5939 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5940 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5941 this, as well as the form without the "IPv6" on the front (but only when
5942 address literals are enabled, of course).
5944 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5946 10. Exim crashed if a message with an empty sender address specified by -f
5947 encountered a router with an errors_to setting. This could be provoked only
5948 by a command such as
5952 where an empty string was supplied; "<>" did not hit this bug.
5954 11. Installed PCRE release 4.5.
5956 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5957 remained set. It is now erased.
5959 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5960 times from message ids (which are base 36 rather than the normal 62).
5962 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5963 were being counted as actual protocol errors, and logged if the log
5964 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5965 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5966 and DATA following a set of rejected RCPTs do not count as protocol errors.
5967 In other words, Exim assumes they were pipelined, though this may not
5968 actually be the case. Of course, in all cases the client gets an
5969 appropriate error code.
5971 15. If a lookup fails in an ACL condition, a message about the failure may
5972 be available; it is used if testing the ACL cannot continue, because most
5973 such messages specify what the cause of the deferral is. However, some
5974 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5975 that caused an old message to be retained and used if a later statement
5976 caused a defer, replacing the real cause of the deferral.
5978 16. If an IP address had so many PTR records that the DNS lookup buffer
5979 was not large enough to hold them, Exim could crash while trying to process
5980 the truncated data. It now detects and logs this case.
5982 17. Further to 4.21/58, another change has been made: if (and only if) the
5983 first line of a message (the first header line) ends with CRLF, a bare LF
5984 in a subsequent header line has a space inserted after it, so as not to
5985 terminate the header.
5987 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5988 unnecessarily, used atoi() instead of strtol(), and didn't check the
5989 termination when getting file sizes from file names by regex.
5991 19. Completely re-implemented the support for maildirsize files, in the light
5992 of a number of problems with the previous contributed implementation
5993 (4.30/29). In particular:
5995 . If the quota is zero, the maildirsize file is maintained, but no quota is
5998 . If the maildir directory does not exist, it is created before any attempt
5999 to write a maildirsize file.
6001 . The quota value in the file is just a cache; if the quota is changed in
6002 the transport, the new value overrides.
6004 . A regular expression is available for excluding directories from the
6007 20. The autoreply transport checks the characters in options that define the
6008 message's headers; it allows continued headers, but it was checking with
6009 isspace() after an embedded newline instead of explicitly looking for a
6012 21. If all the "regular" hosts to which an address was routed had passed their
6013 expiry times, and had not reached their retry times, the address was
6014 bounced, even if fallback hosts were defined. Now Exim should go on to try
6017 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6018 equivalent code in the SMTP transport. Some hosts send humungous responses
6019 to HELO/EHLO, more than 1024 it seems.
6021 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6022 gives compiler warnings in some environments. I've now done it "properly",
6025 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6026 (because of problems with the built-in one) was declared to return uschar *
6027 instead of char *, causing compiler failure.
6029 25. Fixed a file descriptor leak when processing alias/forward files.
6031 26. Fixed a minor format string issue in dbfn.c.
6033 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6035 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6036 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6039 29. When a sender address is verified, it is cached, to save repeating the test
6040 when there is more than one recipient in a message. However, when the
6041 verification involves a callout, it is possible for different callout
6042 options to be set for different recipients. It is too complicated to keep
6043 track of this in the cache, so now Exim always runs a verification when a
6044 callout is required, relying on the callout cache for the optimization.
6045 The overhead is duplication of the address routing, but this should not be
6048 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6049 to be verified with callout=postmaster, and the main callout worked but the
6050 postmaster check failed, the verification correctly failed. However, if a
6051 subsequent RCPT command asked for sender verification *without* the
6052 postmaster check, incorrect caching caused this verification also to fail,
6055 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6056 it was not caching the DNS options (qualify_single, search_parents) that
6057 were used when the lookup failed. A subsequent lookup with different
6058 options therefore always gave the same answer, though there were cases
6059 where it should not have. (Example: a "domains = !$mx_any" option on a
6060 dnslookup router: the "domains" option is always processed without any
6061 widening, but the router might have qualify_single set.) Now Exim uses the
6062 cached value only when the same options are set.
6064 32. Added John Jetmore's "exipick" utility to the distribution.
6066 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6067 than a timeout (e.g. a certificate is required, and is not provided), an
6068 Exim server now closes the connection immediately. Previously it waited for
6069 the client to close - but if the client is SSL, it seems that they each
6070 wait for each other, leading to a delay before one of them times out.
6072 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6073 maintained 0.8.x compatibility because I don't think many are using it, and
6074 it is clearly obsolete.
6076 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6079 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6080 client certificate was expired. A simple patch fixes this, though I don't
6081 understand the full logic of why the verify callback is called multiple
6084 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6085 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6086 which causes problems with some clients (such as the Certicom SSL Plus
6087 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6088 disables the coutermeasure allowing Eudora to connect."
6090 38. Exim was not checking that a write() to a log file succeeded. This could
6091 lead to Bad Things if a log got too big, in particular if it hit a file
6092 size limit. Exim now panics and dies if it cannot write to a log file, just
6093 as it does if it cannot open a log file.
6095 39. Modified OS/Makefile-Linux so that it now contains
6097 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6099 The two -D definitions ensure that Exim is compiled with large file
6100 support, which makes it possible to handle log files that are bigger than
6103 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6104 instance) a domain was checked against a named list that involved a lookup,
6105 causing $domain_data to be set, then another domain was checked against the
6106 same list, then the first domain was re-checked, the value of $domain_data
6107 after the final check could be wrong. In particular, if the second check
6108 failed, it could be set empty. This bug probably also applied to
6111 41. The strip_trailing_dot option was not being applied to the address given
6112 with the -f command-line option.
6114 42. The code for reading a message's header from the spool was incrementing
6115 $received_count, but never initializing it. This meant that the value was
6116 incorrect (doubled) while delivering a message in the same process in which
6117 it was received. In the most common configuration of Exim, this never
6118 happens - a fresh exec is done - but it can happen when
6119 deliver_drop_privilege is set.
