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.
101 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
102 should pack well giving a smaller memory footprint so better caching and
103 therefore performance. Group the declarations where this can't be done so
104 that the byte-sized flag variables are not interspersed among pointer
105 variables, giving a better chance of good packing by the compiler.
107 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
108 non-null, to avoid issues with sites running BATV. Previously reports were
109 sent with an empty envelope sender so looked like bounces.
111 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
112 The ignore_error flag wasn't being returned from the filter subprocess so
113 was not set for later routers. Investigation and fix by Matthias Kurz.
115 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
116 and a msg:complete for the whole, when a message is manually removed using
117 -Mrm. Developement by Matthias Kurz, hacked on by JH.
119 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
120 a "Gnu special" function, asprintf() in the DB utility binary builds; I
121 hope that is portable enough.
123 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
124 requiring a known-CA anchor certificate; make it now rely entirely on the
125 TLSA as an anchor. Checking the name on the leaf cert against the name
126 on the A-record for the host is still done for TA (but not for EE mode).
128 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
129 included in delivery lines for non-proxied connections, when compiled with
130 SUPPORT_SOCKS and running with proxy logging enabled.
136 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
137 When redis_servers is set to a list of > 1 element, and the Redis servers
138 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
139 case of MOVED into a DEFER case instead, thus moving the query onto the
140 next server in the list. For a cluster of N elements, all N servers must
141 be defined in redis_servers.
143 GF/02 Catch and remove uninitialized value warning in exiqsumm
144 Check for existence of @ARGV before looking at $ARGV[0]
146 JH/01 Replace the store_release() internal interface with store_newblock(),
147 which internalises the check required to safely use the old one, plus
148 the allocate and data copy operations duplicated in both (!) of the
149 extant use locations.
151 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
152 modifier. This matches the restriction on the commandline.
154 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
155 Previously only the last row was returned.
157 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
158 we assumed that tags in the header were well-formed, and parsed the
159 element content after inspecting only the first char of the tag.
160 Assumptions at that stage could crash the receive process on malformed
163 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
164 While running the DKIM ACL we operate on the Permanent memory pool so that
165 variables created with "set" persist to the DATA ACL. Also (at any time)
166 DNS lookups that fail create cache records using the Permanent pool. But
167 expansions release any allocations made on the current pool - so a dnsdb
168 lookup expansion done in the DKIM ACL releases the memory used for the
169 DNS negative-cache, and bad things result. Solution is to switch to the
170 Main pool for expansions.
171 While we're in that code, add checks on the DNS cache during store_reset,
172 active in the testsuite.
173 Problem spotted, and debugging aided, by Wolfgang Breyha.
175 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
176 When none of the hosts presented to a transport match an already-open
177 connection, close it and proceed with the list. Previously we would
178 queue the message. Spotted by Lena with Yahoo, probably involving
181 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
182 Previously a spurious "250 OK id=" response was appended to the proper
185 JH/08 The "support for" informational output now, which built with Content
186 Scanning support, has a line for the malware scanner interfaces compiled
187 in. Interface can be individually included or not at build time.
189 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
190 by the template makefile "src/EDITME". The "STREAM" support for an older
191 ClamAV interface method is removed.
193 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
194 rows affected is given instead).
196 JH/11 The runtime Berkeley DB library version is now additionally output by
197 "exim -d -bV". Previously only the compile-time version was shown.
199 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
200 SMTP connection. Previously, when one had more receipients than the
201 first, an abortive onward connection was made. Move to full support for
202 multiple onward connections in sequence, handling cutthrough connection
203 for all multi-message initiating connections.
205 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
206 routers. Previously, a multi-recipient message would fail to match the
207 onward-connection opened for the first recipient, and cause its closure.
209 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
210 a timeout on read on a GnuTLS initiating connection, resulting in the
211 initiating connection being dropped. This mattered most when the callout
212 was marked defer_ok. Fix to keep the two timeout-detection methods
215 JH/15 Relax results from ACL control request to enable cutthrough, in
216 unsupported situations, from error to silently (except under debug)
217 ignoring. This covers use with PRDR, frozen messages, queue-only and
220 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
222 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
223 metadata, resulting in a crash in free().
225 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
226 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
227 Broken also in d185889f4, with init system revamp.
229 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
230 Previously we abruptly closed the connection after reading a malware-
231 found indication; now we go on to read the "scan ok" response line,
234 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
235 ACL. Previously, a crash would result.
237 JH/19 Speed up macro lookups during configuration file read, by skipping non-
238 macro text after a replacement (previously it was only once per line) and
239 by skipping builtin macros when searching for an uppercase lead character.
241 JH/20 DANE support moved from Experimental to mainline. The Makefile control
242 for the build is renamed.
244 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
245 was allocated for every new TLS startup, meaning one per message. Fix
246 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
248 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
249 reported the original. Fix to report (as far as possible) the ACL
250 result replacing the original.
252 JH/23 Fix memory leak during multi-message connections using STARTTLS under
253 OpenSSL. Certificate information is loaded for every new TLS startup,
254 and the resources needed to be freed.
256 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
258 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
261 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
262 DATA response info to the (existing) per-recipient response info for
263 the "C=" log element. It can have useful tracking info from the
264 destination system. Patch from Simon Arlott.
266 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
267 length value. Previously this would segfault.
269 HS/02 Support Avast multiline protoocol, this allows passing flags to
270 newer versions of the scanner.
272 JH/28 Ensure that variables possibly set during message acceptance are marked
273 dead before release of memory in the daemon loop. This stops complaints
274 about them when the debug_store option is enabled. Discovered specifically
275 for sender_rate_period, but applies to a whole set of variables.
276 Do the same for the queue-runner and queue-list loops, for variables set
277 from spool message files. Do the same for the SMTP per-message loop, for
278 certain variables indirectly set in ACL operations.
280 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
281 as a multi-recipient message from a mailinglist manager). The coding had
282 an arbitrary cutoff number of characters while checking for more input;
283 enforced by writing a NUL into the buffer. This corrupted long / fast
284 input. The problem was exposed more widely when more pipelineing of SMTP
285 responses was introduced, and one Exim system was feeding another.
286 The symptom is log complaints of SMTP syntax error (NUL chars) on the
287 receiving system, and refused recipients seen by the sending system
288 (propating to people being dropped from mailing lists).
289 Discovered and pinpointed by David Carter.
291 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
292 replaced by the ${authresults } expansion.
294 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
296 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
297 allows proper process termination in container environments.
299 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
300 Previously the "final dot" had a newline after it; ensure it is CR,LF.
302 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
303 and "err_perm", deprecated since 4.83 when the RFC-defined words
304 "temperror" and "permerror" were introduced.
306 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
307 transport-filters or DKIM-signing. The restriction was lost in the
308 consolidation of verify-callout and delivery SMTP handling.
309 Extend the restriction to also cover ARC-signing.
311 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
312 in defer=pass mode supply a 450 to the initiator. Previously the message
315 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
316 tls_require_ciphers is used as before.
318 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
319 "pass_unscanned". Only tmpfails from the scanner are written to
320 the paniclog, as they may require admin intervention (permission
321 denied, license issues). Other scanner errors (like decompression
322 bombs) do not cause a paniclog entry.
324 JH/36 Fix reinitialisation of DKIM logging variable between messages.
325 Previously it was possible to log spurious information in receive log
328 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
329 triggered odd behaviour from Outlook Express clients.
331 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
334 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
335 since the IETF WG has not yet settled on that versus the original
336 "bare" representation.
338 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
339 Previously the millisecond value corrupted the output.
340 Fix also for syslog_pid=no and log_selector +pid, for which the pid
341 corrupted the output.
347 JH/01 Rework error string handling in TLS interface so that the caller in
348 more cases is responsible for logging. This permits library-sourced
349 string to be attached to addresses during delivery, and collapses
350 pairs of long lines into single ones.
352 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
353 during configuration. Wildcards are allowed and expanded.
355 JH/02 Rework error string handling in DKIM to pass more info back to callers.
356 This permits better logging.
358 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
359 do not close it down and have the child transport start it up again on
360 the passed-on TCP connection. Instead, proxy the child (and any
361 subsequent ones) for TLS via a unix-domain socket channel. Logging is
362 affected: the continued delivery log lines do not have any DNSSEC, TLS
363 Certificate or OCSP information. TLS cipher information is still logged.
365 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
366 identical IP addresses on different listening ports. Will also affect
369 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
370 add noisy ifdef guards to special-case this sillyness.
371 Patch from Bernd Kuhls.
373 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
374 than 255 are no longer allowed.
376 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
377 Disable the session-cache too, which might reduce our load. Since we
378 currrectly use a new context for every connection, both as server and
379 client, there is no benefit for these.
380 GnuTLS appears to not support tickets server-side by default (we don't
381 call gnutls_session_ticket_enable_server()) but client side is enabled
382 by default on recent versions (3.1.3 +) unless the PFS priority string
385 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
386 <https://reproducible-builds.org/specs/source-date-epoch/>.
388 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
389 the check for any unsuccessful recipients did not notice the limit, and
390 erroneously found still-pending ones.
392 JH/08 Pipeline CHUNKING command and data together, on kernels that support
393 MSG_MORE. Only in-clear (not on TLS connections).
395 JH/09 Avoid using a temporary file during transport using dkim. Unless a
396 transport-filter is involved we can buffer the headers in memory for
397 creating the signature, and read the spool data file once for the
398 signature and again for transmission.
400 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
401 4.77 as the kernel support then wasn't solid, having issues in 64bit
402 mode. Now, it's been long enough. Add support for FreeBSD also.
404 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
405 case where the routing stage had gathered several addresses to send to
406 a host before calling the transport for the first, we previously failed
407 to close down TLS in the old transport process before passing the TCP
408 connection to the new process. The new one sent a STARTTLS command
409 which naturally failed, giving a failed delivery and bloating the retry
410 database. Investigation and fix prototype from Wolfgang Breyha.
412 JH/12 Fix check on SMTP command input synchronisation. Previously there were
413 false-negatives in the check that the sender had not preempted a response
414 or prompt from Exim (running as a server), due to that code's lack of
415 awareness of the SMTP input buffering.
417 PP/04 Add commandline_checks_require_admin option.
418 Exim drops privileges sanely, various checks such as -be aren't a
419 security problem, as long as you trust local users with access to their
420 own account. When invoked by services which pass untrusted data to
421 Exim, this might be an issue. Set this option in main configuration
422 AND make fixes to the calling application, such as using `--` to stop
425 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
426 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
427 responses to those, into a single TLS record each way (this usually means
428 a single packet). As a side issue, smtp_enforce_sync now works on TLS
431 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
432 affects you only if you're dancing at the edge of the param size limits.
433 If you are, and this message makes sense to you, then: raise the
434 configured limit or use OpenSSL 1.1. Nothing we can do for older
437 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
438 cmdline element to get the documented default one. Previously it was
441 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
442 are unsuable for non-message-specific reasons - eg. connection timeout,
443 banner-time rejection.
445 JH/16 Fix logging of delivery remote port, when specified by router, under
448 PP/06 Repair manualroute's ability to take options in any order, even if one
449 is the name of a transport.
452 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
454 JH/17 Change the list-building routines interface to use the expanding-string
455 triplet model, for better allocation and copying behaviour.
457 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
458 Previously it was constructed the first time a possibly-matching string
459 was met in the configuration file input during startup; now it is done
462 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
463 compatible one, to avoid the (poorly documented) possibility of a config
464 file in the working directory redirecting the DB files, possibly correpting
465 some existing file. CVE-2017-10140 assigned for BDB.
467 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
468 cache-hot. Previously, although the result was properly cached, the
469 initial verify call returned a defer.
471 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
472 the main verify for receipient in uncached-mode.
474 JH/22 Retire historical build files to an "unsupported" subdir. These are
475 defined as "ones for which we have no current evidence of testing".
477 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
478 if present. Previously it was ignored.
480 JH/24 Start using specified-initialisers in C structure init coding. This is
481 a C99 feature (it's 2017, so now considered safe).
483 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
484 if was a fixed-sized field and bitmask ops via macros; it is now more
487 PP/07 GitHub PR 56: Apply MariaDB build fix.
488 Patch provided by Jaroslav Škarvada.
490 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
491 during Coverity cleanups [4.87 JH/47]
492 Diagnosis and fix provided by Michael Fischer v. Mollard.
494 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
495 the right size to place the terminating semicolon on its own folded
496 line, the header hash was calculated to an incorrect value thanks to
497 the (relaxed) space the fold became.
499 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
500 and confused the parent.
502 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
503 which could crash as a result. This could lead to undeliverable messages.
505 JH/28 Logging: "next input sent too soon" now shows where input was truncated
508 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
509 matters on fast-turnover and PID-randomising systems, which were getting
510 out-of-order delivery.
512 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
513 a possibly-overlapping copy. The symptom was that "Remote host closed
514 connection in response to HELO" was logged instead of the actual 4xx
517 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
518 Previously only that bufferd was discarded, resulting in SYMTP command
521 JH/32 DKIM: when a message has multiple signatures matching an identity given
522 in dkim_verify_signers, run the dkim acl once for each. Previously only
523 one run was done. Bug 2189.
525 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
526 "panic the current process" to "deliberately defer". The panic log is
527 still written with the problem list name; the mail and reject logs now
528 get a temp-reject line for the message that was being handled, saying
529 something like "domains check lookup or other defer". The SMTP 451
530 message is still "Temporary local problem".
532 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
533 A crafted sequence of BDAT commands could result in in-use memory beeing
534 freed. CVE-2017-16943.
536 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
537 from SMTP input. Previously it was always done; now only done for DATA
538 and not BDAT commands. CVE-2017-16944.
540 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
541 to the message (such as an overlong header line). Previously this was
542 not done and we did not exit BDAT mode. Followon from the previous item
543 though a different problem.
549 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
550 than -2003 did; needs libidn2 in addition to libidn.
552 JH/02 The path option on a pipe transport is now expanded before use.
554 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
555 Patch provided by "Björn", documentation fix added too.
557 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
558 missing a wire-to-host endian conversion.
560 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
561 close after a BDAT command line could be taken as a following command,
562 giving a synch failure. Fix by only checking for synch immediately
563 before acknowledging the chunk.
565 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
566 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
567 macro. Patches provided by Josh Soref.
569 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
570 Previously we did not; the RFC seems ambiguous and VRFY is not listed
571 by IANA as a service extension. However, John Klensin suggests that we
574 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
575 the dkim code may be unix-mode line endings rather than smtp wire-format
576 CRLF, so prepend a CR to any bare LF.
578 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
579 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
581 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
582 the first were themselves being wrongly included in the feed into dkim
583 processing; with most chunk sizes in use this resulted in an incorrect
584 body hash calculated value.
586 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
587 DKIM signature block, for verification. Although advised against by
588 standards it is specifically not ruled illegal.
590 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
592 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
593 missing a body hash (the bh= tag).
595 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
596 It seems that HAProxy sends the Proxy Protocol information in clear and
597 only then does a TLS startup, so do the same.
599 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
600 TCP connections (such as for Spamd) unless the daemon successfully set
601 Fast Open mode on its listening sockets. This fixes breakage seen on
602 too-old kernels or those not configured for Fast Open, at the cost of
603 requiring both directions being enabled for TFO, and TFO never being used
604 by non-daemon-related Exim processes.
606 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
607 endings, at least on the first header line. Try to canonify any that get
608 past that check, despite the cost.
610 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
611 now limited to an arbitrary five deep, while parsing addresses with the
612 strip_excess_angle_brackets option enabled.
614 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
615 instead leave the unprompted TLS handshake in socket buffer for the
616 TLS library to consume.
618 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
620 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
622 JH/16 Drop variables when they go out of scope. Memory management drops a whole
623 region in one operation, for speed, and this leaves assigned pointers
624 dangling. Add checks run only under the testsuite which checks all
625 variables at a store-reset and panics on a dangling pointer; add code
626 explicitly nulling out all the variables discovered. Fixes one known
627 bug: a transport crash, where a dangling pointer for $sending_ip_address
628 originally assigned in a verify callout, is re-used.
630 PP/06 Drop '.' from @INC in various Perl scripts.
632 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
634 PP/08 Reduce a number of compilation warnings under clang; building with
635 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
636 should be warning-free.
638 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
640 HS/01 Fix portability problems introduced by PP/08 for platforms where
641 realloc(NULL) is not equivalent to malloc() [SunOS et al].
643 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
644 chunk. This allows us to accept broken chunked messages. We need a more
645 general solution here.
647 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
648 already-broken messages in the queue.
650 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
652 JH/19 Fix reference counting bug in routing-generated-address tracking.
658 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
659 supports it and a size is available (ie. the sending peer gave us one).
661 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
662 years of being deprecated). The replacements are the ACLs
663 acl_smtp_mime and acl_not_smtp_mime.
665 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
666 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
667 or even an in-clear connection were permitted. Now, if the host lookup
668 was dnssec and dane was requested then the host is only used if the
669 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
670 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
671 if one fails this test.
672 This means that a poorly-configured remote DNS will make it incommunicado;
673 but it protects against a DNS-interception attack on it.
675 JH/04 Bug 1810: make continued-use of an open smtp transport connection
676 non-noisy when a race steals the message being considered.
678 JH/05 If main configuration option tls_certificate is unset, generate a
679 self-signed certificate for inbound TLS connections.
681 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
682 in rewrites and routers.
684 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
685 and logged a warning sing 4.83; now they are a configuration file error.
687 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
688 (lacking @domain). Apply the same qualification processing as RCPT.
690 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
692 JH/10 Support ${sha256:} applied to a string (as well as the previous
695 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
696 a cutthrough deliver is pending, as we always want to make a connection.
697 This also avoids re-routing the message when later placing the cutthrough
698 connection after a verify cache hit.
699 Do not update it with the verify result either.
701 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
702 when routing results in more than one destination address.
704 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
705 signing (which inhibits the cutthrough capability). Previously only
706 the presence of an option was tested; now an expansion evaluating as
707 empty is permissible (obviously it should depend only on data available
708 when the cutthrough connection is made).
710 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
711 the relevant preceding SMTP command did not note the pipelining mode.
713 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
714 Previously they were not counted.
716 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
717 as one having no matching records. Previously we deferred the message
718 that needed the lookup.
720 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
721 distinguished as "(=".
723 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
724 for missing MX records. Previously it only worked for missing A records.
726 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
728 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
729 after the data-go-ahead and data-ack. Patch from Jason Betts.
731 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
732 even for a "none" policy. Patch from Tony Meyer.
734 JH/22 Fix continued use of a connection for further deliveries. If a port was
735 specified by a router, it must also match for the delivery to be
738 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
739 When one of the recipients of a message was unsuitable for the connection
740 (has no matching addresses), we lost track of needing to mark it
741 deferred. As a result mail would be lost.
743 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
745 JH/25 Decoding ACL controls is now done using a binary search; the source code
746 takes up less space and should be simpler to maintain. Merge the ACL
747 condition decode tables also, with similar effect.
749 JH/26 Fix problem with one_time used on a redirect router which returned the
750 parent address unchanged. A retry would see the parent address marked as
751 delivered, so not attempt the (identical) child. As a result mail would
754 JH/27 Fix a possible security hole, wherein a process operating with the Exim
755 UID can gain a root shell. Credit to http://www.halfdog.net/ for
756 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
759 JH/28 Enable {spool,log} filesystem space and inode checks as default.
760 Main config options check_{log,spool}_{inodes,space} are now
761 100 inodes, 10MB unless set otherwise in the configuration.
763 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
764 Previously it only applied to the main-section connection policy
767 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
769 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
770 by me. Added RFC7919 DH primes as an alternative.
772 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
773 are not in the system include path.
775 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
776 GnuTLS, when a session startup failed (eg because the client disconnected)
777 Exim did stdio operations after fclose. This was exposed by a recent
778 change which nulled out the file handle after the fclose.
780 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
781 signed directly by the cert-signing cert, rather than an intermediate
782 OCSP-signing cert. This is the model used by LetsEncrypt.
784 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
786 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
787 an incoming connection.
789 HS/02 Bug 1802: Do not half-close the connection after sending a request
792 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
793 fallback to "prime256v1".
795 JH/34 SECURITY: Use proper copy of DATA command in error message.
796 Could leak key material. Remotely exploitable. CVE-2016-9963.
802 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
803 and 3.4.4 - once the server is enabled to respond to an OCSP request
804 it does even when not requested, resulting in a stapling non-aware
805 client dropping the TLS connection.
807 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
808 support variable-length bit vectors. No functional change.
810 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
811 The I= interface field on outgoing lines is now after the H= remote
812 host field, same as incoming lines. There is a separate
813 outgoing_interface log selector which allows you to disable the
816 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
817 If not running log_selector +smtp_connection the mainlog would be held
818 open indefinitely after a "too many connections" event, including to a
819 deleted file after a log rotate. Leave the per net connection logging
820 leaving it open for efficiency as that will be quickly detected by the
821 check on the next write.
823 HS/01 Bug 1671: Fix post transport crash.
824 Processing the wait-<transport> messages could crash the delivery
825 process if the message IDs didn't exist for some reason. When
826 using 'split_spool_directory=yes' the construction of the spool
827 file name failed already, exposing the same netto behaviour.
829 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
830 mime_regex ACL conditions.
832 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
833 to DSN fail messages (bounces): remote IP, remote greeting, remote response
834 to HELO, local diagnostic string.
836 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
837 log line to debug. Even when configured with a tls authenticator many
838 client connections are expected to not authenticate in this way, so
839 an authenticate fail is not an error.
841 HS/02 Add the Exim version string to the process info. This way exiwhat
842 gives some more detail about the running daemon.
844 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
845 matter for fast-change records such as DNSBLs.
847 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
848 retry record, even if constant. There may be multiple transports with
849 different interface settings and the retry behaviour needs to be kept
852 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
854 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
856 JH/10 Bug 840: fix log_defer_output option of pipe transport
858 JH/11 Bug 830: use same host for all RCPTS of a message, even under
859 hosts_randomize. This matters a lot when combined with mua_wrapper.
861 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
862 ${quote_pgsql:<string>} operator.
864 JH/13 Bug 1708: avoid misaligned access in cached lookup.
866 JH/14 Change header file name for freeradius-client. Relevant if compiling
867 with Radius support; from the Gentoo tree and checked under Fedora.
869 JH/15 Bug 1712: Introduce $prdr_requested flag variable
871 JH/16 Bug 1714: Permit an empty string as expansion result for transport
872 option transport_filter, meaning no filtering.
874 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
876 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
877 defaults to "*" (all hosts). The variable is now available when not built
878 with TLS, default unset, mainly to enable keeping the testsuite sane.
879 If a server certificate is not supplied (via tls_certificate) an error is
880 logged, and clients will find TLS connections fail on startup. Presumably
881 they will retry in-clear.
882 Packagers of Exim are strongly encouraged to create a server certificate
883 at installation time.
885 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
886 with the $config_file variable.
888 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
889 in transport context, after the attempt, and per-recipient. The latter type
890 is per host attempted. The event data is the error message, and the errno
891 information encodes the lookup type (A vs. MX) used for the (first) host,
892 and the trailing two digits of the smtp 4xx response.
894 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
895 to write to mainlog (or rejectlog, paniclog) in the window between file
896 creation and permissions/ownership being changed. Particularly affects
897 installations where exicyclog is run as root, rather than exim user;
898 result is that the running daemon panics and dies.
900 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
902 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
903 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
904 "pri" and "weight". Note that the previous implicit priority given by the
905 list order is no longer honoured.
907 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
910 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
911 by defining SUPPORT_SOCKS.
913 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
914 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
915 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
916 variables are renamed to proxy_{local,external}_{address,port}.
918 JH/27 Move Internationalisation support from Experimental to mainline, enabled
919 for a build by defining SUPPORT_I18N
921 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
922 of the query string, and make ${quote_redis:} do that quoting.
924 JH/29 Move Events support from Experimental to mainline, enabled by default
925 and removable for a build by defining DISABLE_EVENT.
927 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
929 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
930 cached by the daemon.
932 JH/32 Move Redis support from Experimental to mainline, enabled for a build
933 by defining LOOKUP_REDIS. The libhiredis library is required.
935 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
936 keys are given for lookup.
938 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
939 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
940 only supported when built with TLS support. The PolarSSL SHA routines
941 are still used when the TLS library is too old for convenient support.
943 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
944 openssl_options), for security. OpenSSL forces this from version 1.1.0
945 server-side so match that on older versions.