6121 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6122 it now includes up to 150 characters of the unexpected data in the log
6125 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6126 and building data strings. The size of both of these buffers was 10 000
6127 bytes - far larger than anybody would *ever* want, thought I. Needless to
6128 say, somebody hit the limit. I have increased the maximum line length to
6129 20 000 and the maximum data length of concatenated lines to 100 000. I have
6130 also fixed two bugs, because there was no checking on these buffers. Tsk,
6131 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6132 buffer is too small.
6134 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6135 lsearch lookups. Now it does.
6137 46. When parsing a route_list item in a manualroute router, a fixed-length
6138 buffer was used for the list of hosts. I made this 1024 bytes long,
6139 thinking that nobody would ever have a list of hosts that long. Wrong.
6140 Somebody had a whole pile of complicated expansion conditions, and the
6141 string was silently truncated, leading to an expansion error. It turns out
6142 that it is easier to change to an unlimited length (owing to other changes
6143 that have happened since this code was originally written) than to build
6144 structure for giving a limitation error. The length of the item that
6145 expands into the list of hosts is now unlimited.
6147 47. The lsearch lookup could not handle data where the length of text line was
6148 more than 4095 characters. Such lines were truncated, leading to shortened
6149 data being returned. It should now handle lines of any length.
6151 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6152 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6155 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6156 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6157 no longer happens in comment lines. A list of replacements is now placed
6158 at the head of all of the source files, except those whose only change is
6159 to replace PERL_COMMAND in the very first #! line.
6161 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6162 messages on the queue, with a bottom-up merge sort, using code contributed
6163 by Michael Haardt. This should make operations like -bp somewhat faster on
6164 large queues. It won't affect queue runners, except when queue_run_in_order
6167 51. Installed eximstats 1.31 in the distribution.
6169 52. Added support for SRV lookups to the dnslookup router.
6171 53. If an ACL referred to $message_body or $message_body_end, the value was not
6172 reset for any messages that followed in the same SMTP session.
6174 54. The store-handling optimization for building very long strings was not
6175 differentiating between the different store pools. I don't think this
6176 actually made any difference in practice, but I've tidied it.
6178 55. While running the routers to verify a sender address, $sender_address
6179 was still set to the sender address. This is wrong, because when routing to
6180 send a bounce to the sender, it would be empty. Therefore, I have changed
6181 it so that, while verifying a sender address, $sender_address is set to <>.
6182 (There is no change to what happens when verifying a recipient address.)
6184 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6185 target A or AAAA records (if not already returned) without resetting the
6186 qualify_single or search_parents options of the DNS resolver. These are
6187 inappropriate in this case because the targets of MX and SRV records must
6188 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6189 target that, when qualified, matched something in the local domain. These
6190 two options are now turned off when doing these lookups.
6192 57. It seems that at least some releases of Reiserfs (which does not have the
6193 concept of a fixed number of inodes) returns zero and not -1 for the
6194 number of available inodes. This interacted badly with check_spool_inodes,
6195 which assumed that -1 was the "no such thing" setting. What I have done is
6196 to check that the total number of inodes is greater than zero before doing
6197 the test of how many are available.
6199 58. When a "warn" ACL statement has a log_message modifier, the message is
6200 remembered, and not repeated. This is to avoid a lot of repetition when a
6201 message has many recipients that cause the same warning to be written.
6202 However, Exim was preserving the list of already written lines for an
6203 entire SMTP session, which doesn't seem right. The memory is now reset if a
6204 new message is started.
6206 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6207 debugging output unless log_rewrite was also set.
6209 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6210 of (int)(handle) when we know that handle contains (void *)(-1).
6212 61. The Exim daemon panic-logs an error return when it closes the incoming
6213 connection. However "connection reset by peer" seems to be common, and
6214 isn't really an error worthy of noting specially, so that particular error
6217 62. When Exim is trying to find all the local interfaces, it used to panic and
6218 die if the ioctl to get the interface flags failed. However, it seems that
6219 on at least one OS (Solaris 9) it is possible to have an interface that is
6220 included in the list of interfaces, but for which you get a failure error
6221 for this call. This happens when the interface is not "plumbed" into a
6222 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6223 failure of the "get flags" call assumes that the interface is down.
6225 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6226 makes life easier for people who are doing arithmetic on fields extracted
6227 from dates, where you often get leading zeros that should not be
6228 interpreted as octal.
6230 64. Added qualify_domain to the redirect router, to override the global
6233 65. If a pathologically long header line contained very many addresses (the
6234 report of this problem mentioned 10 000) and each of them was rewritten,
6235 Exim could use up a very large amount of memory. (It kept on making new
6236 copies of the header line as it rewrote, and never released the old ones.)
6237 At the expense of a bit more processing, the header rewriting function has
6238 been changed so that it no longer eats memory in this way.
6240 66. The generation of the Received: header has been moved from the time that a
6241 message starts to be received, to the time that it finishes. The timestamp
6242 in the Received: header should now be very close to that of the <= log
6243 line. There are two side-effects of this change:
6245 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6246 logged header lines no longer include the local Received: line, because
6247 it has not yet been created. The same applies to a copy of the message
6248 that is returned to a non-SMTP sender when a message is rejected.
6250 (b) When a filter file is tested using -bf, no additional Received: header
6251 is added to the test message. After some thought, I decided that this
6254 This change does not affect the value of $received_for. It is still set
6255 after address rewriting, but before local_scan() is called.
6257 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6259 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6260 gave an unhelpful panic error message, and a defer error. I have managed to
6261 change this behaviour so that it now rejects any supplied certificate,
6262 which seems right, as the list of acceptable certificates is empty.
6264 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6265 gave an unhelpful defer error. I have not managed to make this reject any
6266 supplied certificates, but the error message it gives is "no certificate
6267 supplied", which is not helpful.
6269 70. exigrep's output now also includes lines that are not associated with any
6270 message, but which match the given pattern. Implemented by a patch from
6271 Martin Sluka, which also tidied up the Perl a bit.
6273 71. Recipient callout verification, like sender verification, was using <> in
6274 the MAIL FROM command. This isn't really the right thing, since the actual
6275 sender may affect whether the remote host accepts the recipient or not. I
6276 have changed it to use the actual sender in the callout; this means that
6277 the cache record is now keyed on a recipient/sender pair, not just the
6278 recipient address. There doesn't seem to be a real danger of callout loops,
6279 since a callout by the remote host to check the sender would use <>.
6280 [SEE ABOVE: changed after hitting problems.]
6282 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6283 temporary errors. However, in the case of such a code being given after
6284 the end of a data transmission (i.e. after ".") Exim was failing to write
6285 a retry record for the message. (Yes, there was some broken host that was
6286 actually sending 8xx at this point.)
6288 73. An unknown lookup type in a host list could cause Exim to panic-die when
6289 the list was checked. (An example that provoked this was putting <; in the
6290 middle of a list instead of at the start.) If this happened during a DATA
6291 ACL check, a -D file could be left lying around. This kind of configuration
6292 error no longer causes Exim to die; instead it causes a defer error. The
6293 incident is still logged to the main and panic logs.
6295 74. Buglet left over from Exim 3 conversion. The message "too many messages
6296 in one connection" was written to the rejectlog but not the mainlog, except
6297 when address rewriting (yes!) was being logged.
6299 75. Added write_rejectlog option.
6301 76. When a system filter was run not as root (that is, when system_filter_user
6302 was set), the values of the $n variables were not being returned to the
6303 main process; thus, they were not subsequently available in the $sn
6306 77. Added +return_path_on_delivery log selector.
6308 78. A connection timeout was being treated differently from recipients deferred
6309 when testing hosts_max_try with a message that was older than the host's
6310 retry timeout. (The host should not be counted, thus allowing all hosts to
6311 be tried at least once before bouncing.) This may have been the cause of an
6312 occasionally reported bug whereby a message would remain on the queue
6313 longer than the retry timeout, but would be bounced if a delivery was
6314 forced. I say "may" because I never totally pinned down the problem;
6315 setting up timeout/retry tests is difficult. See also the next item.
6317 79. The ultimate address timeout was not being applied to errors that involved
6318 a combination of host plus message (for example, a timeout on a MAIL
6319 command). When an address resolved to a number of possible hosts, and they
6320 were not all tried for each delivery (e.g. because of hosts_max_try), a
6321 message could remain on the queue longer than the retry timeout.
6323 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6326 81. Fixed an obscure SMTP outgoing bug which required at least the following
6327 conditions: (a) there was another message waiting for the same server;
6328 (b) the server returned 5xx to all RCPT commands in the first message so
6329 that the message was not completed; (c) the server dropped the connection
6330 or gave a negative response to the RSET that Exim sends to abort the
6331 transaction. The observed case was a dropped connection after DATA that had
6332 been sent in pipelining mode. That is, the server had advertised PIPELINING
6333 but was not implementing it correctly. The effect of the bug was incorrect
6334 behaviour, such as trying another host, and this could lead to a crash.
6340 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6341 and daemon.c were passed as pointers to ints; they should have been
6342 pointers to socklen_t variables (which are typically unsigned ints).
6344 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6347 3. Fixed a really odd bug that affected only the testing scheme; patching a
6348 certain fixed string in the binary changed the value of another string that
6349 happened to be identical to the end of the original first string.
6351 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6352 name", it returns that address as the IP address. On some operating
6353 systems (e.g. Solaris), it also passes back the IP address string as the
6354 "host name". However, on others (e.g. Linux), it passes back an empty
6355 string. Exim wasn't checking for this, and was changing the host name to an
6356 empty string, assuming it had been canonicalized.
6358 5. Although rare, it is permitted to have more than one PTR record for a given
6359 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6360 all the names associated with an address, because they do in Solaris.
6361 However, it seems that they do not in Linux for data that comes from the
6362 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6363 I found this out when I moved to a new Linux workstation and tried to run
6364 the Exim test suite.
6366 To get round this problem I have changed the code so that it now does its
6367 own call to the DNS to look up PTR records when searching for a host name.
6368 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6369 addresses that are only in /etc/hosts are still found.
6371 This behaviour is, however, controlled by an option called host_lookup_
6372 order, which defaults to "bydns:byaddr". If people want to use the other
6373 order, or indeed, just use one or the other means of lookup, they can
6374 specify it in this variable.
6376 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6377 some operating systems, this comes back from gethostbyaddr() as an empty
6378 string, and this is what Exim used to test for. However, it seems that in
6379 other systems, "." is yielded. Exim now tests for this case too.
6381 7. The values of check_spool_space and check_log_space are now held internally
6382 as a number of kilobytes instead of an absolute number of bytes. If a
6383 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6384 kilobyte. This means that much larger values can be stored.
6386 8. Exim monitor: an attempt to get the action menu when not actually pointing
6387 at a message produces an empty menu entitled "No message selected". This
6388 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6389 no entries in it ("Shell widget menu has zero width and/or height"). So I
6390 have added a single, blank menu entry in this case.
6392 9. Added ${quote_local_part.
6394 10. MIME decoding is now applied to the contents of Subject: header lines when
6397 11. Now that a reference to $sender_host_address automatically causes a reverse
6398 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6399 host lookup before query-style lookups in lists that might use this
6400 variable. This has therefore been abolished, and the "net-" prefix is no
6401 longer necessary for query-style lookups.
6403 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6404 have been a typo for LFLAGS, so it has been changed.
6406 13. The install script calls Exim with "-C /dev/null" in order to find the
6407 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6408 to be output. However, since Exim outputs its version number before the
6409 error, it didn't break the script. It just looked ugly. I fixed this by
6410 always allowing "-C /dev/null" if the caller is root.