947 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
948 allocation for $value could be released as the expansion processing
949 concluded, but leaving the global pointer active for it.
951 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
952 and to use the domains and local_parts ACL conditions.
954 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
955 incorrectly not doubled on cutthrough transmission, hence seen as a
956 body-termination at the receiving system - resulting in truncated mails.
957 Commonly the sender saw a TCP-level error, and retransmitted the message
958 via the normal store-and-forward channel. This could result in duplicates
959 received - but deduplicating mailstores were liable to retain only the
960 initial truncated version.
962 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
964 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
966 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
967 we're in there, support oversigning also; bug 1309.
969 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
971 HS/04 Add support for keep_environment and add_environment options.
973 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
974 either intentional arithmetic overflow during PRNG, or testing config-
977 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
978 delivery resulted in actual delivery. Cancel cutthrough before DATA
981 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
982 rejects a recipient: pass the reject to the originator.
984 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
985 Many were false-positives and ignorable, but it's worth fixing the
988 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
989 for the new environment-manipulation done at startup. Move the routines
990 from being local to tls.c to being global via the os.c file.
992 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
993 an extract embedded as result-arg for a map, the first arg for extract
994 is unavailable so we cannot tell if this is a numbered or keyed
995 extraction. Accept either.
1001 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1004 JH/02 The smtp transport option "multi_domain" is now expanded.
1006 JH/03 The smtp transport now requests PRDR by default, if the server offers
1009 JH/04 Certificate name checking on server certificates, when exim is a client,
1010 is now done by default. The transport option tls_verify_cert_hostnames
1011 can be used to disable this per-host. The build option
1012 EXPERIMENTAL_CERTNAMES is withdrawn.
1014 JH/05 The value of the tls_verify_certificates smtp transport and main options
1015 default to the word "system" to access the system default CA bundle.
1016 For GnuTLS, only version 3.0.20 or later.
1018 JH/06 Verification of the server certificate for a TLS connection is now tried
1019 (but not required) by default. The verification status is now logged by
1020 default, for both outbound TLS and client-certificate supplying inbound
1023 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1026 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1027 Status Notification (bounce) messages are now MIME format per RFC 3464.
1028 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1029 under the control of the dsn_advertise_hosts option, and routers may
1030 have a dsn_lasthop option.
1032 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1033 default, modifiable by a malware= option. The list separator for
1034 the options can now be changed in the usual way. Bug 68.
1036 JH/10 The smtp_receive_timeout main option is now expanded before use.
1038 JH/11 The incoming_interface log option now also enables logging of the
1039 local interface on delivery outgoing connections.
1041 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1042 if the interface and destination host and port all match.
1044 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1047 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1048 Patch from Andrew Lewis.
1050 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1051 now supports optional time-restrictions, weighting, and priority
1052 modifiers per server. Patch originally by <rommer@active.by>.
1054 JH/16 The spamd_address main option now supports a mixed list of local
1055 and remote servers. Remote servers can be IPv6 addresses, and
1056 specify a port-range.
1058 JH/17 Bug 68: The spamd_address main option now supports an optional
1059 timeout value per server.
1061 JH/18 Bug 1581: Router and transport options headers_add/remove can
1062 now have the list separator specified.
1064 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1067 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1070 JH/21 Support for the A6 type of dns record is withdrawn.
1072 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1073 rather than the verbs used.
1075 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1076 from 255 to 1024 chars.
1078 JH/24 Verification callouts now attempt to use TLS by default.
1080 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1081 are generic router options now. The defaults didn't change.
1083 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1084 Original patch from Alexander Shikoff, worked over by JH.
1086 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1087 files if ZCAT_COMMAND is not executable.
1089 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1091 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1093 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1094 Normally benign, it bites when the pair was led to by a CNAME;
1095 modern usage is to not canonicalize the domain to a CNAME target
1096 (and we were inconsistent anyway for A-only vs AAAA+A).
1098 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1100 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1101 when evaluating $sender_host_dnssec.
1103 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1104 $sender_helo_dnssec variable.
1106 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1108 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1110 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1112 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1113 documented as working, but never had. Support all but $spam_report.
1115 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1116 added for tls authenticator.
1118 HS/03 Add perl_taintmode main config option
1124 TL/01 When running the test suite, the README says that variables such as
1125 no_msglog_check are global and can be placed anywhere in a specific
1126 test's script, however it was observed that placement needed to be near
1127 the beginning for it to behave that way. Changed the runtest perl
1128 script to read through the entire script once to detect and set these
1129 variables, reset to the beginning of the script, and then run through
1130 the script parsing/test process like normal.
1132 TL/02 The BSD's have an arc4random API. One of the functions to induce
1133 adding randomness was arc4random_stir(), but it has been removed in
1134 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1135 function when detected.
1137 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1138 cause callback expansion.
1140 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1141 syntax errors in an expansion can be treated as a string instead of
1142 logging or causing an error, due to the internal use of bool_lax
1143 instead of bool when processing it.
1145 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1146 server certificates when making smtp deliveries.
1148 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1150 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1152 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1154 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1155 Merged patch from Sebastian Wiedenroth.
1157 JH/05 Fix results-pipe from transport process. Several recipients, combined
1158 with certificate use, exposed issues where response data items split
1159 over buffer boundaries were not parsed properly. This eventually
1160 resulted in duplicates being sent. This issue only became common enough
1161 to notice due to the introduction of connection certificate information,
1162 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1164 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1165 size buffer was used, resulting in syntax errors when an expansion
1168 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1169 version 3.3.6 or later.
1171 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1172 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1173 both become "event_action", the variables become $event_name, $event_data
1174 and $event_defer_errno. There is a new variable $verify_mode, usable in
1175 routers, transports and related events. The tls:cert event is now also
1176 raised for inbound connections, if the main configuration event_action
1179 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1180 early OCSP support, but no stapling (appears to be less than 1.0.0).
1182 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1183 server certificate names available under the smtp transport option
1184 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1187 JH/10 Time-related extraction expansions from certificates now use the main
1188 option "timezone" setting for output formatting, and are consistent
1189 between OpenSSL and GnuTLS compilations. Bug 1541.
1191 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1192 encoded parameter in the incoming message. Bug 1558.
1194 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1195 include certificate info, eximon was claiming there were spoolfile
1198 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1200 JH/14 Log delivery-related information more consistently, using the sequence
1201 "H=<name> [<ip>]" wherever possible.
1203 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1204 are problematic for Debian distribution, omit them from the release
1207 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1209 JH/16 Fix string representation of time values on 64bit time_t architectures.
1212 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1213 output list separator was used.
1218 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1219 checkers that were complaining about end of non-void function with no
1222 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1223 This was a regression introduced in 4.83 by another bugfix.
1225 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1227 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1228 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1234 TF/01 Correctly close the server side of TLS when forking for delivery.
1236 When a message was received over SMTP with TLS, Exim failed to clear up
1237 the incoming connection properly after forking off the child process to
1238 deliver the message. In some situations the subsequent outgoing
1239 delivery connection happened to have the same fd number as the incoming
1240 connection previously had. Exim would try to use TLS and fail, logging
1241 a "Bad file descriptor" error.
1243 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1244 utilities have not been installed.
1246 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1247 temporary space as the ACL may create new global variables.
1249 TL/01 LDAP support uses per connection or global context settings, depending
1250 upon the detected version of the libraries at build time.
1252 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1253 to extract and use the src ip:port in logging and expansions as if it
1254 were a direct connection from the outside internet. PPv2 support was
1255 updated based on HAProxy spec change in May 2014.
1257 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1259 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1260 Properly escape header and check for NULL return.
1262 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1265 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1267 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1268 characters in header names, implemented as a verify condition.
1269 Contributed by Michael Fischer v. Mollard.
1271 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1272 results permerror and temperror. Previous values are deprecated but
1273 still accepted. In a future release, err_perm and err_temp will be
1274 completely removed, which will be a backward incompatibility if the
1275 ACL tests for either of these two old results. Patch contributed by
1276 user bes-internal on the mailing list.
1278 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1280 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1281 selectors, in both main and reject logs.
1283 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1286 JH/07 Add malware type "sock" for talking to simple daemon.
1288 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1290 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1291 routers/transports under cutthrough routing.
1293 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1294 numbers. Touch up "bool" conditional to keep the same definition.
1296 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1298 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1300 JH/12 Expand items in router/transport headers_add or headers_remove lists
1301 individually rather than the list as a whole. Bug 1452.
1303 Required for reasonable handling of multiple headers_ options when
1304 they may be empty; requires that headers_remove items with embedded
1305 colons must have them doubled (or the list-separator changed).
1307 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1308 view the policy declared in the DMARC record. Currently, $dmarc_status
1309 is a combined value of both the record presence and the result of the
1312 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1314 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1315 dnslookup router and the smtp transport (applying to the forward
1318 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1319 of ldap servers used for a specific lookup. Patch provided by Heiko
1322 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1323 New variable $lookup_dnssec_authenticated for observability.
1325 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1326 Patch submitted by Lars Timman.
1328 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1330 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1331 Requires trusted mode and valid format message id, aborts otherwise.
1332 Patch contributed by Heiko Schlichting.
1334 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1335 certextract with support for various fields. Bug 1358.
1337 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1338 is requested by default, modifiable by smtp transport option
1341 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1342 operate on certificate variables to give certificate fingerprints
1343 Also new ${sha256:cert_variable}.
1345 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1347 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1348 Christian Aistleitner.
1350 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1352 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1353 file. Patch from Wolfgang Breyha.
1355 JH/25 Expand the coverage of the delivery $host and $host_address to
1356 client authenticators run in verify callout. Bug 1476.
1358 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1359 align with daemon_smtp_ports. Bug 72.
1361 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1362 support and error reporting did not work properly.
1364 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1365 and is readable. Patch from Andrew Colin Kissa.
1367 TL/14 Enhance documentation of ${run expansion and how it parses the
1368 commandline after expansion, particularly in the case when an
1369 unquoted variable expansion results in an empty value.
1371 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1373 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1374 by renaming away from it. Addresses GH issue 15, reported by
1377 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1379 TL/15 SECURITY: prevent double expansion in math comparison functions
1380 (can expand unsanitized data). Not remotely exploitable.
1387 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1389 PP/02 Make -n do something, by making it not do something.
1390 When combined with -bP, the name of an option is not output.
1392 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1395 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1396 $sender_host_name and config options to manage this, and basic check
1399 PP/05 DSCP support for outbound connections and control modifier for inbound.
1401 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1402 (Only plugin which currently uses this is kerberos4, which nobody should
1403 be using, but we should make it available and other future plugins might
1404 conceivably use it, even though it would break NAT; stuff *should* be
1405 using channel bindings instead).
1407 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1408 name; added for Sendmail compatibility; requires admin caller.
1409 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1410 just ignore it); requires trusted caller.
1411 Also parse but ignore: -Ac -Am -X<logfile>
1414 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1416 TL/02 Add +smtp_confirmation as a default logging option.
1418 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1419 Patch by Magnus Holmgren from 2007-02-20.
1421 TL/04 Bugzilla 1281 - Spec typo.
1422 Bugzilla 1283 - Spec typo.
1423 Bugzilla 1290 - Spec grammar fixes.
1425 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1427 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1429 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1430 mailing list by by Dmitry Isaikin.
1432 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1434 JH/02 Support "G" suffix to numbers in ${if comparisons.
1436 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1438 NM/01 Bugzilla 1197 - Spec typo
1439 Bugzilla 1196 - Spec examples corrections
1441 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1443 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1444 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1447 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1448 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1450 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1451 "acl {{name}{arg}...}", and optional args on acl condition
1454 JH/05 Permit multiple router/transport headers_add/remove lines.
1456 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1458 JH/07 Avoid using a waiting database for a single-message-only transport.
1459 Performance patch from Paul Fisher. Bugzilla 1262.
1461 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1464 JH/09 Add $headers_added variable, with content from use of ACL modifier
1465 add_header (but not yet added to the message). Bugzilla 199.
1467 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1468 Pulled from Bugzilla 817 by Wolfgang Breyha.
1470 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1472 (nb: this is the same fix as in Exim 4.80.1)
1474 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1477 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1480 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1482 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1483 advertise SMTP AUTH mechanism to us, instead of a generic
1484 protocol violation error. Also, make Exim more robust to bad
1485 data from the Dovecot auth socket.
1487 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1489 When a queue runner is handling a message, Exim first routes the
1490 recipient addresses, during which it prunes them based on the retry
1491 hints database. After that it attempts to deliver the message to
1492 any remaining recipients. It then updates the hints database using
1495 So if a recipient address works intermittently, it can get repeatedly
1496 deferred at routing time. The retry hints record remains fresh so the
1497 address never reaches the final cutoff time.
1499 This is a fairly common occurrence when a user is bumping up against
1500 their storage quota. Exim had some logic in its local delivery code
1501 to deal with this. However it did not apply to per-recipient defers
1502 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1504 This change adds a proper retry rule check during routing so that the
1505 final cutoff time is checked against the message's age. We only do
1506 this check if there is an address retry record and there is not a
1507 domain retry record; this implies that previous attempts to handle
1508 the address had the retry_use_local_parts option turned on. We use
1509 this as an approximation for the destination being like a local
1510 delivery, as in LMTP.
1512 I suspect this new check makes the old local delivery cutoff check
1513 redundant, but I have not verified this so I left the code in place.
1515 TF/02 Correct gecos expansion when From: is a prefix of the username.
1517 Test 0254 submits a message to Exim with the header
1521 When I ran the test suite under the user fanf2, Exim expanded
1522 the header to contain my full name, whereas it should have added
1523 a Resent-Sender: header. It erroneously treats any prefix of the
1524 username as equal to the username.
1526 This change corrects that bug.
1528 GF/01 DCC debug and logging tidyup
1529 Error conditions log to paniclog rather than rejectlog.
1530 Debug lines prefixed by "DCC: " to remove any ambiguity.
1532 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1534 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1535 Bug spotted by Jeremy Harris; was flawed since initial commit.
1536 Would have resulted in OCSP responses post-SNI triggering an Exim
1537 NULL dereference and crash.
1539 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1541 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1542 Bug detection, analysis and fix by Samuel Thibault.
1543 Bugzilla 1331, Debian bug #698092.
1545 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1547 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1548 Server implementation by Todd Lyons, client by JH.
1549 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1550 config variable "prdr_enable" controls whether the server
1551 advertises the facility. If the client requests PRDR a new
1552 acl_data_smtp_prdr ACL is called once for each recipient, after
1553 the body content is received and before the acl_smtp_data ACL.
1554 The client is controlled by both of: a hosts_try_prdr option
1555 on the smtp transport, and the server advertisement.
1556 Default client logging of deliveries and rejections involving
1557 PRDR are flagged with the string "PRDR".
1559 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1560 fclose(). Diagnosis by Todd Lyons.
1562 PP/17 Update configure.default to handle IPv6 localhost better.
1563 Patch by Alain Williams (plus minor tweaks).
1566 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1567 This is now consistent with GnuTLS, and is now documented: the
1568 previous undocumented portable approach to treating the option as
1569 unset was to force an expansion failure. That still works, and
1570 an empty string is now equivalent.
1572 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1573 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1574 not performing validation itself.
1576 PP/20 Added force_command boolean option to pipe transport.
1577 Patch from Nick Koston, of cPanel Inc.
1579 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1582 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1584 PP/21 Fix eximon continuous updating with timestamped log-files.
1585 Broken in a format-string cleanup in 4.80, missed when I repaired the
1586 other false fix of the same issue.
1587 Report and fix from Heiko Schlichting.
1590 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1591 Report from Prashanth Katuri.
1593 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1594 It's SecureTransport, so affects any MacOS clients which use the
1595 system-integrated TLS libraries, including email clients.
1597 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1598 using a MIME ACL for non-SMTP local injection.
1599 Report and assistance in diagnosis by Warren Baker.
1601 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1603 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1605 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1606 last id that failed so it may be referenced in subsequent ACL's.
1608 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1611 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1612 ldap library initialization, allowing self-signed CA's to be
1613 used. Also properly sets require_cert option later in code by
1614 using NULL (global ldap config) instead of ldap handle (per
1615 session). Bug diagnosis and testing by alxgomz.
1617 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1618 the src/util/ subdirectory.
1620 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1621 renamed to Transport Post Delivery Action by Jeremy Harris, as
1624 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1625 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1626 redis_servers = needs to be configured which will be used by the redis
1627 lookup. Patch from Warren Baker, of The Packet Hub.
1629 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1631 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1632 hostname or reverse DNS when processing a host list. Used suggestions
1633 from multiple comments on this bug.
1635 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1637 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1638 Missed a few lines, added it to make the runtest require no keyboard
1641 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1642 contains upper case chars. Make router use caseful_local_part.
1644 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1645 support when GnuTLS has been built with p11-kit.
1651 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1653 This, or similar/improved, will also be change PP/11 of 4.82.
1659 PP/01 Handle short writes when writing local log-files.
1660 In practice, only affects FreeBSD (8 onwards).
1661 Bugzilla 1053, with thanks to Dmitry Isaikin.
1663 NM/01 Bugzilla 949 - Documentation tweak
1665 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1668 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1670 PP/02 Implemented gsasl authenticator.
1672 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1674 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1675 `pkg-config foo` for cflags/libs.
1677 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1678 with rest of GSASL and with heimdal_gssapi.
1680 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1681 `pkg-config foo` for cflags/libs for the TLS implementation.
1683 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1684 properties get this fed in as external SSF. A number of robustness
1685 and debugging improvements to the cyrus_sasl authenticator.
1687 PP/08 cyrus_sasl server now expands the server_realm option.
1689 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1690 Patch by Jeremy Harris.
1692 PP/10 Added dbmjz lookup type.
1694 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1696 PP/12 MAIL args handles TAB as well as SP, for better interop with
1697 non-compliant senders.
1698 Analysis and variant patch by Todd Lyons.
1700 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1701 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1702 Patch from Dirk Mueller <dmueller@suse.com>
1704 PP/13 tls_peerdn now print-escaped for spool files.
1705 Observed some $tls_peerdn in wild which contained \n, which resulted
1706 in spool file corruption.
1708 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1709 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1710 or write after TLS renegotiation, which otherwise led to messages
1713 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1714 as a tracking header (ie: a signed header comes before the signature).
1715 Patch from Wolfgang Breyha.
1717 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1718 comma-sep list; embedded commas doubled.
1720 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1722 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1724 Report and patch from Dmitry Banschikov.
1726 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1727 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1728 failures. We appear to now support TLS1.1+ with Exim.
1730 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1731 lets Exim select keys and certificates based upon TLS SNI from client.
1732 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1733 before an outbound SMTP session. New log_selector, +tls_sni.
1735 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1736 NULL dereference. Report and patch from Alun Jones.
1738 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1739 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1740 Not seeing resolver debug output on NetBSD, but suspect this is a
1741 resolver implementation change.
1743 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1744 Left warnings. Added "eximon gdb" invocation mode.
1746 PP/21 Defaulting "accept_8bitmime" to true, not false.
1748 PP/22 Added -bw for inetd wait mode support.
1750 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1751 locate the relevant includes and libraries. Made this the default.
1753 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1754 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1756 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1757 This may cause build issues on older platforms.
1759 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1760 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1761 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1762 Added SNI support via GnuTLS too.
1763 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1765 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1767 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1768 Applied second patch from Janne, implementing suggestion to default
1769 multiple-strings-in-record handling to match SPF spec.
1771 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1773 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1774 read-only, out of scope).
1775 Patch from Wolfgang Breyha, report from Stuart Northfield.
1777 PP/29 Fix three issues highlighted by clang analyser static analysis.
1778 Only crash-plausible issue would require the Cambridge-specific
1779 iplookup router and a misconfiguration.
1780 Report from Marcin Mirosław.
1782 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1784 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1785 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1786 As part of this, removing so much warning spew let me fix some minor
1787 real issues in debug logging.
1789 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1790 assignment on my part. Fixed.
1792 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1793 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1794 Janne Snabb (who went above and beyond: thank you).
1796 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1797 string otherwise requires a connection and a bunch more work and it's
1798 relatively easy to get wrong. Should also expose TLS library linkage
1801 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1802 64-bit ${eval} (JH/03).
1804 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1805 GNU libc to support some of the 64-bit stuff, should not lead to
1806 conflicts. Defined before os.h is pulled in, so if a given platform
1807 needs to override this, it can.
1809 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1810 protection layer was required, which is not implemented.
1811 Bugzilla 1254, patch from Wolfgang Breyha.
1813 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1814 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1815 tls_dhparam take prime identifiers. Also unbreak combination of
1816 OpenSSL+DH_params+TLSSNI.
1818 PP/39 Disable SSLv2 by default in OpenSSL support.
1824 PP/01 Solaris build fix for Oracle's LDAP libraries.
1825 Bugzilla 1109, patch from Stephen Usher.
1827 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1829 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1832 TF/02 Fix a couple more cases where we did not log the error message
1833 when unlink() failed. See also change 4.74-TF/03.
1835 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1836 lock up or crash if it happened to be inside a call to libc when it
1837 got a SIGUSR1 from exiwhat.
1839 The SIGUSR1 handler appends the current process status to the process
1840 log which is later printed by exiwhat. It used to use the general
1841 purpose logging code to do this, but several functions it calls are
1842 not safe for signals.
1844 The new output code in the SIGUSR1 handler is specific to the process
1845 log, and simple enough that it's easy to inspect for signal safety.
1846 Removing some special cases also simplifies the general logging code.
1847 Removing the spurious timestamps from the process log simplifies
1850 TF/04 Improved ratelimit ACL condition.
1852 The /noupdate option has been deprecated in favour of /readonly which
1853 has clearer semantics. The /leaky, /strict, and /readonly update modes
1854 are mutually exclusive. The update mode is no longer included in the
1855 database key; it just determines when the database is updated. (This
1856 means that when you upgrade Exim will forget old rate measurements.)
1858 Exim now checks that the per_* options are used with an update mode that
1859 makes sense for the current ACL. For example, when Exim is processing a
1860 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1861 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1862 must specify per_mail/readonly. If you omit the update mode it defaults to
1863 /leaky where that makes sense (as before) or /readonly where required.
1865 The /noupdate option is now undocumented but still supported for
1866 backwards compatibility. It is equivalent to /readonly except that in
1867 ACLs where /readonly is required you may specify /leaky/noupdate or
1868 /strict/noupdate which are treated the same as /readonly.
1870 A useful new feature is the /count= option. This is a generalization
1871 of the per_byte option, so that you can measure the throughput of other
1872 aggregate values. For example, the per_byte option is now equivalent
1873 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1875 The per_rcpt option has been generalized using the /count= mechanism
1876 (though it's more complicated than the per_byte equivalence). When it is
1877 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1878 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1879 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1880 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1881 Note that using per_rcpt with a non-readonly update mode in more than
1882 one ACL will cause the recipients to be double-counted. (The per_mail
1883 and per_byte options don't have this problem.)
1885 The handling of very low rates has changed slightly. If the computed rate
1886 is less than the event's count (usually one) then this event is the first
1887 after a long gap. In this case the rate is set to the same as this event's
1888 count, so that the first message of a spam run is counted properly.
1890 The major new feature is a mechanism for counting the rate of unique
1891 events. The new per_addr option counts the number of different
1892 recipients that someone has sent messages to in the last time period. It
1893 behaves like per_rcpt if all the recipient addresses are different, but
1894 duplicate recipient addresses do not increase the measured rate. Like
1895 the /count= option this is a general mechanism, so the per_addr option
1896 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1897 example, measure the rate that a client uses different sender addresses
1898 with the options per_mail/unique=$sender_address. There are further
1899 details in the main documentation.
1901 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1903 TF/06 Removed a few PCRE remnants.
1905 TF/07 Automatically extract Exim's version number from tags in the git
1906 repository when doing development or release builds.
1908 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1909 Bugzilla 879. Patch from Paul Fisher.
1911 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1912 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1915 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1917 PP/05 Variable $av_failed, true if the AV scanner deferred.
1918 Bugzilla 1078. Patch from John Horne.
1920 PP/06 Stop make process more reliably on build failure.
1921 Bugzilla 1087. Patch from Heiko Schlittermann.
1923 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1924 Bugzilla 1089. Patch from Heiko Schlittermann.
1926 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1927 Bugzilla 1131. Patch from Holger Weiß.
1929 PP/09 Handle IPv6 addresses with SPF.
1930 Bugzilla 860. Patch from Wolfgang Breyha.
1932 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1934 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1937 PP/11 match_* no longer expand right-hand-side by default.