6412 14. Ignore overlarge ACL variable number when reading spool file - insurance
6413 against a later release with more variables having written the file.
6415 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6416 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6417 this, as well as the form without the "IPv6" on the front.
6419 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6420 OS/Makefile-Darwin file.
6422 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6423 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6425 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6426 variable. However, some OS use ints for inodes. I've added cast to long int
6427 to get rid of the compiler warning.
6429 19. I had forgotten to lock out "/../" in configuration file names when
6430 ALT_CONFIG_PREFIX was set.
6432 20. Routers used for verification do not need to specify transports. However,
6433 if such a router generated a host list, and callout was configured, Exim
6434 crashed, because it could not find a port number from the (non-existent)
6435 transport. It now assumes port 25 in this circumstance.
6437 21. Added the -t option to exigrep.
6439 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6440 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6441 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6442 them is set, LOOKUP_LSEARCH is forced.
6444 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6446 24. Added sender and host information to the "rejected by local_scan()" log
6447 line; previously there was no indication of these.
6449 25. Added .include_if_exists.
6451 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6452 Linux. It turns out that not all file systems support this. Apparently some
6453 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6454 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6455 not supported on the file descriptor, is now ignored when Exim is trying to
6456 sync a directory. This applies only to Linux.
6458 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6460 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6461 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6464 29. Added support for maildirsize files from supplied patch (modified a bit).
6466 30. The use of :fail: followed by an empty string could lead Exim to respond to
6467 sender verification failures with (e.g.):
6469 550 Verification failed for <xxx>
6470 550 Sender verify failed
6472 where the first response line was missing the '-' that indicates it is not
6473 the final line of the response.
6475 31. The loop for finding the name of the user that called Exim had a hardwired
6476 limit of 10; it now uses the value of finduser_retries, which is used for
6477 all other user lookups.
6479 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6482 33. Exim was neglecting to zero errno before one call of strtol() when
6483 expanding a string and expecting an integer value. On some systems this
6484 resulted in spurious "integer overflow" errors. Also, it was casting the
6485 result into an int without checking.
6487 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6488 did not work. The code looks as if it has *never* worked, though it appears
6489 to have been documented since at least release 1.62. I have made it work.
6491 35. The "timeout_DNS" error in retry rules, also documented since at least
6492 1.62, also never worked. As it isn't clear exactly what this means, and
6493 clearly it isn't a major issue, I have abolished the feature by treating it
6494 as "timeout", and writing a warning to the main and panic logs.
6496 36. The display of retry rules for -brt wasn't always showing the error code
6499 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6500 timeout_connect_A, timeout_connect_MX.
6502 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6503 to the empty sender.
6505 39. The daemon was not analysing the content of -oX till after it had closed
6506 stderr and disconnected from the controlling terminal. This meant that any
6507 syntax errors were only noted on the panic log, and the return code from
6508 the command was 0. By re-arranging the code a little, I've made the
6509 decoding happen first, so such errors now appear on stderr, and the return
6510 code is 1. However, the actual setting up of the sockets still happens in
6511 the disconnected process, so errors there are still only recorded on the
6514 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6515 connections (as happens on some IP stacks) was logged at start up time as
6516 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6517 it from "IPv6 and IPv4", which means that two separate sockets are being
6520 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6521 says whether AF_INET or AF_INET6 was passed as an argument.
6523 42. Exiwhat output was messed up when time zones were included in log
6526 43. Exiwhat now gives more information about the daemon's listening ports,
6527 and whether -tls-on-connect was used.
6529 44. The "port" option of the smtp transport is now expanded.
6531 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6532 silently ignored. Now an error message is written to the main and panic
6535 46. There's a new ACL modifier called "logwrite" which writes to a log file
6536 as soon as it is encountered.
6538 47. Added $local_user_uid and $local_user_gid at routing time.
6540 48. Exim crashed when trying to verify a sender address that was being
6543 49. Exim was recognizing only a space character after ".include". It now also
6544 recognizes a tab character.
6546 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6547 extracting the relevant information from the specification. The man page no
6548 longer contains scrambled data for the -d option, and I've added a section
6549 at the front about calling Exim under different names.
6551 51. Added "extra_headers" argument to the "mail" command in filter files.
6553 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6556 53. Installed eximstats 1.29.
6558 54. Added transport_filter_timeout as a generic transport option.
6560 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6561 Cc: header lines. This was required by RFC 822, but it not required by RFC
6564 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6565 incoming messages that did not have them. Now it does so only if the
6566 message originates locally, that is, if there is no associated remote host
6567 address. When Resent- header lines are present, this applies to the Resent-
6568 lines rather than the non-Resent- lines.
6570 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6571 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6573 58. Messages for configuration errors now include the name of the main
6574 configuration file - useful now that there may be more than one file in a
6575 list (.included file names were always shown).
6577 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6578 for those rare installations that do not start the daemon as root or run it
6579 setuid root. I've cut out the call to initgroups() if the daemon is not
6582 60. The Exim user and group can now be bound into the binary as text strings
6583 that are looked up at the start of Exim's processing.
6585 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6587 62. Added $mailstore_basename variable.
6589 63. Installed patch to sieve.c from Michael Haardt.
6591 64. When Exim failed to open the panic log after failing to open the main log,
6592 the original message it was trying to log was written to stderr and debug
6593 output, but if they were not available (the usual case in production), it
6594 was lost. Now it is written to syslog before the two lines that record the
6595 failures to open the logs.
6597 65. Users' Exim filters run in subprocesses under the user's uid. It is
6598 possible for a "deliver" command or an alias in a "personal" command to
6599 provoke an address rewrite. If logging of address rewriting is configured,
6600 this fails because the process is not running as root or exim. There may be
6601 a better way of dealing with this, but for the moment (because 4.30 needs
6602 to be released), I have disabled address rewrite logging when running a
6603 filter in a non-root, non-exim process.
6609 1. The buildconfig auxiliary program wasn't quoting the value set for
6610 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6611 not defined. This bug was masked in 4.22 by the effect that was fixed in
6614 2. Some messages that were rejected after a message id was allocated were
6615 shown as "incomplete" by exigrep. It no longer does this for messages that
6616 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6618 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6619 have allow_domain_literals set, the ID did not get logged in the <= line.
6620 Domain literals are now always recognized in Message-ID: header lines.
6622 4. The first argument for a ${extract expansion item is the key name or field
6623 number. Leading and trailing spaces in this item were not being ignored,
6624 causing some misleading effects.
6626 5. When deliver_drop_privilege was set, single queue runner processes started
6627 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6628 same command in the process it spins off) were not dropping privilege.
6630 6. When the daemon running as "exim" started a queue runner, it always
6631 re-executed Exim in the spun-off process. This is a waste of effort when
6632 deliver_drop_privilege is set. The new process now just calls the
6633 queue-runner function directly.
6639 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6642 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6643 ignored. Though the use of -oP was forcing the writing of a pid file, it
6644 was always written to the default place.
6646 3. If the message "no IP address found for host xxxx" is generated during
6647 incoming verification, it is now followed by identification of the incoming
6648 connection (so you can more easily find what provoked it).
6650 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6652 5. Added some features to "harden" Exim a bit more against certain attacks:
6654 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6655 be put in Local/Makefile. This is like the never_users runtime option,
6656 but it cannot be overridden. The default setting is "root".
6658 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6659 prefix string with which any file named in a -C command line option
6662 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6663 is retained for -C and -D only if the caller of Exim is root. Without
6664 it, the exim user may also use -C and -D and retain privilege.
6666 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6667 command line option is disabled.
6669 6. Macro names set by the -D option must start with an upper case letter, just
6670 like macro names defined in the configuration file.
6672 7. Added "dereference=" facility to LDAP.
6674 8. Two instances of the typo "uknown" in the source files are fixed.
6676 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6677 the Configure-Makefile script screwed up while processing it.
6679 10. Incorporated PCRE 4.4.
6681 11. The SMTP synchronization check was not operating right at the start of an
6682 SMTP session. For example, it could not catch a HELO sent before the client
6683 waited for the greeting. There is now a check for outstanding input at the
6684 point when the greeting is written. Because of the duplex, asynchronous
6685 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6686 way, but not yet received, when the check is performed.
6688 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6689 on TCP/IP sockets, because this apparently causes some broken clients to
6692 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6693 unchanged) from the Cygwin maintainer.
6695 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6696 was supported instead of testing for mbx. Effectively a typo.
6698 15. The spa authenticator server code was not checking that the input it
6699 received was valid base64.
6701 16. The debug output line for the "set" modifier in ACLs was not showing the
6702 name of the variable that was being set.
6704 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6706 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6707 lookup on its own. Something else was needed to trigger the lookup. For
6708 example, a match in host_lookup or the need for a host name in a host list.
6709 Now, if $sender_host_name is referenced and the host name has not yet been
6710 looked up, a lookup is performed. If the lookup fails, the variable remains
6711 empty, and $host_lookup_failed is set to "1".
6713 19. Added "eqi" as a case-independent comparison operator.
6715 20. The saslauthd authentication condition could segfault if neither service
6716 nor realm was specified.
6718 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6719 error message that was logged was misleading, and incoming SMTP
6720 connections were dropped. The message is now more accurate, and temporary
6721 errors are given to SMTP connections.
6723 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6724 (except RSET) with a 421 error, until QUIT is received. However, it was
6725 failing to send a response to QUIT.
6727 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6728 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6729 ACL is now run after the helo_try_verify_hosts code.
6731 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6732 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6733 case-independent, so other case variants are also recognized). Apparently
6734 some systems use these upper case variants.
6736 25. If more than two messages were waiting for the same host, and a transport
6737 filter was specified for the transport, Exim sent two messages over the
6738 same TCP/IP connection, and then failed with "socket operation on non-
6739 socket" when it tried to send the third.
6741 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6743 27. The extern definition of crypt16() in expand.c was not being excluded when
6744 the OS had its own crypt16() function.
6746 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6747 as a preferred synonym for return_size_limit, both as an option and as an
6750 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6752 30. Changed the default configuration ACL to relax the local part checking rule
6753 for addresses that are not in any local domains. For these addresses,
6754 slashes and pipe symbols are allowed within local parts, but the sequence
6755 /../ is explicitly forbidden.
6757 31. SPA server authentication was not clearing the challenge buffer before
6760 32. log_message in a "warn" ACL statement was writing to the reject log as
6761 well as to the main log, which contradicts the documentation and doesn't
6762 seem right (because no rejection is happening). So I have stopped it.
6764 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6765 However, I am unable to do any testing of this.
6767 34. Fixed an infelicity in the appendfile transport. When checking directories
6768 for a mailbox, to see if any needed to be created, it was accidentally
6769 using path names with one or more superfluous leading slashes; tracing
6770 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6772 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6773 contents are added to the log line that is written for every discarded
6774 recipient. (Previously a log_message setting was ignored.)
6776 36. The ${quote: operator now quotes the string if it is empty.
6778 37. The install script runs exim in order to find its version number. If for
6779 some reason other than non-existence or emptiness, which it checks, it
6780 could not run './exim', it was installing it with an empty version number,
6781 i.e. as "exim-". This error state is now caught, and the installation is
6784 38. An argument was missing from the function that creates an error message
6785 when Exim fails to connect to the socket for saslauthd authentication.
6786 This could cause Exim to crash, or give a corrupted message.
6788 39. Added isip, isip4, and isip6 to ${if conditions.
6790 40. The ACL variables $acl_xx are now saved with the message, and can be
6791 accessed later in routers, transports, and filters.
6793 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6794 strings in the file are not string-expanded.