1938 New compile-time build option, EXPAND_LISTMATCH_RHS.
1939 New expansion conditions, "inlist", "inlisti".
1941 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1943 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1945 PP/14 fix log_write() format string regression from TF/03.
1946 Bugzilla 1152. Patch from Dmitry Isaikin.
1952 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1954 PP/02 Harmonised TLS library version reporting; only show if debugging.
1955 Layout now matches that introduced for other libraries in 4.74 PP/03.
1957 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1959 PP/04 New "dns_use_edns0" global option.
1961 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1964 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1965 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1967 TK/01 Updated PolarSSL code to 0.14.2.
1968 Bugzilla 1097. Patch from Andreas Metzler.
1970 PP/07 Catch divide-by-zero in ${eval:...}.
1971 Fixes bugzilla 1102.
1973 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1976 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1977 format-string attack -- SECURITY: remote arbitrary code execution.
1979 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1980 time unintentionally subject to list matching rules, letting the header
1981 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1982 arbitrary string expansion). This allowed for information disclosure.
1984 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1985 INT_MIN/-1 -- value coerced to INT_MAX.
1991 NM/01 Workaround for PCRE version dependency in version reporting
1994 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1995 This fixes portability to compilers other than gcc, notably
1996 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1998 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1999 makefiles for portability to HP-UX and POSIX correctness.
2001 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2002 Also via indented variable definition in the Makefile.
2003 (Debugging by Oliver Heesakkers).
2005 PP/02 Restore caching of spamd results with expanded spamd_address.
2006 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2008 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2009 Improves build reliability. Fix from: Frank Elsner
2011 NM/02 Fix wide character breakage in the rfc2047 coding
2012 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2014 NM/03 Allow underscore in dnslist lookups
2015 Fixes bug 1026. Patch from Graeme Fowler
2017 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2018 Code patches from Adam Ciarcinski of NetBSD.
2020 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2023 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2024 is logged, to avoid truncation. Patch from John Horne.
2026 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2027 Patch from Jakob Hirsch.
2029 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2030 SQL string expansion failure details.
2031 Patch from Andrey Oktyabrski.
2033 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2034 Patch from Simon Arlott.
2036 PP/09 New lookups functionality failed to compile on old gcc which rejects
2037 extern declarations in function scope.
2038 Patch from Oliver Fleischmann
2040 PP/10 Use sig_atomic_t for flags set from signal handlers.
2041 Check getgroups() return and improve debugging.
2042 Fixed developed for diagnosis in bug 927 (which turned out to be
2045 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2046 Patch from Mark Zealey.
2048 PP/12 Bugzilla 1056: Improved spamd server selection.
2049 Patch from Mark Zealey.
2051 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2052 Based on patch from Heiko Schlittermann.
2054 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2055 Patch from Uwe Doering, sign-off by Michael Haardt.
2057 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2058 variable declaration deep within a block. Bug and patch from
2061 PP/15 lookups-Makefile IRIX compatibility coercion.
2063 PP/16 Make DISABLE_DKIM build knob functional.
2065 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2066 Patch by Simon Arlott
2068 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2069 variable argument macros. Our copy now differs from upstream.
2075 TF/01 Failure to get a lock on a hints database can have serious
2076 consequences so log it to the panic log.
2078 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2079 controlled using the smtp_confirmation log selector.
2081 TF/03 Include the error message when we fail to unlink a spool file.
2083 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2084 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2085 for maintaining out-of-tree patches for some time.
2087 PP/01 Bugzilla 139: Documentation and portability issues.
2088 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2089 Handle per-OS dynamic-module compilation flags.
2091 PP/02 Let /dev/null have normal permissions.
2092 The 4.73 fixes were a little too stringent and complained about the
2093 permissions on /dev/null. Exempt it from some checks.
2094 Reported by Andreas M. Kirchwitz.
2096 PP/03 Report version information for many libraries, including
2097 Exim version information for dynamically loaded libraries. Created
2098 version.h, now support a version extension string for distributors
2099 who patch heavily. Dynamic module ABI change.
2101 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2102 privilege escalation vulnerability whereby the Exim run-time user
2103 can cause root to append content of the attacker's choosing to
2106 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2109 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2110 If dropping privileges for untrusted macros, we disabled normal logging
2111 on the basis that it would fail; for the Exim run-time user, this is not
2112 the case, and it resulted in successful deliveries going unlogged.
2113 Fixed. Reported by Andreas Metzler.
2119 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2120 only prepend for the Resent-* case. Fixes regression introduced in
2121 Exim 4.70 by NM/22 for Bugzilla 607.
2123 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2124 increasing numbers of administrators be bitten by this.
2126 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2128 PP/03 Bugzilla 994: added openssl_options main configuration option.
2130 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2132 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2134 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2136 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2137 without a peer certificate, leading to a segfault because of an
2138 assumption that peers always have certificates. Be a little more
2139 paranoid. Problem reported by Martin Tscholak.
2141 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2142 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2143 NB: ClamAV planning to remove STREAM in "middle of 2010".
2144 CL also introduces -bmalware, various -d+acl logging additions and
2145 more caution in buffer sizes.
2147 PP/09 Implemented reverse_ip expansion operator.
2149 PP/10 Bugzilla 937: provide a "debug" ACL control.
2151 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2153 PP/12 Bugzilla 973: Implement --version.
2155 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2157 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2159 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2161 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2162 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2163 ignore trailing whitespace.
2165 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2167 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2170 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2171 Notification from Dr Andrew Aitchison.
2173 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2174 ExtendedDetectionInfo response format.
2175 Notification from John Horne.
2177 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2180 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2181 XSL and documented dependency on system catalogs, with examples of how
2184 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2187 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2188 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2189 configuration file which is writeable by the Exim user or group.
2191 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2192 of configuration files to cover files specified with the -C option if
2193 they are going to be used with root privileges, not just the default
2196 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2197 option (effectively making it always true).
2199 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2200 files to be used while preserving root privileges.
2202 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2203 that rogue child processes cannot use them.
2205 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2206 run-time user, instead of root.
2208 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2209 Exim run-time user without dropping privileges.
2211 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2212 result string, instead of calling string_vformat() twice with the same
2215 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2216 for other users. Others should always drop root privileges if they use
2217 -C on the command line, even for a whitelisted configure file.
2219 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2221 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2227 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2228 $data_path, and $header_path variables; fixed documentation bugs and
2231 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2232 exipick to access non-standard spools, including the "frozen" queue
2235 NM/01 Bugzilla 965: Support mysql stored procedures.
2236 Patch from Alain Williams
2238 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2240 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2241 Patch from Andreas Metzler
2243 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2244 Patch from Kirill Miazine
2246 NM/05 Bugzilla 671: Added umask to procmail example.
2248 JJ/03 installed exipick 20100323.0, fixing doc bug
2250 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2251 directory. Notification and patch from Dan Rosenberg.
2253 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2255 TK/02 Improve log output when DKIM signing operation fails.
2257 MH/01 Treat the transport option dkim_domain as a colon separated
2258 list, not as a single string, and sign the message with each element,
2259 omitting multiple occurences of the same signer.
2261 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2262 Bugzilla 985, 986. Patch by Simon Arlott
2264 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2265 Patch by Simon Arlott
2267 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2268 MBX locking. Notification from Dan Rosenberg.
2274 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2276 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2278 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2280 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2282 NM/04 Bugzilla 915: Flag broken perl installation during build.
2288 TK/01 Added patch by Johannes Berg that expands the main option
2289 "spamd_address" if it starts with a dollar sign.
2291 TK/02 Write list of recipients to X-Envelope-Sender header when building
2292 the mbox-format spool file for content scanning (suggested by Jakob
2295 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2296 (http://www.dcc-servers.net/) support via dccifd. Activated by
2297 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2299 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2300 by Mark Daniel Reidel <mr@df.eu>.
2302 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2303 When building exim an external PCRE library is now needed -
2304 PCRE is a system library on the majority of modern systems.
2305 See entry on PCRE_LIBS in EDITME file.
2307 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2308 conversation. Added nologin parameter to request.
2309 Patch contributed by Kirill Miazine.
2311 TF/01 Do not log submission mode rewrites if they do not change the address.
2313 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2315 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2316 log files in place. Contributed by Roberto Lima.
2318 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2320 TF/03 Bugzilla 615: When checking the local_parts router precondition
2321 after a local_part_suffix or local_part_prefix option, Exim now
2322 does not use the address's named list lookup cache, since this
2323 contains cached lookups for the whole local part.
2325 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2326 Robert Millan. Documentation is in experimental-spec.txt.
2328 TF/04 Bugzilla 668: Fix parallel build (make -j).
2330 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2332 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2333 Patch provided by Jan Srzednicki.
2335 TF/05 Leading white space used to be stripped from $spam_report which
2336 wrecked the formatting. Now it is preserved.
2338 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2339 that they are available at delivery time.
2341 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2343 TF/08 TLS error reporting now respects the incoming_interface and
2344 incoming_port log selectors.
2346 TF/09 Produce a more useful error message if an SMTP transport's hosts
2347 setting expands to an empty string.
2349 NM/06 Bugzilla 744: EXPN did not work under TLS.
2350 Patch provided by Phil Pennock.
2352 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2353 Patch provided by Richard Godbee.
2355 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2356 acl_smtp_notquit, added index entry.
2358 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2359 Patch provided by Eugene Bujak.
2361 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2362 accept(). Patch provided by Maxim Dounin.
2364 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2365 Patch provided by Phil Pennock.
2367 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2369 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2370 Patch provided by Brad "anomie" Jorsch.
2372 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2373 Patch provided by Dean Brooks.
2375 TK/05 Add native DKIM support (does not depend on external libraries).
2377 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2378 Patch provided by Graeme Fowler.
2380 NM/16 Bugzilla 851: Documentation example syntax fix.
2382 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2384 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2387 NM/19 Bugzilla 745: TLS version reporting.
2388 Patch provided by Phil Pennock.
2390 NM/20 Bugzilla 167: bool: condition support.
2391 Patch provided by Phil Pennock.
2393 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2394 clients. Patch provided by Phil Pennock.
2396 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2397 Patch provided by Brad "anomie" Jorsch.
2399 NM/23 Bugzilla 687: Fix misparses in eximstats.
2400 Patch provided by Heiko Schlittermann.
2402 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2403 Patch provided by Heiko Schlittermann.
2405 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2406 plus update to original patch.
2408 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2410 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2411 Patch provided by David Brownlee.
2413 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2415 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2417 NM/30 Bugzilla 888: TLS documentation bugfixes.
2419 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2421 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2422 Unlike the original bugzilla I have changed all shell scripts in src tree.
2424 NM/33 Bugzilla 898: Transport filter timeout fix.
2425 Patch by Todd Rinaldo.
2427 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2428 Patch by Serge Demonchaux.
2430 NM/35 Bugzilla 39: Base64 decode bug fixes.
2431 Patch by Jakob Hirsch.
2433 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2435 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2437 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2439 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2445 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2446 ALT-N's libdkim that I have put here:
2447 http://duncanthrax.net/exim-experimental/
2449 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2450 were called 'true' and it seems that C99 defines that as a reserved
2451 keyword to be used with 'bool' variable types. That means you could
2452 not include C99-style headers which use bools without triggering
2453 build errors in sieve.c.
2455 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2456 as mailq or other aliases. Changed the --help handling significantly
2457 to do whats expected. exim_usage() emits usage/help information.
2459 SC/01 Added the -bylocaldomain option to eximstats.
2461 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2463 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2465 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2471 PH/01 Another patch from the Sieve maintainer.
2473 PH/02 When an IPv6 address is converted to a string for single-key lookup
2474 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2475 dots are used instead of colons so that keys in lsearch files need not
2476 contain colons. This was done some time before quoting was made available
2477 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2478 (notated using the quote facility) so as to distinguish them from IPv4
2479 keys. This meant that lookups for IP addresses in host lists did not work
2480 for iplsearch lookups.
2482 This has been fixed by arranging for IPv6 addresses to be expressed with
2483 colons if the lookup type is iplsearch. This is not incompatible, because
2484 previously such lookups could never work.
2486 The situation is now rather anomalous, since one *can* have colons in
2487 ordinary lsearch keys. However, making the change in all cases is
2488 incompatible and would probably break a number of configurations.
2490 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2493 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2494 conversion specification without a maximum field width, thereby enabling
2495 a rogue spamd server to cause a buffer overflow. While nobody in their
2496 right mind would setup Exim to query an untrusted spamd server, an
2497 attacker that gains access to a server running spamd could potentially
2498 exploit this vulnerability to run arbitrary code as the Exim user.
2500 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2501 $primary_hostname instead of what libspf2 thinks the hosts name is.
2503 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2504 a directory entry by the name of the lookup key. Previously, if a
2505 symlink pointed to a non-existing file or a file in a directory that
2506 Exim lacked permissions to read, a lookup for a key matching that
2507 symlink would fail. Now it is enough that a matching directory entry
2508 exists, symlink or not. (Bugzilla 503.)
2510 PH/03 The body_linecount and body_zerocount variables are now exported in the
2513 PH/04 Added the $dnslist_matched variable.
2515 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2516 This means they are set thereafter only if the connection becomes
2519 PH/06 Added the client_condition to authenticators so that some can be skipped
2520 by clients under certain conditions.
2522 PH/07 The error message for a badly-placed control=no_multiline_responses left
2523 "_responses" off the end of the name.
2525 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2527 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2528 (without spaces) instead of just copying the configuration text.
2530 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2532 PH/11 Added $max_received_linelength.
2534 PH/12 Added +ignore_defer and +include_defer to host lists.
2536 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2537 way in which PCRE > 7.0 is built.
2539 PH/14 Implemented queue_only_load_latch.
2541 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2542 MAIL command. The effect was to mangle the value on 64-bit systems.
2544 PH/16 Another patch from the Sieve maintainer.
2546 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2548 PH/18 If a system quota error occurred while trying to create the file for
2549 a maildir delivery, the message "Mailbox is full" was not appended to the
2550 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2551 only to a quota excession during the actual writing of the file.
2553 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2554 characters?) which causes problems in log lines. The DN values are now
2555 passed through string_printing() before being added to log lines.
2557 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2558 and InterBase are left for another time.)
2560 PH/21 Added message_body_newlines option.
2562 PH/22 Guard against possible overflow in moan_check_errorcopy().
2564 PH/23 POSIX allows open() to be a macro; guard against that.
2566 PH/24 If the recipient of an error message contained an @ in the local part
2567 (suitably quoted, of course), incorrect values were put in $domain and
2568 $local_part during the evaluation of errors_copy.
2574 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2575 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2578 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2579 issue a MAIL command.
2581 PH/02 In an ACL statement such as
2583 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2585 if a client was not listed at all, or was listed with a value other than
2586 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2587 the condition was not true (as it should be), so access was not denied.
2588 The bug was that the ! inversion was incorrectly passed on to the second
2589 item. This has been fixed.
2591 PH/03 Added additional dnslists conditions == and =& which are different from
2592 = and & when the dns lookup returns more than one IP address.
2594 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2595 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2597 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2598 FSYNC, which compiles an option called disable_fsync that allows for
2599 bypassing fsync(). The documentation is heavily laced with warnings.
2601 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2603 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2604 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2605 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2606 including adding "make clean"; (3) Added -fPIC when compiling the test
2607 dynamically loaded module, to get rid of a warning.
2609 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2610 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2611 = 0s. The bug is otherwise harmless.
2613 PH/07 There was a bug in the dovecot authenticator such that the value of
2614 $auth1 could be overwritten, and so not correctly preserved, after a
2615 successful authentication. This usually meant that the value preserved by
2616 the server_setid option was incorrect.
2618 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2620 PH/09 Installed PCRE release 7.0.
2622 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2623 run for batched SMTP input. It is now run at the start of every message
2624 in the batch. While fixing this I discovered that the process information
2625 (output by running exiwhat) was not always getting set for -bs and -bS
2626 input. This is fixed, and it now also says "batched" for BSMTP.
2628 PH/11 Added control=no_pipelining.
2630 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2631 patch, slightly modified), and move the expansion of helo_data till after
2632 the connection is made in the smtp transport (so it can use these
2635 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2637 PH/14 Added log_selector = +pid.
2639 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2641 PH/16 Add ${if forany and ${if forall.
2643 PH/17 Added dsn_from option to vary the From: line in DSNs.
2645 PH/18 Flush SMTP output before performing a callout, unless control =
2646 no_callout_flush is set.
2648 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2649 was true (the default) a successful delivery failed to delete the retry
2650 item, thus causing premature timeout of the address. The bug is now
2653 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2655 PH/21 Long custom messages for fakedefer and fakereject are now split up
2656 into multiline responses in the same way that messages for "deny" and
2657 other ACL rejections are.
2659 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2660 with slight modification.
2662 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2663 draft, changing the syntax and factoring some duplicate code".
2665 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2666 for deliveries of the second and subsequent messages over the same SMTP
2669 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2670 ${reduce, with only minor "tidies".
2672 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2674 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2675 expansion side effects.
2677 PH/27 When a message times out after an over-quota error from an Exim-imposed
2678 quota, the bounce message says "mailbox is full". This message was not
2679 being given when it was a system quota that was exceeded. It now should
2682 MH/03 Made $recipients available in local_scan(). local_scan() already has
2683 better access to the recipient list through recipients_list[], but
2684 $recipients can be useful in postmaster-provided expansion strings.
2686 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2687 in the case of a MAIL command with additional options following the
2688 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2689 were accidentally chopped off.
2691 PH/29 SMTP synchronization checks are implemented when a command is read -
2692 there is a check that no more input is waiting when there shouldn't be
2693 any. However, for some commands, a delay in an ACL can mean that it is
2694 some time before the response is written. In this time, more input might
2695 arrive, invalidly. So now there are extra checks after an ACL has run for
2696 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2697 pipelining has not been advertised.
2699 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2701 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2702 returned in the SMTP response when smtp_return_error_details was false.
2703 This has been fixed.
2705 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2706 instead of the C I/O that was originally supplied, because problems were
2707 reported on Solaris.
2709 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2710 Exim which did not show up earlier: it was assuming that a call to
2711 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2712 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2713 was a macro that became an assignment, so it seemed to work. This has
2714 changed to a proper function call with a void return, hence the compile
2715 error. Exim's code has been fixed.
2717 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2720 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2722 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2724 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2725 because it turns out to be too big or there is a timeout) there is no
2726 "Completed" line in the log. When some messages of this type were
2727 selected by exigrep, they were listed as "not completed". Others were
2728 picked up by some special patterns. I have improved the selection
2729 criteria to be more general.
2731 PH/38 The host_find_failed option in the manualroute router can now be set
2732 to "ignore", to completely ignore a host whose IP address cannot be
2733 found. If all hosts are ignored, the behaviour is controlled by the new
2734 host_all_ignored option.
2736 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2737 homing or because of multiple MX records with /mx) generated more than
2738 one IP address, and the following item turned out to be the local host,
2739 all the secondary addresses of the first item were incorrectly removed
2740 from the list, along with the local host and any following hosts (which
2741 is what is supposed to happen).
2743 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2744 whoever called Exim into the -H file. In the case of the daemon it was
2745 behaving confusingly. When first started, it used values for whoever
2746 started the daemon, but after a SIGHUP it used the Exim user (because it
2747 calls itself on a restart). I have changed the code so that it now always
2750 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2751 message are rejected with the same error (e.g. no authentication or bad
2752 sender address), and a DATA command is nevertheless sent (as can happen
2753 with PIPELINING or a stupid MUA), the error message that was given to the
2754 RCPT commands is included in the rejection of the DATA command. This is
2755 intended to be helpful for MUAs that show only the final error to their
2758 PH/42 Another patch from the Sieve maintainer.
2760 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2761 Eximstats - Fixed some broken HTML links and added missing column headers
2763 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2764 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2766 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2768 PH/43 Yet another patch from the Sieve maintainer.
2770 PH/44 I found a way to check for a TCP/IP connection going away before sending
2771 the response to the final '.' that terminates a message, but only in the
2772 case where the client has not sent further data following the '.'
2773 (unfortunately, this is allowed). However, in many cases there won't be
2774 any further data because there won't be any more messages to send. A call
2775 to select() can be used: if it shows that the input is "ready", there is
2776 either input waiting, or the socket has been closed. An attempt to read
2777 the next input character can distinguish the two cases. Previously, Exim
2778 would have sent an OK response which the client would never have see.
2779 This could lead to message repetition. This fix should cure that, at
2780 least in a lot of common cases.
2782 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2783 advertised in response to EHLO.
2789 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2790 fixed by 4.65/MH/01 (is this a record?) are fixed:
2792 (i) An empty string was always treated as zero by the numeric comparison
2793 operators. This behaviour has been restored.
2795 (ii) It is documented that the numeric comparison operators always treat
2796 their arguments as decimal numbers. This was broken in that numbers
2797 starting with 0 were being interpreted as octal.
2799 While fixing these problems I realized that there was another issue that
2800 hadn't been noticed. Values of message_size_limit (both the global option
2801 and the transport option) were treated as octal if they started with 0.
2802 The documentation was vague. These values are now always treated as
2803 decimal, and I will make that clear in the documentation.
2809 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2810 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2813 MH/01 Don't check that the operands of numeric comparison operators are
2814 integers when their expansion is in "skipping" mode (fixes bug
2815 introduced by 4.64-PH/07).
2817 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2818 child addresses, Exim now panics and dies. Previously, because the count
2819 is held in a short int, deliveries were likely to be lost. As such a
2820 large number of recipients for a single message is ridiculous
2821 (performance will be very, very poor), I have chosen to impose a limit
2822 rather than extend the field.
2828 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2829 leftover -K file (the existence of which was triggered by #402).
2830 While we were at it, introduced process PID as part of the -K
2831 filename. This should rule out race conditions when creating
2834 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2835 processing considerably. Previous code took too long for large mails,
2836 triggering a timeout which in turn triggers #401.
2838 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2839 in the DK code in transports.c. sendfile() is not really portable,
2840 hence the _LINUX specificness.
2842 TF/01 In the add_headers option to the mail command in an Exim filter,
2843 there was a bug that Exim would claim a syntax error in any
2844 header after the first one which had an odd number of characters
2847 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2848 callout verification, Exim cached a "reject" for the entire domain. This
2849 is correct for most verifications, but it is not correct for a recipient
2850 verification with use_sender or use_postmaster set, because in that case
2851 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2852 case of MAIL FROM:<> rejection from other early rejections (e.g.
2853 rejection of HELO). When verifying a recipient using a non-null MAIL
2854 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2855 Whatever the result of the callout, the value of the domain cache is
2856 left unchanged (for any other kind of callout, getting as far as trying
2857 RCPT means that the domain itself is ok).
2859 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2862 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2863 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2864 session, the daemon ignores ECONNECT errors and logs others; it now
2865 ignores EPIPE as well.
2867 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2868 (quoted-printable decoding).
2870 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2871 later the small subsequent patch to fix an introduced bug.
2873 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2875 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2877 PH/08 An error is now given if message_size_limit is specified negative.
2879 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2880 to be given (somewhat) arbitrary names.
2882 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2885 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2886 miscellaneous code fixes
2888 PH/10 Added the log_reject_target ACL modifier to specify where to log
2891 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2892 hostname. This is wrong, because it relates to the incoming message (and
2893 probably the interface on which it is arriving) and not to the outgoing
2894 callout (which could be using a different interface). This has been
2895 changed to use the value of the helo_data option from the smtp transport
2896 instead - this is what is used when a message is actually being sent. If
2897 there is no remote transport (possible with a router that sets up host
2898 addresses), $smtp_active_hostname is used.
2900 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2901 tweaks were necessary in order to get it to work (see also 21 below):
2902 (a) The code assumed that strncpy() returns a negative number on buffer
2903 overflow, which isn't the case. Replaced with Exim's string_format()
2905 (b) There were several signed/unsigned issues. I just did the minimum
2906 hacking in of casts. There is scope for a larger refactoring.
2907 (c) The code used strcasecmp() which is not a standard C function.
2908 Replaced with Exim's strcmpic() function.
2909 (d) The code set only $1; it now sets $auth1 as well.
2910 (e) A simple test gave the error "authentication client didn't specify
2911 service in request". It would seem that Dovecot has changed its
2912 interface. Fortunately there's a specification; I followed it and
2913 changed what the client sends and it appears to be working now.
2915 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2918 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2919 address A is aliased to B and C, where B exists and C does not. Without
2920 -v the output is "A verified" because verification stops after a
2921 successful redirection if more than one address is generated. However,
2922 with -v the child addresses are also verified. Exim was outputting "A
2923 failed to verify" and then showing the successful verification for C,
2924 with its parentage. It now outputs "B failed to verify", showing B's
2925 parentage before showing the successful verification of C.
2927 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2928 look up a TXT record in a specific list after matching in a combined
2931 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2932 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2933 they consult the DNS. I had assumed they would set it the way they
2934 wanted; and indeed my experiments on Linux seem to show that in some
2935 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2936 To be on the safe side, however, I have now made the interface to
2937 host_find_byname() similar to host_find_bydns(), with an argument
2938 containing the DNS resolver options. The host_find_byname() function now
2939 sets these options at its start, just as host_find_bydns() does. The smtp
2940 transport options dns_qualify_single and dns_search_parents are passed to
2941 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2942 of host_find_byname() use the default settings of RES_DEFNAMES
2943 (qualify_single) but not RES_DNSRCH (search_parents).
2945 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2946 spool_read_header() do less string testing, by means of a preliminary
2947 switch on the second character of optional "-foo" lines. (This is
2948 overdue, caused by the large number of possibilities that now exist.
2949 Originally there were few.) While I was there, I also converted the
2950 str(n)cmp tests so they don't re-test the leading "-" and the first
2951 character, in the hope this might squeeze out yet more improvement.
2953 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2954 flag allowing group syntax was set by the header_syntax check but not
2955 turned off, possible causing trouble later; (2) The flag was not being
2956 set at all for the header_verify test, causing "group"-style headers to
2957 be rejected. I have now set it in this case, and also caused header_
2958 verify to ignore an empty address taken from a group. While doing this, I
2959 came across some other cases where the code for allowing group syntax
2960 while scanning a header line wasn't quite right (mostly, not resetting
2961 the flag correctly in the right place). These bugs could have caused
2962 trouble for malformed header lines. I hope it is now all correct.
2964 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2965 with the "reply" argument non-NULL. The code, however (which originally
2966 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2967 but it didn't always do it. This confused somebody who was copying the
2968 code for some other use. I have removed all the tests.
2970 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2971 feature that was used to support insecure browsers during the U.S. crypto
2972 embargo. It requires special client support, and Exim is probably the
2973 only MTA that supported it -- and would never use it because real RSA is
2974 always available. This code has been removed, because it had the bad
2975 effect of slowing Exim down by computing (never used) parameters for the
2976 RSA_EXPORT functionality.
2978 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2979 authenticator to fail if there's a tab character in the incoming data
2980 (there should never be unless someone is messing about, as it's supposed
2981 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2982 if the connection is using TLS or if the remote IP is the same as the
2983 local IP, and the "valid-client-cert option" if a client certificate has
2986 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2987 authenticators. This can be used for authorization after authentication
2988 succeeds. (In the case of plaintext, it servers for both authentication
2991 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2992 if any retry times were supplied.
2994 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2995 connection, where there is no client IP address to check. In this
2996 situation, the verify now always succeeds.
2998 PH/25 Applied John Jetmore's -Mset patch.
3000 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3002 PH/27 In a string expansion for a processed (not raw) header when multiple
3003 headers of the same name were present, leading whitespace was being
3004 removed from all of them, but trailing whitespace was being removed only
3005 from the last one. Now trailing whitespace is removed from each header
3006 before concatenation. Completely empty headers in a concatenation (as
3007 before) are ignored.
3009 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3010 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3012 PH/29 [Removed. This was a change that I later backed out, and forgot to
3013 correct the ChangeLog entry (that I had efficiently created) before
3014 committing the later change.]
3016 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3017 address errors (4xx response to RCPT) over the same connection as other
3018 messages routed to the same hosts. Such deliveries are always "forced",
3019 so retry times are not inspected. This resulted in far too many retries
3020 for the affected addresses. The effect occurred only when there were more
3021 hosts than the hosts_max_try setting in the smtp transport when it had
3022 the 4xx errors. Those hosts that it had tried were not added to the list
3023 of hosts for which the message was waiting, so if all were tried, there
3024 was no problem. Two fixes have been applied:
3026 (i) If there are any address or message errors in an SMTP delivery, none
3027 of the hosts (tried or untried) are now added to the list of hosts
3028 for which the message is waiting, so the message should not be a
3029 candidate for sending over the same connection that was used for a
3030 successful delivery of some other message. This seems entirely
3031 reasonable: after all the message is NOT "waiting for some host".
3032 This is so "obvious" that I'm not sure why it wasn't done
3033 previously. Hope I haven't missed anything, but it can't do any
3034 harm, as the worst effect is to miss an optimization.
3036 (ii) If, despite (i), such a delivery is accidentally attempted, the
3037 routing retry time is respected, so at least it doesn't keep
3038 hammering the server.
3040 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3041 in ${readsocket because some servers need this prod.
3043 PH/32 Added some extra debug output when updating a wait-xxx database.
3045 PH/33 The hint "could be header name not terminated by colon", which has been
3046 given for certain expansion errors for a long time, was not being given
3047 for the ${if def:h_colon_omitted{... case.
3049 PH/34 The spec says: "With one important exception, whenever a domain list is
3050 being scanned, $domain contains the subject domain." There was at least
3051 one case where this was not true.
3053 PH/35 The error "getsockname() failed: connection reset by peer" was being
3054 written to the panic log as well as the main log, but it isn't really
3055 panic-worthy as it just means the connection died rather early on. I have
3056 removed the panic log writing for the ECONNRESET error when getsockname()
3059 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3060 runs only) independently of the message's sender address. This meant
3061 that, if the 4xx error was in fact related to the sender, a different
3062 message to the same recipient with a different sender could confuse
3063 things. In particular, this can happen when sending to a greylisting
3064 server, but other circumstances could also provoke similar problems.
3065 I have changed the default so that the retry time for these errors is now
3066 based a combination of the sender and recipient addresses. This change
3067 can be overridden by setting address_retry_include_sender=false in the
3070 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3071 remote server are returned as part of bounce messages. This was not
3072 happening for LMTP over a pipe (the lmtp transport), but now it is the
3073 same for both kinds of LMTP.
3075 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3076 in header lines that were in fact CNAMEs. This is no longer the case.
3078 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3079 and queue runs started by the daemon processed all messages. This has
3080 been fixed so that -R and -S can now usefully be given with -q<time>.
3082 PH/40 Import PCRE release 6.7 (fixes some bugs).
3084 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3086 PH/42 Give an error if -q is specified more than once.
3088 PH/43 Renamed the variables $interface_address and $interface_port as
3089 $received_ip_address and $received_port, to make it clear that these
3090 values apply to message reception, and not to the outgoing interface when
3091 a message is delivered. (The old names remain recognized, of course.)
3093 PH/44 There was no timeout on the connect() call when using a Unix domain
3094 socket in the ${readsocket expansion. There now is.
3096 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3097 be meaningful with "accept".
3099 SC/01 Eximstats V1.43
3100 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3102 SC/02 Eximstats V1.44
3103 Use a glob alias rather than an array ref in the generated
3104 parser. This improves both readability and performance.
3106 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3107 Collect SpamAssassin and rejection statistics.
3108 Don't display local sender or destination tables unless
3109 there is data to show.
3110 Added average volumes into the top table text output.
3112 SC/04 Eximstats V1.46
3113 Collect data on the number of addresses (recipients)
3114 as well as the number of messages.
3116 SC/05 Eximstats V1.47
3117 Added 'Message too big' to the list of mail rejection
3118 reasons (thanks to Marco Gaiarin).
3120 SC/06 Eximstats V1.48
3121 Mainlog lines which have GMT offsets and are too short to
3122 have a flag are now skipped.
3124 SC/07 Eximstats V1.49 (Alain Williams)
3125 Added the -emptyok flag.
3127 SC/08 Eximstats V1.50
3128 Fixes for obtaining the IP address from reject messages.
3130 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3131 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3132 whitespace changes from 4.64-PH/27
3134 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3137 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3138 are found, allow negative numbers in numeric criteria)
3140 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3142 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3143 to match changes made in 4.64-PH/43
3145 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3147 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3148 "domain" is actually an IP address, and force a failure. This locks out
3149 those revolvers/nameservers that support "A-for-A" lookups, in
3150 contravention of the specifications.
3152 PH/48 When a host name was looked up from an IP address, and the subsequent
3153 forward lookup of the name timed out, the host name was left in
3154 $sender_host_name, contrary to the specification.
3156 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3157 restricted to single-key lookups, Exim was not diagnosing an error if
3158 * or *@ was used with a query-style lookup.
3160 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3162 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3163 long ago, but noone interested enough thought of it. Let's just say that
3164 the "1.1" means that there are some new functions that weren't there at
3165 some point in the past.
3167 PH/51 Error processing for expansion failure of helo_data from an smtp
3168 transport during callout processing was broken.
3170 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3171 tested/used via the -bh/-bhc/-bs options.
3173 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3174 bug, fixed in subsequent PCRE releases).
3176 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3177 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3179 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3185 SC/01 Use a glob alias rather than an array ref in eximstats generated
3186 parser. This improves both readability and performance.
3188 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3189 Don't display local sender or destination tables in eximstats unless
3190 there is data to show.
3191 Added average volumes into the eximstats top table text output.
3193 SC/03 Collect data on the number of addresses (recipients) as well
3194 as the number of messages in eximstats.
3196 TF/01 Correct an error in the documentation for the redirect router. Exim
3197 does (usually) call initgroups() when daemonizing.
3199 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3200 with consistent privilege compared to when running as a daemon.
3202 TF/03 Note in the spec that $authenticated_id is not set for local
3203 submissions from trusted users.
3205 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3206 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3208 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3209 by adding some example configuration directives to the default
3210 configuration file. A little bit of work is required to uncomment the
3211 directives and define how usernames and passwords are checked, but
3212 there is now a framework to start from.
3214 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3215 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3216 without this. I don't know how relevant this is to other LDAP libraries.
3218 PH/02 Add the verb name to the "unknown ACL verb" error.
3220 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3222 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3224 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3225 directory not expanded when it should be if an expanded home directory
3226 was set for the address (which is overridden by the transport).
3228 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3231 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3232 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3233 because it is too late at that time, and has no effect.
3235 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3236 security issue with \' (bugzilla #107). I could not use the
3237 PQescapeStringConn() function, because it needs a PGconn value as one of
3240 PH/08 When testing addresses using -bt, indicate those final addresses that
3241 are duplicates that would not cause an additional delivery. At least one
3242 person was confused, thinking that -bt output corresponded to deliveries.
3243 (Suppressing duplicates isn't a good idea as you lose the information
3244 about possibly different redirections that led to the duplicates.)
3246 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3247 systems where poll() doesn't work, in particular OS X.
3249 PH/10 Added more information to debugging output for retry time not reached.
3251 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3252 operations in malware.c.
3254 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3257 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3258 syslog_duplication set false, log lines that would normally be written
3259 both the the main log and to the reject log were not written to syslog at
3262 PH/14 In the default configuration, change the use of "message" in ACL warn
3263 statements to "add_header".
3265 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3266 not followed by a command (e.g. "seen endif").
3268 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3269 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3272 PH/17 Added extra conditions to the default value of delay_warning_condition
3276 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3277 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3278 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3281 The Auto-Submitted: and various List- headers are standardised, whereas I
3282 don't think Precedence: ever was.
3284 PH/18 Refactored debugging code in route_finduser() to show more information,
3285 in particular, the error code if getpwnam() issues one.
3287 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3288 This is apparently needed in addition to the PH/07 change above to avoid
3289 any possible encoding problems.
3291 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3292 but not after initializing Perl.
3294 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3295 output them only if debugging. By default they are written stderr,
3296 apparently, which is not desirable.
3298 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3301 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3304 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3306 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3307 authenticated or an ident call has been made. Suppress the default
3308 values for $authenticated_id and $authenticated_sender (but permit -oMai
3309 and -oMas) when testing with -bh.
3311 PH/24 Re-jigged the order of the tests in the default configuration so that the
3312 tests for valid domains and recipients precede the DNS black list and CSA
3313 tests, on the grounds that those ones are more expensive.
3315 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3316 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3317 This bug exists in every version of Exim that I still have, right back to
3320 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3321 However, an attempt to turn on case-sensitivity in a regex key by
3322 including (?-i) didn't work because the subject string was already
3323 lowercased, and the effects were non-intuitive. It turns out that a
3324 one-line patch can be used to allow (?-i) to work as expected.
3330 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3331 other effects) broke the use of negated acl sub-conditions.
3333 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3336 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3337 "Deny" causes Exim to reject the incoming connection with a 554 error.
3338 Unfortunately, if there is a major crisis, such as a disk failure,
3339 tcp-wrappers gives "deny", whereas what one would like would be some
3340 kind of temporary error. A kludge has been added to help with this.
3341 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3342 554 error is used if errno is still zero or contains ENOENT (which occurs
3343 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3346 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3348 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3349 errors. Otherwise a message that provokes a temporary error (when other
3350 messages do not) can cause a whole host to time out.
3352 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3353 addresses were routed directly to files or pipes from a redirect router.
3354 File deliveries just didn't batch; pipe deliveries might have suffered
3357 PH/06 A failure to get a lock for a hints database would erroneously always say
3358 "Failed to get write lock", even when it was really a read lock.
3360 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3361 of 0600. This has been changed to use the value of the lockfile_mode
3362 option (which defaults to 0600).
3364 PH/08 Applied small patch from the Sieve maintainer.
3366 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3367 folder from quota calculations, a direct delivery into this folder messed
3368 up the contents of the maildirsize file. This was because the regex was
3369 used only to exclude .Trash (or whatever) when the size of the mailbox
3370 was calculated. There was no check that a delivery was happening into an
3371 excluded directory. This bug has been fixed by ignoring all quota
3372 processing for deliveries into excluded directories.
3374 PH/10 Added the maildirfolder_create_regex option to appendfile.
3380 PH/01 The code for finding all the local interface addresses on a FreeBSD
3381 system running IPv6 was broken. This may well have applied to all BSD
3382 systems, as well as to others that have similar system calls. The broken
3383 code found IPv4 interfaces correctly, but gave incorrect values for the
3384 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3385 that it would not match correctly against @[] and not recognize the IPv6
3388 PH/02 The ipliteral router was not recognizing addresses of the form user@
3389 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3391 PH/03 Added disable_ipv6.
3393 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3394 decoded form, because it is most often used to construct To: headers
3395 lines in autoreplies, and the decoded form may well be syntactically
3396 invalid. However, $reply_address has leading white space removed, and all
3397 newlines turned into spaces so that the autoreply transport does not
3400 PH/05 If group was specified without a user on a router, and no group or user
3401 was specified on a transport, the group from the router was ignored.
3403 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3404 for visible compile-time settings that can be used to change these
3405 numbers, for those that want even more. Backwards compatibility with old
3406 spool files has been maintained. However, going back to a previous Exim
3407 release will lost any variables that are in spool files.
3409 PH/07 Two small changes when running in the test harness: increase delay when
3410 passing a TCP/IP connection to a new process, in case the original
3411 process has to generate a bounce, and remove special handling of
3412 127.0.0.2 (sic), which is no longer necessary.
3414 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3415 be the same on different OS.
3417 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3420 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3421 whether --show-vars was specified or not
3423 JJ/02 exipick: Added support for new ACL variable spool format introduced
3426 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3427 syntactically invalid From: or Reply-to: line, and a filter used this to
3428 generate an autoreply, and therefore failed to obtain an address for the
3429 autoreply, Exim could try to deliver to a non-existent relative file
3430 name, causing unrelated and misleading errors. What now happens is that
3431 it logs this as a hard delivery error, but does not attempt to create a
3434 PH/11 The exinext utility has a -C option for testing purposes, but although
3435 the given file was scanned by exinext itself; it wasn't being passed on
3436 when Exim was called.
3438 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3439 an end-of-file indication when reading a command response.
3441 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3442 compiled. In many other places in Exim, IPv6 addresses are always
3443 recognized, so I have changed this. It also means that IPv4 domain
3444 literals of the form [IPV4:n.n.n.n] are now always recognized.
3446 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3447 used if the router is not running as root, for example, when verifying at
3448 ACL time, or when using -bh. The debugging output from this situation was
3449 non-existent - all you got was a failure to exec. I have made two
3452 (a) Failures to set uid/gid, the current directory, or a process leader
3453 in a subprocess such as that created by queryprogram now generate
3454 suitable debugging output when -d is set.
3456 (b) The queryprogram router detects when it is not running as root,
3457 outputs suitable debugging information if -d is set, and then runs
3458 the subprocess without attempting to change uid/gid.
3460 PH/15 Minor change to Makefile for building test_host (undocumented testing
3463 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3464 additional section of a DNS packet that returns MX or SRV records.
3465 Instead, it always explicitly searches for A/AAAA records. This avoids
3466 major problems that occur when a DNS server includes only records of one
3467 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3468 fixed another bug: if SRV records were looked up and the corresponding
3469 address records were *not* found in the additional section, the port
3470 values from the SRV records were lost.
3472 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3473 using the correct key (the original address) when searching the retry
3474 rules in order to find which one to use for generating the retry hint.
3476 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3477 adding the default one. Similarly, if it contains a Reply-To: header, the
3478 errors_reply_to option, if set, is not used.
3480 PH/19 When calculating a retry time, Exim used to measure the "time since
3481 failure" by looking at the "first failed" field in the retry record. Now
3482 it does not use this if it is later than than the arrival time of the
3483 message. Instead it uses the arrival time. This makes for better
3484 behaviour in cases where some deliveries succeed, thus re-setting the
3485 "first failed" field. An example is a quota failure for a huge message
3486 when small messages continue to be delivered. Without this change, the
3487 "time since failure" will always be short, possible causing more frequent
3488 delivery attempts for the huge message than are intended.
3489 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3491 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3492 $1, $2, $3) because the numerical variables can be reset during some
3493 expansion items (e.g. "match"), thereby losing the authentication data.
3495 PH/21 Make -bV show the size of off_t variables so that the test suite can
3496 decide whether to run tests for quotas > 2G.
3498 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3499 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3500 filecount value is greater than 2G or if a quota value is greater than 2G
3501 on a system where the size of off_t is not greater than 4, a panic error
3504 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3505 never match. The debug and -bh output now contains an explicit error
3506 message indicating a malformed IPv4 address or mask.
3508 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3509 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3510 PH/23 above applies.
3512 PH/25 Do not write to syslog when running in the test harness. The only
3513 occasion when this arises is a failure to open the main or panic logs
3514 (for which there is an explicit test).
3516 PH/26 Added the /no_tell option to "control=freeze".
3518 PH/27 If a host name lookup failed very early in a connection, for example, if
3519 the IP address matched host_lookup and the reverse lookup yielded a name
3520 that did not have a forward lookup, an error message of the form "no IP
3521 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3522 could be logged. Now it outputs the IP address instead of "NULL".
3524 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3525 allows the sender and the authenticated sender to be set when
3526 submitting a message from within Exim. Since child_open_exim() is
3527 documented for local_scan(), the new function should be too.
3529 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3530 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3531 results in an empty string is now treated as unset.
3533 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3535 PH/31 Added sender_verify_fail logging option.
3537 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3538 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3539 tidied the source and removed it altogether.
3541 PH/33 When a queue run was abandoned because the load average was too high, a
3542 log line was always written; now it is written only if the queue_run log
3543 selector is set. In addition, the log line for abandonment now contains
3544 information about the queue run such as the pid. This is always present
3545 in "start" and "stop" lines but was omitted from the "abandon" line.
3547 PH/34 Omit spaces between a header name and the colon in the error message that
3548 is given when verify = headers_syntax fails (if there are lots of them,
3549 the message gets confusing).
3551 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3552 names, as there are now some PTR records that contain slashes. This check
3553 is only to protect against broken name servers that fall over on strange
3554 characters, so the fact that it applies to all lookups doesn't matter.
3556 PH/36 Now that the new test suite is complete, we can remove some of the
3557 special code in Exim that was needed for the old test suite. For example,
3558 sorting DNS records because real resolvers return them in an arbitrary
3559 order. The new test suite's fake resolver always returns records in the
3562 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3563 bounces) except when queue_only is set, to avoid logging races between
3564 the different processes.
3566 PH/38 Panic-die if .include specifies a non-absolute path.
3568 PH/39 A tweak to the "H" retry rule from its user.
3570 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3571 a label. They prevented compilation on older perls.
3573 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3574 a warning to be raised on newish perls.
3576 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3577 on queue. Changes to match documented behaviour of showing count of
3578 messages matching specified criteria.
3580 PH/40 Changed the default ident timeout from 30s to 5s.
3582 PH/41 Added support for the use of login_cap features, on those BSD systems
3583 that have them, for controlling the resources used by pipe deliveries.
3585 PH/42 The content-scanning code uses fopen() to create files in which to put
3586 message data. Previously it was not paying any attention to the mode of
3587 the files. Exim runs with umask(0) because the rest of the code creates
3588 files with open(), and sets the required mode explicitly. Thus, these
3589 files were ending up world-writeable. This was not a big issue, because,
3590 being within the spool directory, they were not world-accessible. I have
3591 created a function called modefopen, which takes an additional mode
3592 argument. It sets umask(777), creates the file, chmods it to the required
3593 mode, then resets the umask. All the relevant calls to fopen() in the
3594 content scanning code have been changed to use this function.
3596 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3597 to 24 hours. This avoids potential overflow problems when processing G
3598 and H retry rules. I suspect nobody ever tinkers with this value.
3600 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3602 PH/45 When the plaintext authenticator is running as a client, the server's
3603 challenges are checked to ensure they are valid base64 strings. By
3604 default, the authentication attempt is cancelled if an invalid string is
3605 received. Setting client_ignore_invalid_base64 true ignores these errors.
3606 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3607 they are received. Thus, the responses can be made to depend on the
3608 challenges. If an invalid string is ignored, an empty string is placed in
3611 PH/46 Messages that are created by the autoreply transport now contains a
3612 References: header, in accordance with RFCs 2822 and 3834.
3614 PH/47 Added authenticated_sender_force to the smtp transport.
3616 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3618 PH/49 Installed latest patch from the Sieve maintainer.
3620 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3621 was also set, the appendfile transport was unnecessarily scanning a
3622 directory of message files (e.g. for maildir delivery) to find the count
3623 of files (along with the size), even though it did not need this
3624 information. It now does the scan only if it needs to find either the
3625 size of the count of files.
3627 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3629 PH/52 Two bugs concerned with error handling when the smtp transport is
3632 (i) Exim was not creating retry information for temporary errors given
3633 for individual recipients after the DATA command when the smtp transport
3634 was used in LMTP mode. This meant that they could be retried too
3635 frequently, and not timed out correctly.
3637 (ii) Exim was setting the flag that allows error details to be returned
3638 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3639 recipients that were returned after the DATA command.
3641 PH/53 This is related to PH/52, but is more general: for any failing address,
3642 when detailed error information was permitted to be returned to the
3643 sender, but the error was temporary, then after the final timeout, only
3644 "retry timeout exceeded" was returned. Now it returns the full error as
3645 well as "retry timeout exceeded".
3647 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3648 do this, and (what is worse) MTAs that accept it.
3650 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3651 will now be deprecated.
3653 PH/56 New os.c-cygwin from the Cygwin maintainer.
3655 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3656 formats (previously only available in exim formats via -bpr, -bpru,
3657 and -bpra. Now also available in native and exiqgrep formats)
3659 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3660 with very large, slow to parse queues
3662 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3664 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3666 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3667 responses to authentication challenges, though it was showing the
3668 challenges; (ii) I've removed the CR characters from the debug output for
3671 PH/58 Allow for the insertion of a newline as well as a space when a string
3672 is turned into more than one encoded-word during RFC 2047 encoding. The
3673 Sieve code now uses this.
3675 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3676 data_4xx, lost_connection, tls_required.
3678 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3679 message was being sent as an SMTP response.
3681 PH/61 Add -l and -k options to exicyclog.
3683 PH/62 When verifying, if an address was redirected to one new address, so that
3684 verification continued, and the new address failed or deferred after
3685 having set something in $address_data, the value of $address_data was not
3686 passed back to the ACL. This was different to the case when no
3687 redirection occurred. The value is now passed back in both cases.
3689 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3690 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3691 use login_cap.h, so on its own it isn't the distinguishing feature. The
3692 new name refers directly to the setclassresources() function.
3694 PH/65 Added configuration files for NetBSD3.
3696 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3698 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3699 is preferred over IPv4.