6796 42. If a MAIL command specified a SIZE value that was too large to fit into an
6797 int variable, the check against message_size_limit failed. Such values are
6798 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6799 day this will have to be increased, but I don't think I want to be around
6800 when emails are that large.
6807 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6808 iconv() is not standard in FreeBSD.
6810 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6811 IPv6 enabled. The observed symptom was a segmentation fault on return from
6812 the function os_common_find_running_interfaces() in src/os.c.
6814 3. In the check_special_case() function in daemon.c I had used "errno" as an
6815 argument name, which causes warnings on some systems. This was basically a
6816 typo, since it was named "eno" in the comments!
6818 4. The code that waits for the clock to tick (at a resolution of some fraction
6819 of a second) so as to ensure message-id uniqueness was always waiting for
6820 at least one whole tick, when it could have waited for less. [This is
6821 almost certainly not relevant at current processor speeds, where it is
6822 unlikely to ever wait at all. But we try to future-proof.]
6824 5. The function that sleeps for a time interval that includes fractions of a
6825 second contained a race. It did not block SIGALRM between setting the
6826 timer, and suspending (a couple of lines later). If the interval was short
6827 and the sigsuspend() was delayed until after it had expired, the suspension
6828 never ended. On busy systems this could lead to processes getting stuck for
6831 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6832 process, before it forks any delivery processes. The open lookup caching
6833 mechanism meant that the open file or database connection was passed into
6834 the delivery process. The problem was that delivery processes always tidy
6835 up cached lookup data. This could cause a problem for the next delivery
6836 process started by the queue runner, because the external queue runner
6837 process does not know about the closure. So the next delivery process
6838 still has data in the lookup cache. In the case of a file lookup, there was
6839 no problem because closing a file descriptor in a subprocess doesn't affect
6840 the parent. However, if the lookup was caching a connection to a database,
6841 the connection was closed, and the second delivery process was likely to
6842 see errors such as "PGSQL: query failed: server closed the connection
6843 unexpectedly". The problem has been fixed by closing all cached lookups
6844 in a queue runner before running a delivery process.
6846 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6847 seem to have the "const" qualifier which it has on other OS. I've
6850 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6851 *of the same type* (client or server) with the same public name that an
6852 error should be diagnosed.
6854 9. When Exim looked up a host name for an IP address, but failed to find the
6855 original IP address when looking up the host name (a safety check), it
6856 output the message "<ip address> does not match any IP for NULL", which was
6857 confusing, to say the least. The bug was that the host name should have
6858 appeared instead of "NULL".
6860 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6861 user that is built into the binary, and the -C or -D options is used, root
6862 privilege is dropped before the configuration file is read. In addition,
6863 logging is switched to stderr instead of the normal log files. If the
6864 configuration then re-defines the Exim user, the unprivileged environment
6865 is probably not what is expected, so Exim logs a panic warning message (but
6868 However, if deliver_drop_privilege is set, the unprivileged state may well
6869 be exactly what is intended, so the warning has been cut out in that case,
6870 and Exim is allowed to try to write to its normal log files.
6876 1. smtp_return_error_details was not giving details for temporary sender
6877 or receiver verification errors.
6879 2. Diagnose a configuration error if two authenticators have the same public
6882 3. Exim used not to create the message log file for a message until the first
6883 delivery attempt. This could be confusing when incoming messages were held
6884 for policy or load reasons. The message log file is now created at the time
6885 the message is received, and an initial "Received" line is written to it.
6887 4. The automatically generated man page for command line options had a minor
6888 bug that caused no ill effects; however, a more serious problem was that
6889 the procedure for building the man page automatically didn't always
6890 operate. Consequently, release 4.20 contains an out-of-date version. This
6891 shouldn't happen again.
6893 5. When building Exim with embedded Perl support, the script that builds the
6894 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6895 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6897 6. The freeze_tell option was not being used for messages that were frozen on
6898 arrival, either by an ACL or by local_scan().
6900 7. Added the smtp_incomplete_transaction log selector.
6902 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6903 was accepting AUTH without a new EHLO.
6905 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6906 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6909 10. Logging of TCP/IP connections (when configured) now happens in the main
6910 daemon process instead of the child process, so that the TCP/IP connection
6911 count is more accurate (but it can never be perfect).
6913 11. The use of "drop" in a nested ACL was not being handled correctly in the
6914 outer ACL. Now, if condition failure induced by the nested "drop" causes
6915 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6916 or "require"), the connection is dropped.
6918 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6919 that yield "discard" can now be used with an "accept" or a "discard" verb,
6920 but an error is generated for any others (because I can't see a useful way
6921 to define what should happen).
6923 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6924 are now processed in the same way as lines in the Exim configuration file.
6925 In particular, continuation lines are supported.
6927 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6929 15. Added -ti meaning -t -i.
6931 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6932 domains, and warn by logging if others are found.
6934 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6935 returned from GIFCONF when Exim is trying to find the list of interfaces on
6936 a host. The code in os.c has been modified to copy each ifreq to an aligned
6937 structure in all cases.
6939 Also, in some cases, the returned ifreq's were being copied to a 'struct
6940 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6941 means the last couple of bytes of an IPv6 address could be chopped if the
6942 ifreq contained only a normal sockaddr (14 bytes storage).
6944 18. Named domain lists were not supported in the hosts_treat_as_local option.
6945 An entry such as +xxxx was not recognized, and was treated as a literal
6948 19. Ensure that header lines added by a DATA ACL are included in the reject log
6949 if the ACL subsequently rejects the message.
6951 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6952 MD5 (which is deprecated).
6954 21. When testing a filter file using -bf, Exim was writing a message when it
6955 took the sender from a "From " line in the message, but it was not doing so
6956 when it took $return_path from a Return-Path: header line. It now does.
6958 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6959 with a valid header line field name (a series of printing characters
6960 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6962 23. Changed "disc" in the source to "disk" to conform to the documentation and
6963 the book and for uniformity.
6965 24. Ignore Sendmail's -Ooption=value command line item.