3701 PH/68 The bounce_return_message and bounce_return_body options were not being
3702 honoured for bounces generated during the reception of non-SMTP messages.
3703 In particular, this applied to messages rejected by the ACL. This bug has
3704 been fixed. However, if bounce_return_message is true and bounce_return_
3705 body is false, the headers that are returned for a non-SMTP message
3706 include only those that have been read before the error was detected.
3707 (In the case of an ACL rejection, they have all been read.)
3709 PH/69 The HTML version of the specification is now built in a directory called
3710 spec_html instead of spec.html, because the latter looks like a path with
3711 a MIME-type, and this confuses some software.
3713 PH/70 Catch two compiler warnings in sieve.c.
3715 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3716 function verify_get_ident() calls ip_connect() to connect a socket, but
3717 if the "connect()" function timed out, ip_connect() used to close the
3718 socket. However, verify_get_ident() also closes the socket later, and in
3719 between Exim writes to the log, which may get opened at this point. When
3720 the socket was closed in ip_connect(), the log could get the same file
3721 descriptor number as the socket. This naturally causes chaos. The fix is
3722 not to close the socket in ip_connect(); the socket should be closed by
3723 the function that creates it. There was only one place in the code where
3724 this was missing, in the iplookup router, which I don't think anybody now
3725 uses, but I've fixed it anyway.
3727 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3728 well as to direct DNS lookups. Otherwise the handling of names in host
3729 lists is inconsistent and therefore confusing.
3735 PH/01 Two changes to the default runtime configuration:
3737 (1) Move the checks for relay_from_hosts and authenticated clients from
3738 after to before the (commented out) DNS black list checks.
3740 (2) Add control=submission to the relay_from_hosts and authenticated
3741 clients checks, on the grounds that messages accepted by these
3742 statements are most likely to be submissions.
3744 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3746 (1) Generate an error if the third argument for the ${prvs expansion is
3749 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3752 (3) Reset the variables that are obtained from the first argument of
3753 ${prvscheck and used in the second argument before leaving the code,
3754 because their memory is reclaimed, so using them afterwards may do
3757 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3758 one (it's much easier than Tom thought :-).
3760 (5) Because of (4), we can now allow for the use of $prvscheck_result
3761 inside the third argument.
3763 PH/03 For some reason, the default setting of PATH when running a command from
3764 a pipe transport was just "/usr/bin". I have changed it to
3767 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3768 anything to be listed in the output from -bV.
3770 PH/05 When a filter generated an autoreply, the entire To: header line was
3771 quoted in the delivery log line, like this:
3773 => >A.N.Other <ano@some.domain> <original@ddress> ...
3775 This has been changed so that it extracts the operative address. There
3776 may be more than one such address. If so, they are comma-separated, like
3779 => >ano@some.domain,ona@other.domain <original@ddress> ...
3781 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3782 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3783 not being looked up in the rDNS to get a host name, Exim was showing the
3784 IP address twice in Received: lines, even though the IP addresses were
3785 identical. For example:
3787 Received: from [1.2.3.4] (helo=[1.2.3.4])
3789 However, if the real host name was known, it was omitting the HELO data
3790 if it matched the actual IP address. This has been tidied up so that it
3791 doesn't show the same IP address twice.
3793 PH/07 When both +timestamp and +memory debugging was on, the value given by
3794 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3795 called by the debug printing, thereby overwriting the timestamp buffer.
3796 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3798 PH/08 When the original message was included in an autoreply transport, it
3799 always said "this is a copy of the message, including all the headers",
3800 even if body_only or headers_only was set. It now gives an appropriate
3803 PH/09 Applied a patch from the Sieve maintainer which:
3805 o fixes some comments
3806 o adds the (disabled) notify extension core
3807 o adds some debug output for the result of if/elsif tests
3808 o points to the current vacation draft in the documentation
3809 and documents the missing references header update
3813 o fixes a bug in processing the envelope test (when testing
3814 multiple envelope elements, the last element determined the
3817 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3818 Electronic Mail") by including:
3820 Auto-submitted: auto-generated
3822 in the messages that it generates (bounce messages and others, such as
3823 warnings). In the case of bounce messages for non-SMTP messages, there was
3824 also a typo: it was using "Auto_submitted" (underscore instead of
3825 hyphen). Since every message generated by Exim is necessarily in response
3826 to another message, thes have all been changed to:
3828 Auto-Submitted: auto-replied
3830 in accordance with these statements in the RFC:
3832 The auto-replied keyword:
3834 - SHOULD be used on messages sent in direct response to another
3835 message by an automatic process,
3837 - MUST NOT be used on manually-generated messages,
3839 - MAY be used on Delivery Status Notifications (DSNs) and Message
3840 Disposition Notifications (MDNs),
3842 - MUST NOT be used on messages generated by automatic or periodic
3843 processes, except for messages which are automatic responses to
3846 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3847 to the default Received: header definition.
3849 PH/12 Added log selector acl_warn_skipped (default on).
3851 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3852 variables because (a) they are in the wrong storage pool and (b) even if
3853 they were copied, it wouldn't work properly because of the caching.
3855 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3856 checking when decoding. Apparently there are clients that generate
3857 overlong encoded strings. Why am I not surprised?
3859 PH/15 If the first argument of "${if match_address" was not empty, but did not
3860 contain an "@" character, Exim crashed. Now it writes a panic log message
3861 and treats the condition as false.
3863 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3865 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3866 extension "envelope-auth". The code is finished and in agreement with
3867 other implementations, but there is no documentation so far and in fact,
3868 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3869 not changing the active code.
3871 Print executed "if" and "elsif" statements when debugging is used. This
3872 helps a great deal to understand what a filter does.
3874 Document more things not specified clearly in RFC3028. I had all this
3875 sorted out, when out of a sudden new issues came to my mind. Oops."
3877 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3880 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3881 canonical form (as documented). However, after a host name lookup from
3882 the IP address, check_host() was doing a simple string comparison with
3883 addresses acquired from the DNS when checking that the found name did
3884 have the original IP as one of its addresses. Since any found IPv6
3885 addresses are likely to be in abbreviated form, the comparison could
3886 fail. Luckily, there already exists a function for doing the comparison
3887 by converting both addresses to binary, so now that is used instead of
3888 the text comparison.
3890 PH/20 There was another similar case to PH/19, when a complete host name was
3891 given in a host list; looking up its IP address could give an abbreviated
3892 form, whereas the current host's name might or might not be abbreviated.
3893 The same fix has been applied.
3899 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3900 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3903 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3904 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3906 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3908 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3909 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3910 Also, for "Auto-Submitted", treat anything other than "no" as
3911 non-personal, in accordance with RFC 3834. (Previously it treated
3912 anything starting "auto-" as non-personal.)
3914 TF/01 The control=submission/name=... option had a problem with syntax
3915 errors if the name included a slash character. The /name= option
3916 now slurps the rest of the string, so it can include any characters
3917 but it must come last in the list of options (after /sender_retain
3920 PH/05 Some modifications to the interface to the fake nameserver for the new
3928 TK/01 Added the "success_on_redirect" address verification option. See
3929 NewStuff for rationale and an example.
3931 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3933 PH/02 Patch to exigrep to allow it to work on syslog lines.
3935 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3936 fread() to skip over the body file's header line, because in Cygwin the
3937 header line is locked and is inaccessible.
3939 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3940 co-exist for some time) to make it clear that it is the Exim ID that is
3941 referenced, not the Message-ID: header line.
3943 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3944 string_format() function, because snprintf() does not exist on all
3947 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3948 ${stat: expansion item.
3950 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3951 protocol synchronization error", to keep the pedants happy.
3953 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3954 well as for IRIX systems, when gcc is being used. See the host.c source
3957 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3959 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3962 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3963 transport and to the smtp transport in LMTP mode.
3965 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3967 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3968 run a filter in a subprocess. This could lead to confusion in subsequent
3969 lookups in the parent process. There should also be a search_tidyup() at
3970 the end of the subprocess.
3972 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3973 only if the host matched helo_try_verify_hosts, which caused the
3974 verification to occur when the EHLO/HELO command was issued. The ACL just
3975 tested the remembered result. Now, if a previous verification attempt has
3976 not happened, "verify = helo" does it there and then.
3978 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3980 TK/03 Fix log output including CR from clamd.
3982 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3983 exist provoked a memory error which could cause a segfault.
3985 PH/15 Installed PCRE 6.2
3987 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3989 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3990 of the problem. Specifically, suggested +O2 rather than +O1 for the
3993 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3995 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3996 record, the port was not correctly logged when the outgoing_port log
3997 selector was set (it logged the transort's default port).
3999 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4000 fallback_hosts, and "hosts" in the smtp transport.
4002 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4003 host errors such as "Connection refused".
4005 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4006 authentication with radiusclient 0.4.9:
4008 - Error returned from rc_read_config was caught wrongly
4009 - Username/password not passed on to radius server due to wrong length.
4011 The presumption is that some radiusclient API changes for 4.51/PH/17
4012 were not taken care of correctly. The code is still untested by me (my
4013 Linux distribution still has 0.3.2 of radiusclient), but it was
4014 contributed by a Radius user.
4016 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4017 expanding the "port" option of the smtp transport.
4019 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4020 while reading a MIME header. Thanks to Tom Hughes for a patch.
4022 PH/24 Include config.h inside local_scan.h so that configuration settings are
4025 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4026 that in an ACL for RCPT (for example), you can examine exactly what was
4029 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4030 commands, but it was not correctly comparing the address with the actual
4031 client host address. Thus, it would show the EHLO address in Received:
4032 header lines when this was not necessary.
4034 PH/27 Added the % operator to ${eval:}.
4036 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4037 it should be ignoring failures (because with -C, for example, it has lost
4038 privilege). It wasn't ignoring creation failures other than "already
4041 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4044 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4045 because an input line was too long, either on its own, or by virtue of
4046 too many continuations, the temporary file was not being removed, and the
4047 return code was incorrect.
4049 PH/31 Missing "BOOL" in function definition in filtertest.c.
4051 PH/32 Applied Sieve patches from the maintainer.
4053 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4055 PH/33 Added "verify = not_blind".
4057 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4058 Local/Makefile (with some defaults set). These are used in built scripts
4059 such as exicyclog, but they have never been used in the exim_install
4060 script (though there are many overriding facilities there). I have
4061 arranged that the exim_install script now takes note of these two
4064 PH/35 Installed configuration files for Dragonfly.
4066 PH/36 When a locally submitted message by a trusted user did not contain a
4067 From: header, and the sender address was obtained from -f or from an SMTP
4068 MAIL command, and the trusted user did not use -F to supply a sender
4069 name, $originator_name was incorrectly used when constructing a From:
4070 header. Furthermore, $originator_name was used for submission mode
4071 messages from external hosts without From: headers in a similar way,
4072 which is clearly wrong.
4074 PH/37 Added control=suppress_local_fixups.
4076 PH/38 When log_selector = +received_sender was set, and the addition of the
4077 sender made the log line's construction buffer exactly full, or one byte
4078 less than full, an overflow happened when the terminating "\n" was
4081 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4082 when the result of a list match is failure because a DNS lookup failed.
4084 PH/40 RM_COMMAND is now used in the building process.
4086 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4087 the "build-* directories that it finds.
4089 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4090 address in a domain literal was a prefix of an interface address.
4092 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4093 when verifying a sender address, unless rewrite_headers is false.
4095 PH/44 Wrote a long comment about why errors_to addresses are verified as
4096 recipients, not senders.
4098 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4099 the ratelimit ACL was added.
4101 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4103 PH/46 Added extra information about PostgreSQL errors to the error string.
4105 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4106 avoiding the need to install special zones in a real server. This is
4107 backwards compatible; if it can't find the fake resolver, it drops back.
4108 Thus, both old and new test suites can be run.
4110 TF/02 Added util/ratelimit.pl
4112 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4113 clock is set back in time.
4115 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4116 Candler <B.Candler@pobox.com>.
4118 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4119 OK for addresses that are the result of redirections.
4121 PH/48 A number of further additions for the benefit of the new test suite,
4122 including a fake gethostbyname() that interfaces to the fake DNS resolver
4125 TF/06 The fix for widen_domains has also been applied to qualify_single and
4126 search_parents which are the other dnslookup options that can cause
4129 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4132 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4134 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4135 service" if it encountered an MX record with an empty target hostname.
4136 The message is now "an MX or SRV record indicated no SMTP service".
4138 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4139 if the DNS of the sending site is misconfigured. This is quite a
4140 common situation. This change restores the behaviour of treating a
4141 helo verification defer as a failure.
4143 PH/51 If self=fail was set on a router, the bounce message did not include the
4144 actual error message.
4150 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4152 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4153 command itself ended in error, the underlying message about the transport
4154 filter timeout was being overwritten with the pipe command error. Now the
4155 underlying error message should be appended to the second error message.
4157 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4159 PH/02 Reduce the amount of output that "make" produces by default. Full output
4160 can still be requested.
4162 PH/03 The warning log line about a condition test deferring for a "warn" verb
4163 was being output only once per connection, rather than after each
4164 occurrence (because it was using the same function as for successful
4165 "warn" verbs). This seems wrong, so I have changed it.
4167 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4168 it should not have, which might have caused a crash in the right
4169 circumstances, but probably never did.
4171 PH/04 Installed a modified version of Tony Finch's patch to make submission
4172 mode fix the return path as well as the Sender: header line, and to
4173 add a /name= option so that you can make the user's friendly name appear
4176 TF/03 Added the control = fakedefer ACL modifier.
4178 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4179 Mark Lowes for thorough testing.
4181 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4183 TK/03 Merged latest SRS patch from Miles Wilton.
4185 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4186 with the definition in sysexits.h (which is #included earlier).
4187 Fortunately, Exim does not actually use EX_OK. The code used to try to
4188 preserve the sysexits.h value, by assuming that macro definitions were
4189 scanned for macro replacements. I have been disabused of this notion,
4190 so now the code just undefines EX_OK before #including unistd.h.
4192 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4193 in the smtp transport. When a block could not be written in a single
4194 write() function, the timeout was being re-applied to each part-write.
4195 This seems wrong - if the receiver was accepting one byte at a time it
4196 would take for ever. The timeout is now adjusted when this happens. It
4197 doesn't have to be particularly precise.
4199 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4200 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4202 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4203 without a domain if the check to <postmaster@domain> fails.
4205 SC/01 Eximstats: added -xls and the ability to specify output files
4206 (patch written by Frank Heydlauf).
4208 SC/02 Eximstats: use FileHandles for outputting results.
4210 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4212 SC/04 Eximstats: fixed display of large numbers with -nvr option
4214 SC/05 Eximstats: fixed merging of reports with empty tables.
4216 SC/06 Eximstats: added the -include_original_destination flag
4218 SC/07 Eximstats: removed tabs and trailing whitespace.
4220 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4222 TK/06 MBOX spool code: Add real "From " MBOX separator line
4223 so the .eml file is really in mbox format (even though
4224 most programs do not really care). Patch from Alex Miller.
4226 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4227 The latter is generated from $received_to and is only set if the
4228 message has one envelope recipient. SA can use these headers,
4229 obviously out-of-the-box. Patch from Alex Miller.
4231 PH/08 The ${def test on a variable was returning false if the variable's
4232 value was "0", contrary to what the specification has always said!
4233 The result should be true unless the variable is empty.
4235 PH/09 The syntax error of a character other than { following "${if
4236 def:variable_name" (after optional whitespace) was not being diagnosed.
4237 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4238 accidental colon was present, for example, could give incorrect results.
4240 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4241 result is used (not including appendfile, where other changes are about
4244 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4245 This involved changing a lot of size variables from int to off_t. It
4246 should work with maildirs and everything.
4248 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4249 spamd dying while we are connected to it.
4251 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4254 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4255 function for BDB 4.3.
4257 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4259 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4260 the sources for PCRE have been re-arranged and more files are now
4263 PH/15 The code I had for printing potentially long long variables in PH/11
4264 above was not the best (it lost precision). The length of off_t variables
4265 is now inspected at build time, and an appropriate printing format (%ld
4266 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4267 to be "long long int" or "long int". This is needed for the internal
4268 formatting function string_vformat().
4270 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4271 the configuration file to be ":syslog", then the script "guesses" where
4272 the logs files are, rather than using the compiled in default. In our
4273 case the guess is not the same as the compiled default, so the script
4274 suddenly stopped working when I started to use syslog. The patch checks
4275 to see if log_file_path is "". If so, it attempts to read it from exim
4276 with no configuration file to get the compiled in version, before it
4277 falls back to the previous guessing code."
4279 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4280 implementing BATV in an Exim configuration. See NewStuff for the gory
4283 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4284 Makefile that are specific to HP-UX.
4286 PH/18 If the "use_postmaster" option was set for a recipient callout together
4287 with the "random" option, the postmaster address was used as the MAIL
4288 FROM address for the random test, but not for the subsequent recipient
4289 test. It is now used for both.
4291 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4292 patch removes a few documentation additions to RFC 3028, because the
4293 latest draft now contains them. It adds the new en;ascii-case comparator
4294 and a new error check for 8bit text in MIME parts. Comparator and
4295 require names are now matched exactly. I enabled the subaddress
4296 extension, but it is not well tested yet (read: it works for me)."
4298 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4299 rework some of the code of TK/09 above to avoid the hardwired use of
4300 "%lld" and "long long". Replaced the call to snprintf() with a call to
4303 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4304 records point to non-existent hosts", "retry timeout exceeded", and
4305 "retry time not reached for any host after a long failure period".
4307 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4308 experimental DomainKeys support:
4310 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4311 (2) On an error such as an illegally used "control", the wrong name for
4312 the control was given.
4314 These problems did NOT occur unless DomainKeys support was compiled.
4316 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4318 PH/24 Added ${if match_ip condition.
4320 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4321 databases so that it will be absolutely obvious if a crash occurs in the
4322 DB library. This is a regular occurrence (often caused by mis-matched
4325 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4326 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4327 picked up on a user's system that detects such things. There doesn't seem
4328 to be a gcc warning option for this - only an attribute that has to be
4329 put on the function's prototype. It seems that in Fedora Core 4 they have
4330 set this on a number of new functions. No doubt there will be more in due
4333 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4334 specify a transport. However, if an address that was verified by such a
4335 router was the subject of a callout, Exim crashed because it tried to
4336 read the rcpt_include_affixes from the non-existent transport. Now it
4337 just assumes that the setting of that option is false. This bug was
4338 introduced by 4.51/PH/31.
4340 PH/28 Changed -d+all to exclude +memory, because that information is very
4341 rarely of interest, but it makes the output a lot bigger. People tend to
4342 do -d+all out of habit.
4344 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4345 code in os-type was giving problems when libc.so lives in lib64, like on
4348 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4349 aren't the modern standard, and it seems that some systems' include files
4350 don't always have them. Exim was already checking for some of the newer
4351 ones like T_AAAA, and defining it itself. I've added checks for all the
4352 record types that Exim uses.
4354 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4355 not automatically generating a new one, as it is supposed to. This
4356 prevented TLS from working. If the file did exist, but contained invalid
4357 data, a new version was generated, as expected. It was only the case of a
4358 non-existent file that was broken.
4360 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4361 with a change in libdomainkeys > 0.64.
4363 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4364 from DNS. If the selector record carries the flag, it now has
4365 precedence over the domain-wide flag.
4367 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4369 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4370 the use of a port name that isn't defined in /etc/services) occurred, the
4371 message was deferred as in a normal delivery, and thus remained on the
4372 spool, instead of being failed because of the mua_wrapper setting. This
4373 is now fixed, and I tidied up some of the mua_wrapper messages at the
4376 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4377 the messages in a hash of arrays rather than using individual hashes.
4378 This is a bit cleaner and results in dramatic memory savings, albeit
4379 at a slight CPU cost.
4381 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4382 as requested by Marc Sherman.
4384 SC/10 Eximstats: added histograms for user specified patterns as requested
4387 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4389 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4390 fopen() in the content-scanning modules that did not already have it.
4396 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4397 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4399 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4401 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4403 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4404 to test Sieve filters that use "vacation".
4406 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4407 that changes the way the GnuTLS parameters are stored in the cache file.
4408 The new format can be generated externally. For backward compatibility,
4409 if the data in the cache doesn't make sense, Exim assumes it has read an
4410 old-format file, and it generates new data and writes a new file. This
4411 means that you can't go back to an older release without removing the
4414 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4415 work if there are any delivery delays because "one_time" forces the
4416 parent to be marked "delivered", so its unseen clone is never tried
4417 again. For this reason, Exim now forbids the simultaneous setting of
4420 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4421 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4422 entry if you want to know the details. The fix, however, neglected to
4423 consider the case when local delivery batching is involved. The test for
4424 "previously delivered" was not happening when checking to see if an
4425 address could be batched with a previous (undelivered) one; under
4426 certain circumstances this could lead to multiple deliveries to the same
4429 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4430 in its include files, and this causes problems building Exim.
4432 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4433 header_syntax) but Exim was just ignoring anything given after a slash.
4434 In particular, this caused confusion with an attempt to use "verify =
4435 reverse_host_lookup/defer_ok". An error is now given when options are
4436 supplied for verify items that do not have them. (Maybe reverse_host_
4437 lookup should have a defer_ok option, but that's a different point.)
4439 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4440 defined by RFC 821) to 2048, because there were problems with some AUTH
4441 commands, and RFC 1869 says the size should be increased for extended
4442 SMTP commands that take arguments.
4444 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4447 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4448 "unknown" error; now it says that the functionality isn't in the binary.
4450 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4451 an address' error message when a string expansion fails (syntax or
4452 whatever). Otherwise the password may appear in the log. Following change
4453 PH/42 below, there is no longer a chance of it appearing in a bounce
4456 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4458 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4459 of its addresses was ever tried. (Bugzilla bug #2.)
4461 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4462 the result incorrectly in the debug output. (It correctly added a newline
4463 to what was transported.)
4465 TF/01 Added $received_time.
4467 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4468 commented out examples of how to interface to a virus scanner and to
4469 SpamAssassin. Also added commented examples of av_scanner and
4470 spamd_address settings.
4472 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4473 and controls are allowed in which ACLs. There were a couple of minor
4474 errors. Some of the entries in the conditions table (which is a table of
4475 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4476 negation of where the condition IS allowed.
4478 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4480 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4481 header file does not have a version number, so I've had to invent a new
4482 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4483 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4484 radiusclient), but it was contributed by a Radius user.
4486 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4487 files or directories, for OpenSSL.
4489 PH/19 When an Exim process that is running as root has to create an Exim log
4490 file, it does so in a subprocess that runs as exim:exim so as to get the
4491 ownership right at creation (otherwise, other Exim processes might see
4492 the file with the wrong ownership). There was no test for failure of this
4493 fork() call, which would lead to the process getting stuck as it waited
4494 for a non-existent subprocess. Forks do occasionally fail when resources
4495 run out. I reviewed all the other calls to fork(); they all seem to check
4498 PH/20 When checking for unexpected SMTP input at connect time (before writing
4499 the banner), Exim was not dealing correctly with a non-positive return
4500 from the read() function. If the client had disconnected by this time,
4501 the result was a log entry for a synchronization error with an empty
4502 string after "input=" when read() returned zero. If read() returned -1
4503 (an event I could not check), uninitialized data bytes were printed.
4504 There were reports of junk text (parts of files, etc) appearing after
4507 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4509 PH/22 Added support for macro redefinition, and (re)definition in between
4510 driver and ACL definitions.
4512 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4513 forgetting to use the resulting value; it was using the unexpanded value.
4515 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4516 hadn't been configured. The fix is from Juergen Kreileder, who
4517 understands it better than I do:
4519 "Here's what I see happening with three configured cyrus_sasl
4520 authenticators configured (plain, login, cram-md5):
4522 On startup auth_cyrus_sasl_init() gets called for each of these.
4523 This means three calls to sasl_listmech() without a specified mech_list.
4524 => SASL tests which mechs of all available mechs actually work
4525 => three warnings about OTP not working
4526 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4528 With the patch, sasl_listmech() also gets called three times. But now
4529 SASL's mech_list option is set to the server_mech specified in the the
4530 authenticator. Or in other words, the answer from sasl_listmech()
4531 gets limited to just the mech you're testing for (which is different
4533 => the return list contains just 'plain' or 'login', 'cram-md5' or
4534 nothing depending on the value of ob->server_mech.
4536 I've just tested the patch: Authentication still works fine,
4537 unavailable mechs specified in the exim configuration are still
4538 caught, and the auth.log warnings about OTP are gone."
4540 PH/25 When debugging is enabled, the contents of the command line are added
4541 to the debugging output, even when log_selector=+arguments is not
4544 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4545 answer is "GNU", and only if the return is "GNU/something" is the answer
4548 PH/27 $acl_verify_message is now set immediately after the failure of a
4549 verification in an ACL, and so is available in subsequent modifiers. In
4550 particular, the message can be preserved by coding like this:
4552 warn !verify = sender
4553 set acl_m0 = $acl_verify_message
4555 Previously, $acl_verify_message was set only while expanding "message"
4556 and "log_message" when a very denied access.
4558 PH/28 Modified OS/os.c-Linux with
4560 -#ifndef OS_LOAD_AVERAGE
4561 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4563 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4564 nomenclature these days.)
4566 PH/29 Installed patch from the Sieve maintainer that adds the options
4567 sieve_useraddress and sieve_subaddress to the redirect router.
4569 PH/30 In these circumstances:
4570 . Two addresses routed to the same list of hosts;
4571 . First host does not offer TLS;
4572 . First host accepts first address;
4573 . First host gives temporary error to second address;
4574 . Second host offers TLS and a TLS session is established;
4575 . Second host accepts second address.
4576 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4577 and peerdn, if requested) that were in fact used only for the second
4580 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4581 attention to any local part prefix or suffix that was matched by the
4582 router that accepted the address. It now behaves in the same way as it
4583 does for delivery: the affixes are removed from the local part unless
4584 rcpt_include_affixes is set on the transport.
4586 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4587 timeout during the DATA phase of an incoming message.
4589 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4590 applied a patch sanctioned by the Sieve maintainer.
4592 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4593 the uid or gid is negative. A case of a negative gid caused this to be
4594 noticed. The fix allows for either to be negative.
4596 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4597 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4600 PH/36 The change PH/12 above was broken. Fixed it.
4602 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4603 the grounds that routing the same address twice would always produce the
4604 same answer. This might have been true once, but it is certainly no
4605 longer true now. Routing a child address may depend on the previous
4606 routing that produced that child. Some complicated redirection strategies
4607 went wrong when messages had multiple recipients, and made Exim's
4608 behaviour dependent on the order in which the addresses were given.
4610 I have moved the duplicate checking until after the routing is complete.
4611 Exim scans the addresses that are assigned to local and remote
4612 transports, and removes any duplicates. This means that more work will be
4613 done, as duplicates will always all be routed, but duplicates are
4614 presumably rare, so I don't expect this is of any significance.
4616 For deliveries to pipes, files, and autoreplies, the duplicate checking
4617 still happens during the routing process, since they are not going to be
4620 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4621 It corrects a timeout issue with spamd. This is Ian's comment: "The
4622 background is that sometimes spamd either never reads data from a
4623 connection it has accepted, or it never writes response data. The exiscan
4624 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4625 blindly assumes that writes won't block so it may never time out."
4627 PH/39 Allow G after quota size as well as K and M.
4629 PH/40 The value set for $authenticated_id in an authenticator may not contain
4630 binary zeroes or newlines because the value is written to log lines and
4631 to spool files. There was no check on this. Now the value is run through
4632 the string_printing() function so that such characters are converted to
4633 printable escape sequences.
4635 PH/41 $message_linecount is a new variable that contains the total number of
4636 lines in the message. Compare $body_linecount, which is the count for the
4639 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4640 bounce and delay warning messages, except in certain special cases, which
4643 (a) An SMTP error message from a remote host;
4644 (b) A message specified in a :fail: redirection;
4645 (c) A message specified in a "fail" command in a system filter;
4646 (d) A message specified in a FAIL return from the queryprogram router;
4647 (e) A message specified by the cannot_route_message router option.
4649 In these cases only, Exim does include the error details in bounce and
4650 warning messages. There are also a few cases where bland messages such
4651 as "unrouteable address" or "local delivery error" are given.
4653 PH/43 $value is now also set for the "else" part of a ${run expansion.
4655 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4656 being worked on, but at least Exim now implements the latest version to
4659 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4660 process to complete was treated as a delivery failure, a timeout while
4661 writing the message to the pipe was logged, but erroneously treated as a
4662 successful delivery. Such timeouts include transport filter timeouts. For
4663 consistency with the overall process timeout, these timeouts are now
4664 treated as errors, giving rise to delivery failures by default. However,
4665 there is now a new Boolean option for the pipe transport called
4666 timeout_defer, which, if set TRUE, converts the failures into defers for
4667 both kinds of timeout. A transport filter timeout is now identified in
4670 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4671 systems where "make" and "gmake" are different, calling "gmake" at top
4672 level broke things. I've arranged for the value of $(MAKE) to be passed
4673 from the Makefile to this script so that it can call the same version of
4677 A note about Exim versions 4.44 and 4.50
4678 ----------------------------------------
4680 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4681 changes of various kinds. As a consequence, a big documentation update was
4682 needed. This delayed the release for rather longer than seemed good, especially
4683 in the light of a couple of (minor) security issues. Therefore, the changes
4684 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4685 release. So 4.44 and 4.50 are in effect two different branches that both start
4688 I have left the 4.50 change log unchanged; it contains all the changes since
4689 4.43. The change log for 4.44 is below; many of its items are identical to
4690 those for 4.50. This seems to be the most sensible way to preserve the
4691 historical information.
4697 1. Minor wording change to the doc/README.SIEVE file.
4699 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4700 computation of the current number of files was incorrect.
4702 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4703 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4706 4. Give more explanation in the error message when the command for a transport
4707 filter fails to execute.
4709 5. There are several places where Exim runs a non-Exim command in a
4710 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4711 was being done only for the command run by the queryprogram router. It is
4712 now done for all such subprocesses. The other cases are: ${run, transport
4713 filters, and the commands run by the lmtp and pipe transports.
4715 6. Added CONFIGURE_GROUP build-time option.
4717 7. Some older OS have a limit of 256 on the maximum number of file
4718 descriptors. Exim was using setrlimit() to set 1000 as a large value
4719 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4720 systems. I've change it so that if it can't get 1000, it tries for 256.
4722 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4723 was an oversight, and furthermore, ever since the addition of extra
4724 controls (e.g. 4.43/32), the checks on when to allow different forms of
4725 "control" were broken. There should now be diagnostics for all cases when a
4726 control that does not make sense is encountered.
4728 9. Added the /retain_sender option to "control=submission".
4730 10. $recipients is now available in the predata ACL (oversight).
4732 11. Tidy the search cache before the fork to do a delivery from a message
4733 received from the command line. Otherwise the child will trigger a lookup
4734 failure and thereby defer the delivery if it tries to use (for example) a
4735 cached ldap connection that the parent has called unbind on.
4737 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4738 of $address_data from the recipient verification was clobbered by the
4739 sender verification.
4741 13. The value of address_data from a sender verification is now available in
4742 $sender_address_data in subsequent conditions in the ACL statement.
4744 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4746 15. Added a new option "connect=<time>" to callout options, to set a different
4749 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4750 was its contents. (It was OK if the option was not defined at all.)
4752 17. A "Completed" log line is now written for messages that are removed from
4753 the spool by the -Mrm option.
4755 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4756 information about exactly what failed.
4758 19. Added -dd to debug only the daemon process.
4760 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4761 handles timeouts, both on the server side and network timeouts. Renamed the
4762 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4764 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4765 It is now set to "smtps".
4767 22. $host_address is now set to the target address during the checking of
4768 ignore_target_hosts.
4770 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4771 being passed; this would have caused $sender_host_name to have been used if
4772 matching the list had actually called for a host name (not very likely,
4773 since this list is usually IP addresses). A host name is now passed as
4776 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4777 code that specifies a non-restarting handler (typically sigaction() in
4778 modern systems) in an attempt to fix a rare and obscure crash bug.
4780 25. Narrowed the window for a race in the daemon that could cause it to ignore
4781 SIGCHLD signals. This is not a major problem, because they are used only to
4782 wake it up if nothing else does.
4784 26. A malformed maildirsize file could cause Exim to calculate negative values
4785 for the mailbox size or file count. Odd effects could occur as a result.
4786 The maildirsize information is now recalculated if the size or filecount
4789 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4790 support for a long time. Removed HAVE_SYS_VFS_H.
4792 28. Installed the latest version of exipick from John Jetmore.
4794 29. In an address list, if the pattern was not a regular expression, an empty
4795 subject address (from a bounce message) matched only if the pattern was an
4796 empty string. Non-empty patterns were not even tested. This was the wrong
4797 because it is perfectly reasonable to use an empty address as part of a
4798 database query. An empty address is now tested by patterns that are
4799 lookups. However, all the other forms of pattern expect the subject to
4800 contain a local part and a domain, and therefore, for them, an empty
4801 address still always fails if the pattern is not itself empty.
4803 30. Exim went into a mad DNS loop when attempting to do a callout where the
4804 host was specified on an smtp transport, and looking it up yielded more
4805 than one IP address.
4807 31. Re-factored the code for checking spool and log partition space into a
4808 function that finds that data and another that does the check. The former
4809 is then used to implement four new variables: $spool_space, $log_space,
4810 $spool_inodes, and $log_inodes.
4812 32. The RFC2047 encoding function was originally intended for short strings
4813 such as real names; it was not keeping to the 75-character limit for
4814 encoded words that the RFC imposes. It now respects the limit, and
4815 generates multiple encoded words if necessary. To be on the safe side, I
4816 have increased the buffer size for the ${rfc2047: expansion operator from
4819 33. It is now permitted to omit both strings after an "if" condition; if the
4820 condition is true, the result is "true". As before, when the second string
4821 is omitted, a false condition yields an empty string. This makes it less
4822 cumbersome to write custom ACL and router conditions.
4824 34. Failure to deliver a bounce message always caused it to be frozen, even if
4825 there was an errors_to setting on the router. The errors_to setting is now
4828 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4829 canonical form (fully expanded) before being placed in
4830 $sender_host_address.
4832 36. The table in the code that translates DNS record types into text (T_A to
4833 "A" for instance) was missing entries for NS and CNAME. It is just possible
4834 that this could have caused confusion if both these types were looked up
4835 for the same domain, because the text type is used as part of Exim's
4836 per-process caching. But the chance of anyone hitting this buglet seems
4839 37. The dnsdb lookup has been extended in a number of ways.
4841 (1) There is a new type, "zns", which walks up the domain tree until it
4842 finds some nameserver records. It should be used with care.
4844 (2) There is a new type, "mxh", which is like "mx" except that it returns
4845 just the host names, not the priorities.
4847 (3) It is now possible to give a list of domains (or IP addresses) to be
4848 looked up. The behaviour when one of the lookups defers can be
4849 controlled by a keyword.
4851 (4) It is now possible to specify the separator character for use when
4852 multiple records are returned.
4854 38. The dnslists ACL condition has been extended: it is now possible to supply
4855 a list of IP addresses and/or domains to be looked up in a particular DNS
4858 39. Added log_selector=+queue_time_overall.
4860 40. When running the queue in the test harness, wait just a tad after forking a
4861 delivery process, to get repeatability of debugging output.
4863 41. Include certificate and key file names in error message when GnuTLS fails
4864 to set them up, because the GnuTLS error message doesn't include the name
4865 of the failing file when there is a problem reading it.
4867 42. Allow both -bf and -bF in the same test run.
4869 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4871 44. The "Exiscan patch" is now merged into the mainline Exim source.
4873 45. Sometimes the final signoff response after QUIT could fail to get
4874 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4875 < 0 before doing a fflush(). This bug looks as though it goes back to the
4876 introduction of TLS in release 3.20, but "sometimes" must have been rare
4877 because the tests only now provoked it.
4879 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4880 (this can affect the format of dates).
4882 47. exim_tidydb, when checking for the continued existence of a message for
4883 which it has found a message-specific retry record, was not finding
4884 messages that were in split spool directories. Consequently, it was
4885 deleting retry records that should have stayed in existence.
4887 48. Steve fixed some bugs in eximstats.
4889 49. The SPA authentication driver was not abandoning authentication and moving
4890 on to the next authenticator when an expansion was forced to fail,
4891 contradicting the general specification for all authenticators. Instead it
4892 was generating a temporary error. It now behaves as specified.
4894 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4895 (the order specifies the preference for clients). The order is now AES256,
4896 AES128, 3DES, ARCFOUR128.
4898 51. Small patch to Sieve code - explicitly set From: when generating an
4901 52. Exim crashed if a remote delivery caused a very long error message to be
4902 recorded - for instance if somebody sent an entire SpamAssassin report back
4903 as a large number of 550 error lines. This bug was coincidentally fixed by
4904 increasing the size of one of Exim's internal buffers (big_buffer) that
4905 happened as part of the Exiscan merge. However, to be on the safe side, I
4906 have made the code more robust (and fixed the comments that describe what
4909 53. Now that there can be additional text after "Completed" in log lines (if
4910 the queue_time_overall log selector is set), a one-byte patch to exigrep
4911 was needed to allow it to recognize "Completed" as not the last thing in
4914 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4915 patch that reportedly fixes this has been added. I am not expert enough to
4916 create a test for it. This is what the patch creator wrote:
4918 "I found a little strange behaviour of ldap code when working with
4919 Windows 2003 AD Domain, where users was placed in more than one
4920 Organization Units. When I tried to give exim partial DN, the exit code
4921 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4922 But simultaneously result of request was absolutely normal ldap result,
4923 so I produce this patch..."
4925 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4926 REFERENCE, so I have modified the code to exclude the patch when that macro
4929 55. Some experimental protocols are using DNS PTR records for new purposes. The
4930 keys for these records are domain names, not reversed IP addresses. The
4931 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4932 leaves it alone. Component reversal etc. now happens only for IP addresses.
4935 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4937 57. Double the size of the debug message buffer (to 2048) so that more of very
4938 long debug lines gets shown.
4940 58. The exicyclog utility now does better if the number of log files to keep
4941 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4943 59. Two changes related to the smtp_active_hostname option:
4945 (1) $smtp_active_hostname is now available as a variable.
4946 (2) The default for smtp_banner uses $smtp_active_hostname instead
4947 of $primary_hostname.
4949 60. The host_aton() function is supposed to be passed a string that is known
4950 to be a valid IP address. However, in the case of IPv6 addresses, it was
4951 not checking this. This is a hostage to fortune. Exim now panics and dies
4952 if the condition is not met. A case was found where this could be provoked
4953 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4954 components; fortuitously, this particular loophole had already been fixed
4955 by change 4.50/55 above.
4957 If there are any other similar loopholes, the new check in host_aton()
4958 itself should stop them being exploited. The report I received stated that
4959 data on the command line could provoke the exploit when Exim was running as
4960 exim, but did not say which command line option was involved. All I could
4961 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4962 running as the user.
4965 61. There was a buffer overflow vulnerability in the SPA authentication code
4966 (which came originally from the Samba project). I have added a test to the
4967 spa_base64_to_bits() function which I hope fixes it.
4970 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4971 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4973 63. The daemon start-up calls getloadavg() while still root for those OS that
4974 need the first call to be done as root, but it missed one case: when
4975 deliver_queue_load_max is set with deliver_drop_privilege. This is
4976 necessary for the benefit of the queue runner, because there is no re-exec
4977 when deliver_drop_privilege is set.
4979 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4980 This has been fixed.
4982 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4983 was in use, was not putting the data itself into the right store pool;
4984 consequently, it could be overwritten for a subsequent message in the same
4985 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4988 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4990 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4991 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4992 the function were treating the return as a boolean value, which happened to
4993 work because 0=false and not-0=true, but is not correct code.
4995 68. The host_aton() function was not handling scoped IPv6 addresses (those
4996 with, for example, "%eth0" on the end) correctly.
4998 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4999 negated items (that is, ~something) in unsigned ints. Some compilers
5000 apparently mutter when there is no cast.
5002 70. If an address verification called from an ACL failed, and did not produce a
5003 user-specific message (i.e. there was only a "system" message), nothing was
5004 put in $acl_verify_message. In this situation, it now puts the system
5007 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5008 session; change 4.31/43 added the unwanted input to the log line - except
5009 that it did not do this in the start of session case. It now does.
5011 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5012 This is wrong and can cause the other end to generate a synchronization
5013 error if it is another Exim or anything else that does the synchronization
5014 check. A QUIT command is no longer sent after a timeout.
5016 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5017 during host lookups.
5019 74. The defer_ok option of callout verification was not working if it was used
5020 when verifying addresses in header lines, that is, for this case:
5022 verify = header_sender/callout=defer_ok
5024 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5025 those file descriptors could be used for SMTP connections. If anything
5026 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5027 could be sent to the SMTP client, causing chaos. The daemon now opens
5028 stdin, stdout, and stderr to /dev/null when it puts itself into the
5031 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5032 log by default. The user can override this with suitable Perl magic.
5034 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5035 the log message when discard triggers, was not working for the DATA ACL or
5036 for the non-SMTP ACL.
5038 78. Error message wording change in sieve.c.
5040 79. If smtp_accept_max_per_host was set, the number of connections could be
5041 restricted to fewer than expected, because the daemon was trying to set up
5042 a new connection before checking whether the processes handling previous
5043 connections had finished. The check for completed processes is now done
5044 earlier. On busy systems, this bug wouldn't be noticed because something
5045 else would have woken the daemon, and it would have reaped the completed
5048 80. If a message was submitted locally by a user whose login name contained one
5049 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5050 It caused a spool format error. I have fixed the spool reading code. A
5051 related problem was that the "from" clause in the Received: line became
5052 illegal because of the space(s). It is now covered by ${quote_local_part.
5054 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5057 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5059 83. Patch from Sieve maintainer to fix unterminated string problem in
5060 "vacation" handling.
5062 84. Some minor changes to the Linux configuration files to help with other
5063 OS variants using glibc.
5065 85. One more patch for Sieve to update vacation handling to latest spec.
5068 ----------------------------------------------------
5069 See the note above about the 4.44 and 4.50 releases.
5070 ----------------------------------------------------
5076 1. Change 4.43/35 introduced a bug that caused file counts to be
5077 incorrectly computed when quota_filecount was set in an appendfile
5080 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5081 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5084 3. Give more explanation in the error message when the command for a transport
5085 filter fails to execute.
5087 4. There are several places where Exim runs a non-Exim command in a
5088 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5089 was being done only for the command run by the queryprogram router. It is
5090 now done for all such subprocesses. The other cases are: ${run, transport
5091 filters, and the commands run by the lmtp and pipe transports.
5093 5. Some older OS have a limit of 256 on the maximum number of file
5094 descriptors. Exim was using setrlimit() to set 1000 as a large value
5095 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5096 systems. I've change it so that if it can't get 1000, it tries for 256.
5098 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5099 was an oversight, and furthermore, ever since the addition of extra
5100 controls (e.g. 4.43/32), the checks on when to allow different forms of
5101 "control" were broken. There should now be diagnostics for all cases when a
5102 control that does not make sense is encountered.
5104 7. $recipients is now available in the predata ACL (oversight).
5106 8. Tidy the search cache before the fork to do a delivery from a message
5107 received from the command line. Otherwise the child will trigger a lookup
5108 failure and thereby defer the delivery if it tries to use (for example) a
5109 cached ldap connection that the parent has called unbind on.
5111 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5112 of $address_data from the recipient verification was clobbered by the
5113 sender verification.
5115 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5116 was its contents. (It was OK if the option was not defined at all.)
5118 11. A "Completed" log line is now written for messages that are removed from
5119 the spool by the -Mrm option.
5121 12. $host_address is now set to the target address during the checking of
5122 ignore_target_hosts.
5124 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5125 being passed; this would have caused $sender_host_name to have been used if
5126 matching the list had actually called for a host name (not very likely,
5127 since this list is usually IP addresses). A host name is now passed as
5130 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5131 code that specifies a non-restarting handler (typically sigaction() in
5132 modern systems) in an attempt to fix a rare and obscure crash bug.
5134 15. Narrowed the window for a race in the daemon that could cause it to ignore
5135 SIGCHLD signals. This is not a major problem, because they are used only to
5136 wake it up if nothing else does.
5138 16. A malformed maildirsize file could cause Exim to calculate negative values
5139 for the mailbox size or file count. Odd effects could occur as a result.
5140 The maildirsize information is now recalculated if the size or filecount
5143 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5144 support for a long time. Removed HAVE_SYS_VFS_H.
5146 18. Updated exipick to current release from John Jetmore.
5148 19. Allow an empty sender to be matched against a lookup in an address list.
5149 Previously the only cases considered were a regular expression, or an
5152 20. Exim went into a mad DNS lookup loop when doing a callout where the
5153 host was specified on the transport, if the DNS lookup yielded more than
5156 21. The RFC2047 encoding function was originally intended for short strings
5157 such as real names; it was not keeping to the 75-character limit for
5158 encoded words that the RFC imposes. It now respects the limit, and
5159 generates multiple encoded words if necessary. To be on the safe side, I
5160 have increased the buffer size for the ${rfc2047: expansion operator from
5163 22. Failure to deliver a bounce message always caused it to be frozen, even if
5164 there was an errors_to setting on the router. The errors_to setting is now
5167 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5168 canonical form (fully expanded) before being placed in
5169 $sender_host_address.
5171 24. Updated eximstats to version 1.33
5173 25. Include certificate and key file names in error message when GnuTLS fails
5174 to set them up, because the GnuTLS error message doesn't include the name
5175 of the failing file when there is a problem reading it.
5177 26. Expand error message when OpenSSL has problems setting up cert/key files.
5180 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5181 (this can affect the format of dates).
5183 28. exim_tidydb, when checking for the continued existence of a message for
5184 which it has found a message-specific retry record, was not finding
5185 messages that were in split spool directories. Consequently, it was
5186 deleting retry records that should have stayed in existence.
5188 29. eximstats updated to version 1.35
5189 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5190 1.35 - bugfix such that pie charts by volume are generated correctly
5192 30. The SPA authentication driver was not abandoning authentication and moving
5193 on to the next authenticator when an expansion was forced to fail,
5194 contradicting the general specification for all authenticators. Instead it
5195 was generating a temporary error. It now behaves as specified.
5197 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5198 (the order specifies the preference for clients). The order is now AES256,
5199 AES128, 3DES, ARCFOUR128.
5201 31. Small patch to Sieve code - explicitly set From: when generating an
5204 32. Exim crashed if a remote delivery caused a very long error message to be
5205 recorded - for instance if somebody sent an entire SpamAssassin report back
5206 as a large number of 550 error lines. This bug was coincidentally fixed by
5207 increasing the size of one of Exim's internal buffers (big_buffer) that
5208 happened as part of the Exiscan merge. However, to be on the safe side, I
5209 have made the code more robust (and fixed the comments that describe what
5212 33. Some experimental protocols are using DNS PTR records for new purposes. The
5213 keys for these records are domain names, not reversed IP addresses. The
5214 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5215 leaves it alone. Component reversal etc. now happens only for IP addresses.
5218 34. The host_aton() function is supposed to be passed a string that is known
5219 to be a valid IP address. However, in the case of IPv6 addresses, it was
5220 not checking this. This is a hostage to fortune. Exim now panics and dies
5221 if the condition is not met. A case was found where this could be provoked
5222 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5223 components; fortuitously, this particular loophole had already been fixed
5224 by change 4.50/55 or 4.44/33 above.
5226 If there are any other similar loopholes, the new check in host_aton()
5227 itself should stop them being exploited. The report I received stated that
5228 data on the command line could provoke the exploit when Exim was running as
5229 exim, but did not say which command line option was involved. All I could
5230 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5231 running as the user.
5234 35. There was a buffer overflow vulnerability in the SPA authentication code
5235 (which came originally from the Samba project). I have added a test to the
5236 spa_base64_to_bits() function which I hope fixes it.
5239 36. The daemon start-up calls getloadavg() while still root for those OS that
5240 need the first call to be done as root, but it missed one case: when
5241 deliver_queue_load_max is set with deliver_drop_privilege. This is
5242 necessary for the benefit of the queue runner, because there is no re-exec
5243 when deliver_drop_privilege is set.
5245 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5246 was in use, was not putting the data itself into the right store pool;
5247 consequently, it could be overwritten for a subsequent message in the same
5248 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5251 38. Sometimes the final signoff response after QUIT could fail to get
5252 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5253 < 0 before doing a fflush(). This bug looks as though it goes back to the
5254 introduction of TLS in release 3.20, but "sometimes" must have been rare
5255 because the tests only now provoked it.
5261 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5262 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5263 Now it returns an integer. A number of calls were still expecting a T/F
5264 return. Fortuitously, in all cases, the tests worked in OK situations,
5265 which is the norm. However, things would have gone wrong on any write
5266 failures on the smtp file descriptor. This function is used when sending
5267 messages over SMTP and also when doing verify callouts.
5269 2. When Exim is called to do synchronous delivery of a locally submitted
5270 message (the -odf or -odi options), it no longer closes stderr before doing
5273 3. Implemented the mua_wrapper option.
5275 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5277 5. Implemented the functions header_remove(), header_testname(),
5278 header_add_at_position(), and receive_remove_recipient(), and exported them
5281 6. If an ACL "warn" statement specified the addition of headers, Exim already
5282 inserted X-ACL-Warn: at the start if there was no header name. However, it
5283 was not making this test for the second and subsequent header lines if
5284 there were newlines in the string. This meant that an invalid header could
5285 be inserted if Exim was badly configured.
5287 7. Allow an ACL "warn" statement to add header lines at the start or after all
5288 the Received: headers, as well as at the end.
5290 8. Added the rcpt_4xx retry error code.
5292 9. Added postmaster_mailfrom=xxx to callout verification option.
5294 10. Added mailfrom=xxxx to the callout verification option, for verify=
5297 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5298 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5300 12. Inserted some casts to stop certain compilers warning when using pointer
5301 differences as field lengths or precisions in printf-type calls (mostly
5302 affecting debugging statements).
5304 13. Added optional readline() support for -be (dynamically loaded).
5306 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5307 same clock tick as a message's arrival, so that its received time was the
5308 same as the "first fail" time on the retry record, and that message
5309 remained on the queue past the ultimate address timeout, every queue runner
5310 would try a delivery (because it was past the ultimate address timeout) but
5311 after another failure, the ultimate address timeout, which should have then
5312 bounced the address, did not kick in. This was a "< instead of <=" error;
5313 in most cases the first failure would have been in the next clock tick
5314 after the received time, and all would be well.
5316 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5317 being recognized when the domain list was tested by the match_domain
5318 condition in an expansion string.
5320 16. Added the ${str2b64: operator.
5322 17. Exim was always calling setrlimit() to set a large limit for the number of
5323 processes, without checking whether the existing limit was already
5324 adequate. (It did check for the limit on file descriptors.) Furthermore,
5325 errors from getrlimit() and setrlimit() were being ignored. Now they are
5326 logged to the main and panic logs, but Exim does carry on, to try to do its
5327 job under whatever limits there are.
5329 18. Imported PCRE 5.0.
5331 19. Trivial typo in log message " temporarily refused connection" (the leading
5334 20. If the log selector return_path_on_delivery was set and an address was
5335 redirected to /dev/null, the delivery process crashed because it assumed
5336 that a return path would always be set for a "successful" delivery. In this
5337 case, the whole delivery is bypassed as an optimization, and therefore no
5340 21. Internal re-arrangement: the function for sending a challenge and reading
5341 a response while authentication was assuming a zero-terminated challenge
5342 string. It's now changed to take a pointer and a length, to allow for
5343 binary data in such strings.
5345 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5347 23. Exim was not respecting finduser_retries when seeking the login of the
5348 uid under which it was called; it was always trying 10 times. (The default
5349 setting of finduser_retries is zero.) Also, it was sleeping after the final
5350 failure, which is pointless.
5352 24. Implemented tls_on_connect_ports.
5354 25. Implemented acl_smtp_predata.
5356 26. If the domain in control=submission is set empty, Exim assumes that the
5357 authenticated id is a complete email address when it generates From: or
5358 Sender: header lines.
5360 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5361 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5362 chown and chgrp in /bin and hostname in /usr/bin.
5364 28. Exim was keeping the "process log" file open after each use, just as it
5365 does for the main log. This opens the possibility of it remaining open for
5366 long periods when the USR1 signal hits a daemon. Occasional processlog
5367 errors were reported, that could have been caused by this. Anyway, it seems
5368 much more sensible not to leave this file open at all, so that is what now
5371 29. The long-running daemon process does not normally write to the log once it
5372 has entered its main loop, and it closes the log before doing so. This is
5373 so that log files can straightforwardly be renamed and moved. However,
5374 there are a couple of unusual error situations where the daemon does write
5375 log entries, and I had neglected to close the log afterwards.
5377 30. The text of an SMTP error response that was received during a remote
5378 delivery was being truncated at 512 bytes. This is too short for some of
5379 the long messages that one sometimes sees. I've increased the limit to
5382 31. It is now possible to make retry rules that apply only when a message has a
5383 specific sender, in particular, an empty sender.
5385 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5386 it possible to be selective about when SMTP synchronization is enforced.
5388 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5390 32. Implemented hosts_connection_nolog.
5392 33. Added an ACL for QUIT.
5394 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5397 35. Added mailbox_size and mailbox_filecount to appendfile.
5399 36. Added control = no_multiline_responses to ACLs.
5401 37. There was a bug in the logic of the code that waits for the clock to tick
5402 in the case where the clock went backwards by a substantial amount such
5403 that the microsecond fraction of "now" was more than the microsecond
5404 fraction of "then" (but the whole seconds number was less).
5406 38. Added support for the libradius Radius client library this is found on
5407 FreeBSD (previously only the radiusclient library was supported).
5413 1. When certain lookups returned multiple values in the form name=value, the
5414 quoting of the values was not always being done properly. Specifically:
5415 (a) If the value started with a double quote, but contained no whitespace,
5417 (b) If the value contained whitespace other than a space character (i.e.
5418 tabs or newlines or carriage returns) it was not quoted.
5419 This fix has been applied to the mysql and pgsql lookups by writing a
5420 separate quoting function and calling it from the lookup code. The fix
5421 should probably also be applied to nisplus, ibase and oracle lookups, but
5422 since I cannot test any of those, I have not disturbed their existing code.
5424 2. A hit in the callout cache for a specific address caused a log line with no
5425 reason for rejecting RCPT. Now it says "Previous (cached) callout
5426 verification failure".
5428 3. There was an off-by-one bug in the queryprogram router. An over-long
5429 return line was truncated at 256 instead of 255 characters, thereby
5430 overflowing its buffer with the terminating zero. As well as fixing this, I
5431 have increased the buffer size to 1024 (and made a note to document this).
5433 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5434 when Exim is waiting for an SMTP response from a remote server, Exim
5435 restarts its select() call on the socket, thereby resetting its timeout.
5436 This is not a problem when such interrupts are rare. Somebody set up a cron
5437 job to run exiwhat every 2 minutes, which is less than the normal select()
5438 timeout (5 or 10 minutes). This meant that the select() timeout never
5439 kicked in because it was always reset. I have fixed this by comparing the
5440 time when an interrupt arrives with the time at the start of the first call
5441 to select(). If more time than the timeout has elapsed, the interrupt is
5442 treated as a timeout.
5444 5. Some internal re-factoring in preparation for the addition of Sieve
5445 extensions (by MH). In particular, the "personal" test is moved to a
5446 separate function, and given an option for scanning Cc: and Bcc: (which is
5447 not set for Exim filters).
5449 6. When Exim created an email address using the login of the caller as the
5450 local part (e.g. when creating a From: or Sender: header line), it was not
5451 quoting the local part when it contained special characters such as @.
5453 7. Installed new OpenBSD configuration files.
5455 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5456 try to make them clearer.
5458 9. Callout options, other than the timeout value, were being ignored when
5459 verifying sender addresses in header lines. For example, when using
5461 verify = header_sender/callout=no_cache
5463 the cache was (incorrectly) being used.
5465 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5466 only the Cygwin environment.
5468 11. When return_path_on_delivery was set as a log selector, if different remote
5469 addresses in the same message used different return paths and parallel
5470 remote delivery occurred, the wrong values would sometimes be logged.
5471 (Whenever a remote delivery process finished, the return path value from
5472 the most recently started remote delivery process was logged.)
5474 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5475 lines when AUTH and/or TLS are in use. This is the "received protocol"
5476 field. Exim used to use "asmtp" for authenticated SMTP, without any
5477 indication (in the protocol name) for TLS use. Now it follows the RFC and
5478 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5479 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5480 names appear in log lines as well as in Received: header lines.
5482 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5483 extensions, and comparison tests, and to fix some bugs.
5485 14. Changes to the "personal" filter test:
5487 (1) The test was buggy in that it was just doing the equivalent of
5488 "contains" tests on header lines. For example, if a user's address was
5489 anne@some.where, the "personal" test would incorrectly be true for
5491 To: susanne@some.where
5493 This test is now done by extracting each address from the header in turn,
5494 and checking the entire address. Other tests that are part of "personal"
5495 are now done using regular expressions (for example, to check local parts
5496 of addresses in From: header lines).
5498 (2) The list of non-personal local parts in From: addresses has been
5499 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5500 taken from the Sieve specification recommendations.
5502 (3) If the message contains any header line starting with "List-" it is
5503 treated as non-personal.
5505 (4) The test for "circular" in the Subject: header line has been removed
5506 because it now seems ill-conceived.
5508 15. Minor typos in src/EDITME comments corrected.
5510 16. Installed latest exipick from John Jetmore.
5512 17. If headers_add on a router specified a text string that was too long for
5513 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5514 of string_sprintf() is now avoided.
5516 18. $message_body_size was not set (it was always zero) when running the DATA
5517 ACL and the local_scan() function.
5519 19. For the "mail" command in an Exim filter, no default was being set for
5520 the once_repeat time, causing a random time value to be used if "once" was
5521 specified. (If the value happened to be <= 0, no repeat happened.) The
5522 default is now 0s, meaning "never repeat". The "vacation" command was OK
5523 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5524 (I found it when inspecting the code).
5526 20. There is now an overall timeout for performing a callout verification. It
5527 defaults to 4 times the callout timeout, which applies to individual SMTP
5528 commands during the callout. The overall timeout applies when there is more
5529 than one host that can be tried. The timeout is checked before trying the
5530 next host. This prevents very long delays if there are a large number of
5531 hosts and all are timing out (e.g. when the network connections are timing
5532 out). The value of the overall timeout can be changed by specifying an
5533 additional sub-option for "callout", called "maxwait". For example:
5535 verify = sender/callout=5s,maxwait=20s
5537 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5538 to the end before writing, but this should make it even safer).
5540 22. Exim was forgetting that it had advertised PIPELINING for the second and
5541 subsequent messages on an SMTP connection. It was also not resetting its
5542 memory on STARTTLS and an internal HELO.
5544 23. When Exim logs an SMTP synchronization error within a session, it now
5545 records whether PIPELINING has been advertised or not.
5547 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5548 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5549 rather than long int.
5551 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5553 26. Added the never_mail option to autoreply.
5559 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5560 crash if the getsockname() call failed; this can happen if a connection is
5561 closed very soon after it is established. The problem was simply in the
5562 order in which certain operations were done, causing Exim to try to write
5563 to the SMTP stream before it had set up the file descriptor. The bug has
5564 been fixed by making things happen in the correct order.
5570 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5571 before the connection was closed, thus losing the rejection response.
5573 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5574 some early Solaris releases, but causes trouble in current releases where
5575 socklen_t is defined.
5577 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5580 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5583 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5584 that releasing the top of it at the end releases what was used for sub-
5585 expansions (unless the block got too big). However, discard this block if
5586 the first thing is a variable or header, so that we can use its block when
5587 it is dynamic (useful for very large $message_headers, for example).
5589 6. Lookups now cache *every* query, not just the most recent. A new, separate
5590 store pool is used for this. It can be recovered when all lookup caches are
5591 flushed. Lookups now release memory at the end of their result strings.
5592 This has involved some general refactoring of the lookup sources.
5594 7. Some code has been added to the store_xxx() functions to reduce the amount
5595 of flapping under certain conditions.
5597 8. log_incoming_interface used to affect only the <= reception log lines. Now
5598 it causes the local interface and port to be added to several more SMTP log
5599 lines, for example "SMTP connection from", and rejection lines.
5601 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5603 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5605 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5606 could be overwritten at the end of the current message (or the start of a
5607 new message if it was set in a HELO ACL). The value is now preserved for
5608 the duration of the SMTP connection.
5610 12. If a transport had a headers_rewrite setting, and a matching header line
5611 contained an unqualified address, that address was qualified, even if it
5612 did not match any rewriting rules. The underlying bug was that the values
5613 of the flags that permit the existence of unqualified sender and recipient
5614 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5615 non-local messages, and by -bnq for local messages) were not being
5616 preserved with the message after it was received.
5618 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5619 "next input=" as part of the text comprising the host identity instead of
5620 the correct text. The code was using the same buffer for two different
5621 strings. However, depending on which order the printing function evaluated
5622 its arguments, the bug did not always show up. Under Linux, for example, my
5623 test suite worked just fine.
5625 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5626 doesn't work with some older versions of Perl. It has been changed to "my",
5627 which in any case is probably the better facility to use.
5629 15. A really picky compiler found some instances of statements for creating
5630 error messages that either had too many or two few arguments for the format
5633 16. The size of the buffer for calls to the DNS resolver has been increased
5634 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5635 for addresses that have a lot of PTR records. This alleviates a problem; it
5636 does not fully solve it.
5638 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5639 buffer now truncates the list and logs the incident, which is the same
5640 action as happens when Exim is looking up a host name and its aliases.
5641 Previously in this situation something unpredictable would happen;
5642 sometimes it was "internal error: store_reset failed".
5644 18. If a server dropped the connection unexpectedly when an Exim client was
5645 using GnuTLS and trying to read a response, the client delivery process
5646 crashed while trying to generate an error log message.
5648 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5649 string, for example:
5651 warn message = H1: something\nH2: something
5653 the text was added as a single header line from Exim's point of view
5654 though it ended up OK in the delivered message. However, searching for the
5655 second and subsequent header lines using $h_h2: did not work. This has been
5656 fixed. Similarly, if a system filter added multiple headers in this way,
5657 the routers could not see them.
5659 20. Expanded the error message when iplsearch is called with an invalid key to
5660 suggest using net-iplsearch in a host list.
5662 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5663 ACLs are no longer actually imposed (and a message to that effect is
5666 22. If a "gecos" field in a passwd entry contained escaped characters, in
5667 particular, if it contained a \" sequence, Exim got it wrong when building
5668 a From: or a Sender: header from that name. A second bug also caused
5669 incorrect handling when an unquoted " was present following a character
5670 that needed quoting.
5672 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5673 was not being matched caselessly.
5675 24. Arranged for all hyphens in the exim.8 source to be escaped with
5678 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5679 properly. Recipient callout cache records were still being keyed to include
5680 the sender, even when use_sender was set false. This led to far more
5681 callouts that were necessary. The sender is no longer included in the key
5682 when use_sender is false.
5684 26. Added "control = submission" modifier to ACLs.
5686 27. Added the ${base62d: operator to decode base 62 numbers.
5688 28. dnsdb lookups can now access SRV records.
5690 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5691 the configuration file.
5693 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5694 (-v) mode. This makes the output for a verbose queue run more intelligible.
5696 31. Added a use_postmaster feature to recipient callouts.
5698 32. Added the $body_zerocount variable, containing the number of binary zero
5699 bytes in the message body.
5701 33. The time of last modification of the "new" subdirectory is now used as the
5702 "mailbox time last read" when there is a quota error for a maildir
5705 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5707 35. Added +ignore_unknown as a special item in host lists.
5709 36. Code for decoding IPv6 addresses in host lists is now included, even if
5710 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5711 address was recognized as an IP address, but was then not correctly decoded
5712 into binary, causing unexpected and incorrect effects when compared with
5719 1. Very minor rewording of debugging text in manualroute to say "list of
5720 hosts" instead of "hostlist".
5722 2. If verify=header_syntax was set, and a header line with an unqualified
5723 address (no domain) and a large number of spaces between the end of the
5724 name and the colon was received, the reception process suffered a buffer
5725 overflow, and (when I tested it) crashed. This was caused by some obsolete
5726 code that should have been removed. The fix is to remove it!
5728 3. When running in the test harness, delay a bit after writing a bounce
5729 message to get a bit more predictability in the log output.
5731 4. Added a call to search_tidyup() just before forking a reception process. In
5732 theory, someone could use a lookup in the expansion of smtp_accept_max_
5733 per_host which, without the tidyup, could leave open a database connection.
5735 5. Added the variables $recipient_data and $sender_data which get set from a
5736 lookup success in an ACL "recipients" or "senders" condition, or a router
5737 "senders" option, similar to $domain_data and $local_part_data.
5739 6. Moved the writing of debug_print from before to after the "senders" test
5742 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5743 problems for message scanning, either using a data ACL, or using
5744 local_scan() because the Received: header was not generated till after they
5745 were called (in order to set the time as the time of reception completion).
5746 I have revised the way this works. The header is now generated after the
5747 body is received, but before the ACL or local_scan() are called. After they
5748 are run, the timestamp in the header is updated.
5754 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5755 before starting a queue runner without re-exec. This happened only when
5756 deliver_drop_privilege was set or when the Exim user was set to root. The
5757 effect of the bug was that timeouts during subsequent deliveries caused
5758 crashes instead of being properly handled. The handler is now left at its
5759 default (and expected) setting.
5761 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5762 message, again when deliver_drop_privilege is set or Exim is run as root.
5763 The bug described in (1) was not present in this case, but the tidying up
5764 of the other signals was missing. I have made the two cases consistent.
5766 3. The ignore_target_hosts setting on a manualroute router was being ignored
5767 for hosts that were looked up using the /MX notation.
5769 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5772 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5773 operated on the sender address. After changing the $sender_address to <>
5774 for the sender address verify, Exim was re-instated it as the original
5775 (before rewriting) address, but remembering that it had rewritten it, so it
5776 wasn't rewriting it again. This bug also had the effect of breaking the
5777 sender address verification caching when the sender address was rewritten.
5779 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5780 This has been changed so that if the ip literal address matches
5781 ignore_target_hosts, the router declines.
5783 7. Added expansion conditions match_domain, match_address, and match_local_
5784 part (NOT match_host).
5786 8. The placeholder for the Received: header didn't have a length field set.
5788 9. Added code to Exim itself and to exim_lock to test for a specific race
5789 condition that could lead to file corruption when using MBX delivery. The
5790 issue is with the lockfile that is created in /tmp. If this file is removed
5791 after a process has opened it but before that process has acquired a lock,
5792 there is the potential for a second process to recreate the file and also
5793 acquire a lock. This could lead to two Exim processes writing to the file
5794 at the same time. The added code performs the same test as UW imapd; it
5795 checks after acquiring the lock that its file descriptor still refers to
5796 the same named file.
5798 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5799 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5802 11. Added the smtp_active_hostname option. If used, this will typically be made
5803 to depend on the incoming interface address. Because $interface_address is
5804 not set up until the daemon has forked a reception process, error responses
5805 that can happen earlier (such as "too many connections") no longer contain
5808 12. If an expansion in a condition on a "warn" statement fails because a lookup
5809 defers, the "warn" statement is abandoned, and the next ACL statement is
5810 processed. Previously this caused the whole ACL to be aborted.
5812 13. Added the iplsearch lookup type.
5814 14. Added ident_timeout as a log selector.
5816 15. Added tls_certificate_verified as a log selector.
5818 16. Added a global option tls_require_ciphers (compare the smtp transport
5819 option of the same name). This controls incoming TLS connections.
5821 17. I finally figured out how to make tls_require_ciphers do a similar thing
5822 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5823 before starting the TLS session.
5825 18. Tabs are now shown as \t in -bP output.
5827 19. If the log selector return_path_on_delivery was set, Exim crashed when
5828 bouncing a message because it had too many Received: header lines.
5830 20. If two routers both had headers_remove settings, and the first one included
5831 a superfluous trailing colon, the final name in the first list and the
5832 first name in the second list were incorrectly joined into one item (with a
5833 colon in the middle).
5839 1. Added -C and -D options to the exinext utility, mainly to make it easier
5840 to include in the automated testing, but these could be helpful when
5841 multiple configurations are in use.
5843 2. The exinext utility was not formatting the output nicely when there was
5844 an alternate port involved in the retry record key, nor when there was a
5845 message id as well (for retries that were specific to a specific message
5846 and a specific host). It was also confused by IPv6 addresses, because of
5847 the additional colons they contain. I have fixed the IPv4 problem, and
5848 patched it up to do a reasonable job for IPv6.
5850 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5851 delivery, the log line now contains "pipelined" if PIPELINING was used.
5853 4. An SMTP transport process used to panic and die if the bind() call to set
5854 an explicit outgoing interface failed. This has been changed; it is now
5855 treated in the same way as a connect() failure.
5857 5. A reference to $sender_host_name in the part of a conditional expansion
5858 that was being skipped was still causing a DNS lookup. This no longer
5861 6. The def: expansion condition was not recognizing references to header lines
5862 that used bh_ and bheader_.
5864 7. Added the _cache feature to named lists.
5866 8. The code for checking quota_filecount in the appendfile transport was
5867 allowing one more file than it should have been.
5869 9. For compatibility with Sendmail, the command line option
5877 and sets the incoming protocol and host name (for trusted callers). The
5878 host name and its colon can be omitted when only the protocol is to be set.
5879 Note the Exim already has two private options, -pd and -ps, that refer to
5880 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5881 "s", but I don't think that's a major issue.
5883 10. A number of refactoring changes to the code, none of which should affect
5886 (a) The number of logging options was getting close to filling up the
5887 32-bit word that was used as a bit map. I have split them into two classes:
5888 those that are passed in the argument to log_write(), and those that are
5889 only ever tested independently outside of that function. These are now in
5890 separate 32-bit words, so there is plenty of room for expansion again.
5891 There is no change in the user interface or the logging behaviour.
5893 (b) When building, for example, log lines, the code previously used a
5894 macro that called string_cat() twice, in order to add two strings. This is
5895 not really sufficiently general. Furthermore, there was one instance where
5896 it was actually wrong because one of the argument was used twice, and in
5897 one call a function was used. (As it happened, calling the function twice
5898 did not affect the overall behaviour.) The macro has been replaced by a
5899 function that can join an arbitrary number of extra strings onto a growing
5902 (c) The code for expansion conditions now uses a table and a binary chop
5903 instead of a serial search (which was left over from when there were very
5904 few conditions). Also, it now recognizes conditions like "pam" even when
5905 the relevant support is not compiled in: a suitably worded error message is
5906 given if an attempt is made to use such a condition.
5908 11. Added ${time_interval:xxxxx}.
5910 12. A bug was causing one of the ddress fields not to be passed back correctly
5911 from remote delivery subprocesses. The field in question was not being
5912 subsequently used, so this caused to problems in practice.
5914 13. Added new log selectors queue_time and deliver_time.
5916 14. Might have fixed a bug in maildirsizefile handling that threw up
5917 "unexpected character" debug warnings, and recalculated the data
5918 unnecessarily. In any case, I expanded the warning message to give more
5921 15. Added the message "Restricted characters in address" to the statements in
5922 the default ACL that block characters like @ and % in local parts.
5924 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5925 Three changes have been made:
5927 (a) There was a serious bug; a negative response to MAIL caused the whole
5928 recipient domain to be cached as invalid, thereby blocking all messages
5929 to all local parts at the same domain, from all senders. This bug has
5930 been fixed. The domain is no longer cached after a negative response to
5931 MAIL if the sender used is not empty.
5933 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5936 (c) A new callout option, "use_sender" has been added for people who want
5937 the modified behaviour.
5943 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5946 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5947 indeed breaks things for older releases.
5949 3. Added additional logging to the case where there is a problem reading data
5950 from a filter that is running in a subprocess using a pipe, in order to
5951 try to track down a specific problem.
5953 4. Testing facility fudge: when running in the test harness and attempting
5954 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5955 getting "No route to host". Convert this to a timeout.
5957 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5960 6. Some OS don't have socklen_t but use size_t instead. This affects the
5961 fifth argument of getsockopt() amongst other things. This is now
5962 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5963 can be set for individual OS. I have set it for SunOS5, OSF1, and
5964 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5965 some earlier ones do not.
5967 7. Change 4.30/15 was not doing the test caselessly.
5969 8. The standard form for an IPv6 address literal was being rejected by address
5970 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5971 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5972 this, as well as the form without the "IPv6" on the front (but only when
5973 address literals are enabled, of course).
5975 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5977 10. Exim crashed if a message with an empty sender address specified by -f
5978 encountered a router with an errors_to setting. This could be provoked only
5979 by a command such as
5983 where an empty string was supplied; "<>" did not hit this bug.