6967 25. When execve() failed while trying to run a command in a pipe transport,
6968 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6969 could be confused with a return value of 69 from the command itself. This
6970 has been changed to 127, the value the shell returns if it is asked to run
6971 a non-existent command. The wording for the related log line suggests a
6972 non-existent command as the problem.
6974 26. If received_header_text expands to an empty string, do not add a Received:
6975 header line to the message. (Well, it adds a token one on the spool, but
6976 marks it "old" so that it doesn't get used or transmitted.)
6978 27. Installed eximstats 1.28 (addition of -nt option).
6980 28. There was no check for failure on the call to getsockname() in the daemon
6981 code. This can fail if there is a shortage of resources on the system, with
6982 ENOMEM, for example. A temporary error is now given on failure.
6984 29. Contrary to the C standard, it seems that in some environments, the
6985 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6986 program. Exim now does this explicitly; it affects the formatting of
6987 timestamps using strftime().
6989 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6990 complaints. I've now initialized all the variables, to avoid this.
6992 32. Header lines added by a system filter were not being "seen" during
6993 transport-time rewrites.
6995 33. The info_callback() function passed to OpenSSL is set up with type void
6996 (*)(SSL *, int, int), as described somewhere. However, when calling the
6997 function (actually a macro) that sets it up, the type void(*)() is
6998 expected. I've put in a cast to prevent warnings from picky compilers.
7000 34. If a DNS black list lookup found a CNAME record, but there were no A
7001 records associated with the domain it pointed at, Exim crashed.
7003 35. If a DNS black list lookup returned more than one A record, Exim ignored
7004 all but the first. It now scans all returned addresses if a particular IP
7005 value is being sought. In this situation, the contents of the
7006 $dnslist_value variable are a list of all the addresses, separated by a
7009 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7010 to accept a host name and all its aliases if the forward lookup for any of
7011 them yielded the IP address of the incoming connection. Now it accepts only
7012 those names whose forward lookup yields the correct IP address. Any other
7013 names are discarded. This closes a loophole whereby a rogue DNS
7014 administrator could create reverse DNS records to break through a
7015 wildcarded host restriction in an ACL.
7017 37. If a user filter or a system filter that ran in a subprocess used any of
7018 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7019 the wrong values were passed to the pipe command ($thisaddress had the
7020 value of $0, $0 had the value of $1, etc). This bug was introduced by
7021 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7023 38. Improved the line breaking for long SMTP error messages from ACLs.
7024 Previously, if there was no break point between 40 and 75 characters, Exim
7025 left the rest of the message alone. Two changes have been made: (a) I've
7026 reduced the minimum length to 35 characters; (b) if it can't find a break
7027 point between 35 and 75 characters, it looks ahead and uses the first one
7028 that it finds. This may give the occasional overlong line, but at least the
7029 remaining text gets split now.
7031 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7032 file descriptors might be low, and that setting 1000 would always raise it.
7033 It turns out that in some environments, the limit is already over 1000 and
7034 that lowering it causes trouble. So now Exim takes care not to decrease it.
7036 40. When delivering a message, the value of $return_path is set to $sender_
7037 address at the start of routing (routers may change the value). By an
7038 oversight, this default was not being set up when an address was tested by
7039 -bt or -bv, which affected the outcome if any router or filter referred to
7042 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7043 log only when "message" or "log_message" are set. However, if one of the
7044 conditions was an address verification, or a call to a nested ACL, the
7045 messages generated by the underlying test were being passed through. This
7046 no longer happens. The underlying message is available in $acl_verify_
7047 message for both "message" and "log_message" expansions, so it can be
7048 passed through if needed.
7050 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7051 new expansion $bh_ to give the encoded byte string without charset
7052 translation. Translation happens only if iconv() is available; HAVE_ICONV
7053 indicates this at build time. HEADERS_CHARSET gives the charset to
7054 translate to; headers_charset can change it in the configuration, and
7055 "headers charset" can change it in an individual filter file.
7057 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7058 that creates RFC 2047 encoded "words" labels them as that charset instead
7059 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7060 expansion operator; (ii) when Exim creates a From: line for a local
7061 message; (iii) when a header line is rewritten to include a "phrase" part.
7063 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7064 buggy, causing it to skip the first lines of messages whose message ID
7065 ended in 'D'. This would not have bitten before Exim release 4.14, because
7066 message IDs were unlikely to end in 'D' before then. The effect was to have
7067 incorrect size information for certain domains.
7069 45. #include "config.h" was missing at the start of the crypt16.c module. This
7070 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7073 46. If there was a timeout during a "random" callout check, Exim treated it as
7074 a failure of the random address, and carried on sending RSET and the real
7075 address. If the delay was just some slowness somewhere, the response to the
7076 original RCPT would be taken as a response to RSET and so on, causing
7077 mayhem of various kinds.
7079 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7080 when I implemented it. It didn't allow for the fact that some option values
7081 may legitimately be negative (e.g. size_addition), and it didn't even do
7082 the right test for positive values.
7084 48. Domain names in DNS records are case-independent. Exim always looks them up
7085 in lower case. Some resolvers return domain names in exactly the case they
7086 appear in the zone file, that is, they may contain uppercase letters. Not
7087 all resolvers do this - some return always lower case. Exim was treating a
7088 change of case by a resolver as a change of domain, similar to a widening
7089 of a domain abbreviation. This triggered its re-routing code and so it was
7090 trying to route what was effectively the same domain again. This normally
7091 caused routing to fail (because the router wouldn't handle the domain
7092 twice). Now Exim checks for this case specially, and just changes the
7093 casing of the domain that it ultimately uses when it transmits the message
7096 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7099 50. If a filter generated a file delivery with a non-absolute name (possible if
7100 no home directory exists for the router), the forbid_file option was not
7103 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7104 the existing equality matching.
7106 52. Exim was using ints instead of ino_t variables in some places where it was
7107 dealing with inode numbers.
7109 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7110 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7111 called TMPDIR, and if it finds it is different, it changes its value.