5985 11. Installed PCRE release 4.5.
5987 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5988 remained set. It is now erased.
5990 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5991 times from message ids (which are base 36 rather than the normal 62).
5993 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5994 were being counted as actual protocol errors, and logged if the log
5995 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5996 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5997 and DATA following a set of rejected RCPTs do not count as protocol errors.
5998 In other words, Exim assumes they were pipelined, though this may not
5999 actually be the case. Of course, in all cases the client gets an
6000 appropriate error code.
6002 15. If a lookup fails in an ACL condition, a message about the failure may
6003 be available; it is used if testing the ACL cannot continue, because most
6004 such messages specify what the cause of the deferral is. However, some
6005 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6006 that caused an old message to be retained and used if a later statement
6007 caused a defer, replacing the real cause of the deferral.
6009 16. If an IP address had so many PTR records that the DNS lookup buffer
6010 was not large enough to hold them, Exim could crash while trying to process
6011 the truncated data. It now detects and logs this case.
6013 17. Further to 4.21/58, another change has been made: if (and only if) the
6014 first line of a message (the first header line) ends with CRLF, a bare LF
6015 in a subsequent header line has a space inserted after it, so as not to
6016 terminate the header.
6018 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6019 unnecessarily, used atoi() instead of strtol(), and didn't check the
6020 termination when getting file sizes from file names by regex.
6022 19. Completely re-implemented the support for maildirsize files, in the light
6023 of a number of problems with the previous contributed implementation
6024 (4.30/29). In particular:
6026 . If the quota is zero, the maildirsize file is maintained, but no quota is
6029 . If the maildir directory does not exist, it is created before any attempt
6030 to write a maildirsize file.
6032 . The quota value in the file is just a cache; if the quota is changed in
6033 the transport, the new value overrides.
6035 . A regular expression is available for excluding directories from the
6038 20. The autoreply transport checks the characters in options that define the
6039 message's headers; it allows continued headers, but it was checking with
6040 isspace() after an embedded newline instead of explicitly looking for a
6043 21. If all the "regular" hosts to which an address was routed had passed their
6044 expiry times, and had not reached their retry times, the address was
6045 bounced, even if fallback hosts were defined. Now Exim should go on to try
6048 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6049 equivalent code in the SMTP transport. Some hosts send humungous responses
6050 to HELO/EHLO, more than 1024 it seems.
6052 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6053 gives compiler warnings in some environments. I've now done it "properly",
6056 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6057 (because of problems with the built-in one) was declared to return uschar *
6058 instead of char *, causing compiler failure.
6060 25. Fixed a file descriptor leak when processing alias/forward files.
6062 26. Fixed a minor format string issue in dbfn.c.
6064 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6066 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6067 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6070 29. When a sender address is verified, it is cached, to save repeating the test
6071 when there is more than one recipient in a message. However, when the
6072 verification involves a callout, it is possible for different callout
6073 options to be set for different recipients. It is too complicated to keep
6074 track of this in the cache, so now Exim always runs a verification when a
6075 callout is required, relying on the callout cache for the optimization.
6076 The overhead is duplication of the address routing, but this should not be
6079 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6080 to be verified with callout=postmaster, and the main callout worked but the
6081 postmaster check failed, the verification correctly failed. However, if a
6082 subsequent RCPT command asked for sender verification *without* the
6083 postmaster check, incorrect caching caused this verification also to fail,
6086 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6087 it was not caching the DNS options (qualify_single, search_parents) that
6088 were used when the lookup failed. A subsequent lookup with different
6089 options therefore always gave the same answer, though there were cases
6090 where it should not have. (Example: a "domains = !$mx_any" option on a
6091 dnslookup router: the "domains" option is always processed without any
6092 widening, but the router might have qualify_single set.) Now Exim uses the
6093 cached value only when the same options are set.
6095 32. Added John Jetmore's "exipick" utility to the distribution.
6097 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6098 than a timeout (e.g. a certificate is required, and is not provided), an
6099 Exim server now closes the connection immediately. Previously it waited for
6100 the client to close - but if the client is SSL, it seems that they each
6101 wait for each other, leading to a delay before one of them times out.
6103 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6104 maintained 0.8.x compatibility because I don't think many are using it, and
6105 it is clearly obsolete.
6107 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6110 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6111 client certificate was expired. A simple patch fixes this, though I don't
6112 understand the full logic of why the verify callback is called multiple
6115 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6116 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6117 which causes problems with some clients (such as the Certicom SSL Plus
6118 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6119 disables the coutermeasure allowing Eudora to connect."
6121 38. Exim was not checking that a write() to a log file succeeded. This could
6122 lead to Bad Things if a log got too big, in particular if it hit a file
6123 size limit. Exim now panics and dies if it cannot write to a log file, just
6124 as it does if it cannot open a log file.
6126 39. Modified OS/Makefile-Linux so that it now contains
6128 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6130 The two -D definitions ensure that Exim is compiled with large file
6131 support, which makes it possible to handle log files that are bigger than
6134 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6135 instance) a domain was checked against a named list that involved a lookup,
6136 causing $domain_data to be set, then another domain was checked against the
6137 same list, then the first domain was re-checked, the value of $domain_data
6138 after the final check could be wrong. In particular, if the second check
6139 failed, it could be set empty. This bug probably also applied to
6142 41. The strip_trailing_dot option was not being applied to the address given
6143 with the -f command-line option.
6145 42. The code for reading a message's header from the spool was incrementing
6146 $received_count, but never initializing it. This meant that the value was
6147 incorrect (doubled) while delivering a message in the same process in which
6148 it was received. In the most common configuration of Exim, this never
6149 happens - a fresh exec is done - but it can happen when
6150 deliver_drop_privilege is set.
6152 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6153 it now includes up to 150 characters of the unexpected data in the log
6156 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6157 and building data strings. The size of both of these buffers was 10 000
6158 bytes - far larger than anybody would *ever* want, thought I. Needless to
6159 say, somebody hit the limit. I have increased the maximum line length to
6160 20 000 and the maximum data length of concatenated lines to 100 000. I have
6161 also fixed two bugs, because there was no checking on these buffers. Tsk,
6162 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6163 buffer is too small.
6165 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6166 lsearch lookups. Now it does.
6168 46. When parsing a route_list item in a manualroute router, a fixed-length
6169 buffer was used for the list of hosts. I made this 1024 bytes long,
6170 thinking that nobody would ever have a list of hosts that long. Wrong.
6171 Somebody had a whole pile of complicated expansion conditions, and the
6172 string was silently truncated, leading to an expansion error. It turns out
6173 that it is easier to change to an unlimited length (owing to other changes
6174 that have happened since this code was originally written) than to build
6175 structure for giving a limitation error. The length of the item that
6176 expands into the list of hosts is now unlimited.
6178 47. The lsearch lookup could not handle data where the length of text line was
6179 more than 4095 characters. Such lines were truncated, leading to shortened
6180 data being returned. It should now handle lines of any length.
6182 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6183 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6186 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6187 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6188 no longer happens in comment lines. A list of replacements is now placed
6189 at the head of all of the source files, except those whose only change is
6190 to replace PERL_COMMAND in the very first #! line.
6192 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6193 messages on the queue, with a bottom-up merge sort, using code contributed
6194 by Michael Haardt. This should make operations like -bp somewhat faster on
6195 large queues. It won't affect queue runners, except when queue_run_in_order
6198 51. Installed eximstats 1.31 in the distribution.
6200 52. Added support for SRV lookups to the dnslookup router.
6202 53. If an ACL referred to $message_body or $message_body_end, the value was not
6203 reset for any messages that followed in the same SMTP session.
6205 54. The store-handling optimization for building very long strings was not
6206 differentiating between the different store pools. I don't think this
6207 actually made any difference in practice, but I've tidied it.
6209 55. While running the routers to verify a sender address, $sender_address
6210 was still set to the sender address. This is wrong, because when routing to
6211 send a bounce to the sender, it would be empty. Therefore, I have changed
6212 it so that, while verifying a sender address, $sender_address is set to <>.
6213 (There is no change to what happens when verifying a recipient address.)
6215 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6216 target A or AAAA records (if not already returned) without resetting the
6217 qualify_single or search_parents options of the DNS resolver. These are
6218 inappropriate in this case because the targets of MX and SRV records must
6219 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6220 target that, when qualified, matched something in the local domain. These
6221 two options are now turned off when doing these lookups.
6223 57. It seems that at least some releases of Reiserfs (which does not have the
6224 concept of a fixed number of inodes) returns zero and not -1 for the
6225 number of available inodes. This interacted badly with check_spool_inodes,
6226 which assumed that -1 was the "no such thing" setting. What I have done is
6227 to check that the total number of inodes is greater than zero before doing
6228 the test of how many are available.
6230 58. When a "warn" ACL statement has a log_message modifier, the message is
6231 remembered, and not repeated. This is to avoid a lot of repetition when a
6232 message has many recipients that cause the same warning to be written.
6233 However, Exim was preserving the list of already written lines for an
6234 entire SMTP session, which doesn't seem right. The memory is now reset if a
6235 new message is started.
6237 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6238 debugging output unless log_rewrite was also set.
6240 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6241 of (int)(handle) when we know that handle contains (void *)(-1).
6243 61. The Exim daemon panic-logs an error return when it closes the incoming
6244 connection. However "connection reset by peer" seems to be common, and
6245 isn't really an error worthy of noting specially, so that particular error
6248 62. When Exim is trying to find all the local interfaces, it used to panic and
6249 die if the ioctl to get the interface flags failed. However, it seems that
6250 on at least one OS (Solaris 9) it is possible to have an interface that is
6251 included in the list of interfaces, but for which you get a failure error
6252 for this call. This happens when the interface is not "plumbed" into a
6253 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6254 failure of the "get flags" call assumes that the interface is down.
6256 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6257 makes life easier for people who are doing arithmetic on fields extracted
6258 from dates, where you often get leading zeros that should not be
6259 interpreted as octal.
6261 64. Added qualify_domain to the redirect router, to override the global
6264 65. If a pathologically long header line contained very many addresses (the
6265 report of this problem mentioned 10 000) and each of them was rewritten,
6266 Exim could use up a very large amount of memory. (It kept on making new
6267 copies of the header line as it rewrote, and never released the old ones.)
6268 At the expense of a bit more processing, the header rewriting function has
6269 been changed so that it no longer eats memory in this way.
6271 66. The generation of the Received: header has been moved from the time that a
6272 message starts to be received, to the time that it finishes. The timestamp
6273 in the Received: header should now be very close to that of the <= log
6274 line. There are two side-effects of this change:
6276 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6277 logged header lines no longer include the local Received: line, because
6278 it has not yet been created. The same applies to a copy of the message
6279 that is returned to a non-SMTP sender when a message is rejected.
6281 (b) When a filter file is tested using -bf, no additional Received: header
6282 is added to the test message. After some thought, I decided that this
6285 This change does not affect the value of $received_for. It is still set
6286 after address rewriting, but before local_scan() is called.
6288 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6290 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6291 gave an unhelpful panic error message, and a defer error. I have managed to
6292 change this behaviour so that it now rejects any supplied certificate,
6293 which seems right, as the list of acceptable certificates is empty.
6295 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6296 gave an unhelpful defer error. I have not managed to make this reject any
6297 supplied certificates, but the error message it gives is "no certificate
6298 supplied", which is not helpful.
6300 70. exigrep's output now also includes lines that are not associated with any
6301 message, but which match the given pattern. Implemented by a patch from
6302 Martin Sluka, which also tidied up the Perl a bit.
6304 71. Recipient callout verification, like sender verification, was using <> in
6305 the MAIL FROM command. This isn't really the right thing, since the actual
6306 sender may affect whether the remote host accepts the recipient or not. I
6307 have changed it to use the actual sender in the callout; this means that
6308 the cache record is now keyed on a recipient/sender pair, not just the
6309 recipient address. There doesn't seem to be a real danger of callout loops,
6310 since a callout by the remote host to check the sender would use <>.
6311 [SEE ABOVE: changed after hitting problems.]
6313 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6314 temporary errors. However, in the case of such a code being given after
6315 the end of a data transmission (i.e. after ".") Exim was failing to write
6316 a retry record for the message. (Yes, there was some broken host that was
6317 actually sending 8xx at this point.)
6319 73. An unknown lookup type in a host list could cause Exim to panic-die when
6320 the list was checked. (An example that provoked this was putting <; in the
6321 middle of a list instead of at the start.) If this happened during a DATA
6322 ACL check, a -D file could be left lying around. This kind of configuration
6323 error no longer causes Exim to die; instead it causes a defer error. The
6324 incident is still logged to the main and panic logs.
6326 74. Buglet left over from Exim 3 conversion. The message "too many messages
6327 in one connection" was written to the rejectlog but not the mainlog, except
6328 when address rewriting (yes!) was being logged.
6330 75. Added write_rejectlog option.
6332 76. When a system filter was run not as root (that is, when system_filter_user
6333 was set), the values of the $n variables were not being returned to the
6334 main process; thus, they were not subsequently available in the $sn
6337 77. Added +return_path_on_delivery log selector.
6339 78. A connection timeout was being treated differently from recipients deferred
6340 when testing hosts_max_try with a message that was older than the host's
6341 retry timeout. (The host should not be counted, thus allowing all hosts to
6342 be tried at least once before bouncing.) This may have been the cause of an
6343 occasionally reported bug whereby a message would remain on the queue
6344 longer than the retry timeout, but would be bounced if a delivery was
6345 forced. I say "may" because I never totally pinned down the problem;
6346 setting up timeout/retry tests is difficult. See also the next item.
6348 79. The ultimate address timeout was not being applied to errors that involved
6349 a combination of host plus message (for example, a timeout on a MAIL
6350 command). When an address resolved to a number of possible hosts, and they
6351 were not all tried for each delivery (e.g. because of hosts_max_try), a
6352 message could remain on the queue longer than the retry timeout.
6354 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6357 81. Fixed an obscure SMTP outgoing bug which required at least the following
6358 conditions: (a) there was another message waiting for the same server;
6359 (b) the server returned 5xx to all RCPT commands in the first message so
6360 that the message was not completed; (c) the server dropped the connection
6361 or gave a negative response to the RSET that Exim sends to abort the
6362 transaction. The observed case was a dropped connection after DATA that had
6363 been sent in pipelining mode. That is, the server had advertised PIPELINING
6364 but was not implementing it correctly. The effect of the bug was incorrect
6365 behaviour, such as trying another host, and this could lead to a crash.
6371 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6372 and daemon.c were passed as pointers to ints; they should have been
6373 pointers to socklen_t variables (which are typically unsigned ints).
6375 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6378 3. Fixed a really odd bug that affected only the testing scheme; patching a
6379 certain fixed string in the binary changed the value of another string that
6380 happened to be identical to the end of the original first string.
6382 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6383 name", it returns that address as the IP address. On some operating
6384 systems (e.g. Solaris), it also passes back the IP address string as the
6385 "host name". However, on others (e.g. Linux), it passes back an empty
6386 string. Exim wasn't checking for this, and was changing the host name to an
6387 empty string, assuming it had been canonicalized.
6389 5. Although rare, it is permitted to have more than one PTR record for a given
6390 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6391 all the names associated with an address, because they do in Solaris.
6392 However, it seems that they do not in Linux for data that comes from the
6393 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6394 I found this out when I moved to a new Linux workstation and tried to run
6395 the Exim test suite.
6397 To get round this problem I have changed the code so that it now does its
6398 own call to the DNS to look up PTR records when searching for a host name.
6399 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6400 addresses that are only in /etc/hosts are still found.
6402 This behaviour is, however, controlled by an option called host_lookup_
6403 order, which defaults to "bydns:byaddr". If people want to use the other
6404 order, or indeed, just use one or the other means of lookup, they can
6405 specify it in this variable.
6407 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6408 some operating systems, this comes back from gethostbyaddr() as an empty
6409 string, and this is what Exim used to test for. However, it seems that in
6410 other systems, "." is yielded. Exim now tests for this case too.
6412 7. The values of check_spool_space and check_log_space are now held internally
6413 as a number of kilobytes instead of an absolute number of bytes. If a
6414 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6415 kilobyte. This means that much larger values can be stored.
6417 8. Exim monitor: an attempt to get the action menu when not actually pointing
6418 at a message produces an empty menu entitled "No message selected". This
6419 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6420 no entries in it ("Shell widget menu has zero width and/or height"). So I
6421 have added a single, blank menu entry in this case.
6423 9. Added ${quote_local_part.
6425 10. MIME decoding is now applied to the contents of Subject: header lines when
6428 11. Now that a reference to $sender_host_address automatically causes a reverse
6429 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6430 host lookup before query-style lookups in lists that might use this
6431 variable. This has therefore been abolished, and the "net-" prefix is no
6432 longer necessary for query-style lookups.
6434 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6435 have been a typo for LFLAGS, so it has been changed.
6437 13. The install script calls Exim with "-C /dev/null" in order to find the
6438 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6439 to be output. However, since Exim outputs its version number before the
6440 error, it didn't break the script. It just looked ugly. I fixed this by
6441 always allowing "-C /dev/null" if the caller is root.
6443 14. Ignore overlarge ACL variable number when reading spool file - insurance
6444 against a later release with more variables having written the file.
6446 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6447 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6448 this, as well as the form without the "IPv6" on the front.
6450 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6451 OS/Makefile-Darwin file.
6453 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6454 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6456 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6457 variable. However, some OS use ints for inodes. I've added cast to long int
6458 to get rid of the compiler warning.
6460 19. I had forgotten to lock out "/../" in configuration file names when
6461 ALT_CONFIG_PREFIX was set.
6463 20. Routers used for verification do not need to specify transports. However,
6464 if such a router generated a host list, and callout was configured, Exim
6465 crashed, because it could not find a port number from the (non-existent)
6466 transport. It now assumes port 25 in this circumstance.
6468 21. Added the -t option to exigrep.
6470 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6471 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6472 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6473 them is set, LOOKUP_LSEARCH is forced.
6475 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6477 24. Added sender and host information to the "rejected by local_scan()" log
6478 line; previously there was no indication of these.
6480 25. Added .include_if_exists.
6482 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6483 Linux. It turns out that not all file systems support this. Apparently some
6484 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6485 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6486 not supported on the file descriptor, is now ignored when Exim is trying to
6487 sync a directory. This applies only to Linux.
6489 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6491 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6492 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6495 29. Added support for maildirsize files from supplied patch (modified a bit).
6497 30. The use of :fail: followed by an empty string could lead Exim to respond to
6498 sender verification failures with (e.g.):
6500 550 Verification failed for <xxx>
6501 550 Sender verify failed
6503 where the first response line was missing the '-' that indicates it is not
6504 the final line of the response.
6506 31. The loop for finding the name of the user that called Exim had a hardwired
6507 limit of 10; it now uses the value of finduser_retries, which is used for
6508 all other user lookups.
6510 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6513 33. Exim was neglecting to zero errno before one call of strtol() when
6514 expanding a string and expecting an integer value. On some systems this
6515 resulted in spurious "integer overflow" errors. Also, it was casting the
6516 result into an int without checking.
6518 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6519 did not work. The code looks as if it has *never* worked, though it appears
6520 to have been documented since at least release 1.62. I have made it work.
6522 35. The "timeout_DNS" error in retry rules, also documented since at least
6523 1.62, also never worked. As it isn't clear exactly what this means, and
6524 clearly it isn't a major issue, I have abolished the feature by treating it
6525 as "timeout", and writing a warning to the main and panic logs.
6527 36. The display of retry rules for -brt wasn't always showing the error code
6530 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6531 timeout_connect_A, timeout_connect_MX.
6533 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6534 to the empty sender.
6536 39. The daemon was not analysing the content of -oX till after it had closed
6537 stderr and disconnected from the controlling terminal. This meant that any
6538 syntax errors were only noted on the panic log, and the return code from
6539 the command was 0. By re-arranging the code a little, I've made the
6540 decoding happen first, so such errors now appear on stderr, and the return
6541 code is 1. However, the actual setting up of the sockets still happens in
6542 the disconnected process, so errors there are still only recorded on the
6545 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6546 connections (as happens on some IP stacks) was logged at start up time as
6547 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6548 it from "IPv6 and IPv4", which means that two separate sockets are being
6551 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6552 says whether AF_INET or AF_INET6 was passed as an argument.
6554 42. Exiwhat output was messed up when time zones were included in log
6557 43. Exiwhat now gives more information about the daemon's listening ports,
6558 and whether -tls-on-connect was used.
6560 44. The "port" option of the smtp transport is now expanded.
6562 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6563 silently ignored. Now an error message is written to the main and panic
6566 46. There's a new ACL modifier called "logwrite" which writes to a log file
6567 as soon as it is encountered.
6569 47. Added $local_user_uid and $local_user_gid at routing time.
6571 48. Exim crashed when trying to verify a sender address that was being
6574 49. Exim was recognizing only a space character after ".include". It now also
6575 recognizes a tab character.
6577 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6578 extracting the relevant information from the specification. The man page no
6579 longer contains scrambled data for the -d option, and I've added a section
6580 at the front about calling Exim under different names.
6582 51. Added "extra_headers" argument to the "mail" command in filter files.
6584 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6587 53. Installed eximstats 1.29.
6589 54. Added transport_filter_timeout as a generic transport option.
6591 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6592 Cc: header lines. This was required by RFC 822, but it not required by RFC
6595 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6596 incoming messages that did not have them. Now it does so only if the
6597 message originates locally, that is, if there is no associated remote host
6598 address. When Resent- header lines are present, this applies to the Resent-
6599 lines rather than the non-Resent- lines.
6601 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6602 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6604 58. Messages for configuration errors now include the name of the main
6605 configuration file - useful now that there may be more than one file in a
6606 list (.included file names were always shown).
6608 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6609 for those rare installations that do not start the daemon as root or run it
6610 setuid root. I've cut out the call to initgroups() if the daemon is not
6613 60. The Exim user and group can now be bound into the binary as text strings
6614 that are looked up at the start of Exim's processing.
6616 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6618 62. Added $mailstore_basename variable.
6620 63. Installed patch to sieve.c from Michael Haardt.
6622 64. When Exim failed to open the panic log after failing to open the main log,
6623 the original message it was trying to log was written to stderr and debug
6624 output, but if they were not available (the usual case in production), it
6625 was lost. Now it is written to syslog before the two lines that record the
6626 failures to open the logs.
6628 65. Users' Exim filters run in subprocesses under the user's uid. It is
6629 possible for a "deliver" command or an alias in a "personal" command to
6630 provoke an address rewrite. If logging of address rewriting is configured,
6631 this fails because the process is not running as root or exim. There may be
6632 a better way of dealing with this, but for the moment (because 4.30 needs
6633 to be released), I have disabled address rewrite logging when running a
6634 filter in a non-root, non-exim process.
6640 1. The buildconfig auxiliary program wasn't quoting the value set for
6641 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6642 not defined. This bug was masked in 4.22 by the effect that was fixed in
6645 2. Some messages that were rejected after a message id was allocated were
6646 shown as "incomplete" by exigrep. It no longer does this for messages that
6647 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6649 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6650 have allow_domain_literals set, the ID did not get logged in the <= line.
6651 Domain literals are now always recognized in Message-ID: header lines.
6653 4. The first argument for a ${extract expansion item is the key name or field
6654 number. Leading and trailing spaces in this item were not being ignored,
6655 causing some misleading effects.
6657 5. When deliver_drop_privilege was set, single queue runner processes started
6658 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6659 same command in the process it spins off) were not dropping privilege.
6661 6. When the daemon running as "exim" started a queue runner, it always
6662 re-executed Exim in the spun-off process. This is a waste of effort when
6663 deliver_drop_privilege is set. The new process now just calls the
6664 queue-runner function directly.
6670 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6673 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6674 ignored. Though the use of -oP was forcing the writing of a pid file, it
6675 was always written to the default place.
6677 3. If the message "no IP address found for host xxxx" is generated during
6678 incoming verification, it is now followed by identification of the incoming
6679 connection (so you can more easily find what provoked it).
6681 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6683 5. Added some features to "harden" Exim a bit more against certain attacks:
6685 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6686 be put in Local/Makefile. This is like the never_users runtime option,
6687 but it cannot be overridden. The default setting is "root".
6689 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6690 prefix string with which any file named in a -C command line option
6693 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6694 is retained for -C and -D only if the caller of Exim is root. Without
6695 it, the exim user may also use -C and -D and retain privilege.
6697 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