7113 54. The smtp_printf() function is now made available to local_scan() so
7114 additional output lines can be written before returning. There is also an
7115 smtp_fflush() function to enable the detection of a dropped connection.
7116 The variables smtp_input and smtp_batched_input are exported to
7119 55. Changed the default runtime configuration: the message "Unknown user"
7120 has been removed from the ACL, and instead placed on the localuser router,
7121 using the cannot_route_message feature. This means that any verification
7122 failures that generate their own messages won't get overridden. Similarly,
7123 the "Unrouteable address" message that was in the ACL for unverifiable
7124 relay addresses has also been removed.
7126 56. Added hosts_avoid_esmtp to the smtp transport.
7128 57. The exicyclog script was not checking for the esoteric option
7129 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7130 will work only if exicyclog is run under the appropriate euid.
7132 58. Following a discussion on the list, the rules by which Exim recognises line
7133 endings on incoming messages have been changed. The -dropcr and drop_cr
7134 options are now no-ops, retained only for backwards compatibility. The
7135 following line terminators are recognized: LF CRLF CR. However, special
7136 processing applies to CR:
7138 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7139 nor a local message in the state where . is a terminator.
7141 (ii) If a bare CR is encountered in a header line, an extra space is added
7142 after the line terminator so as not to end the header. The reasoning
7143 behind this is that bare CRs in header lines are most likely either
7144 to be mistakes, or people trying to play silly games.
7146 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7147 was being incorrectly given as 18 bytes larger than it should have been.
7148 This is a VOB (very old bug).
7150 60. This may never have affected anything current, but just in case it has:
7151 When the local host is found other than at the start of a list of hosts,
7152 the local host, those with the same MX, and any that follow, are discarded.
7153 When the list in question was part of a longer list of hosts, the following
7154 hosts (not currently being processed) were also being discarded. This no
7155 longer happens. I'm not sure if this situation could ever has previously
7158 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7161 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7162 header. This is something that is recommended in a new Internet Draft, and
7163 is something that is documented as being done by Sendmail. There are two
7164 possible values. For messages generated by the autoreply transport, Exim
7167 Auto-Submitted: auto-replied
7169 whereas for all other generated messages (e.g. bounces) it adds
7171 Auto-Submitted: auto-generated
7173 63. The "personal" condition in filters now includes a test for the
7174 Auto-Submitted: header. If it contains the string "auto-" the message it
7175 not considered personal.
7177 64. Added rcpt_include_affixes as a generic transport option.
7179 65. Added queue_only_override (default true).
7181 66. Added the syslog_duplication option.
7183 67. If what should have been the first header line of a message consisted of
7184 a space followed by a colon, Exim was mis-interpreting it as a header line.
7185 It isn't of course - it is syntactically invalid and should therefore be
7186 treated as the start of the message body. The misbehaviour could have
7187 caused a number of strange effects, including loss of data in subsequent
7188 header lines, and spool format errors.
7190 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7191 client host had authenticated. This control can now be exercised by an ACL
7192 for more flexibility.
7194 69. By default, callouts do not happen when testing with -bh. There is now a
7195 variant, -bhc, which does actually run the callout code, including
7196 consulting and updating the callout cache.
7198 70. Added support for saslauthd authentication, courtesy of Alexander
7201 71. If statvfs() failed on the spool or log directories while checking their
7202 size for availability, Exim confusingly gave the error "space shortage".
7203 Furthermore, in debugging mode it crashed with a floating point exception.
7204 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7205 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7206 serious problem, Exim now writes to the main and panic logs when this
7207 happens, with details of the failure. It then refuses to accept the
7208 incoming message, giving the message "spool directory problem" or "log
7209 directory problem" with a 421 code for SMTP messages.
7211 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7212 0, 1, and 2 exist, because some OS complain for execs without them (see
7213 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7214 descriptors. However, the code omitted to check that the open succeeded,
7215 causing mysterious errors if for some reason the permissions on /dev/null
7216 got screwed. Now Exim writes a message to the main and panic logs, and
7217 bombs out if it can't open /dev/null.
7219 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7220 interact so that it is all more flexible. It is supposed to remain
7221 backwards compatible. Also added extra_local_interfaces.
7223 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7224 to bomb out with an assertion failure - to the client this appears as a
7225 connection drop. This problem occurs in the part of the code that was taken
7226 from the Samba project. Fortunately, the assertion is in a very simple
7227 function, so I have fixed this by reproducing the function inline in the
7228 one place where it is called, and arranging for authentication to fail
7229 instead of killing the process with assert().
7231 75. The SPA client code was not working when the server requested OEM rather
7232 than Unicode encoding.
7234 76. Added code to make require_files with a specific uid setting more usable in
7235 the case where statting the file as root fails - usually a non-root-mounted
7236 NFS file system. When this happens and the failure is EACCES, Exim now
7237 forks a subprocess and does the per-uid checking as the relevant uid.
7239 77. Added process_log_path.
7241 78. If log_file_path was not explicitly set, a setting of check_log_space or
7242 check_log_inodes was ignored.
7244 79. If a space check for the spool or log partitions fails, the incident is now
7245 logged. Of course, in the latter case the data may get lost...
7247 80. Added the %p formatting code to string_format() so that it can be used to
7248 print addresses in debug_print(). Adjusted all the address printing in the
7249 debugging in store.c to use %p rather than %d.
7251 81. There was a concern that a line of code in smtp_in.c could overflow a
7252 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7253 initially expressed, the concern was not well-founded, because trailing
7254 spaces are removed early. However, if the trailing spaces were followed by
7255 a NULL, they did not get removed, so the overflow was possible. Two fixes
7258 (a) I re-wrote the offending code in a cleaner fashion.
7259 (b) If an incoming SMTP command contains a NULL character, it is rejected
7262 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7263 runs initgroups(), so that if the Exim user is in any additional groups,
7264 they will be used during message reception.
7270 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob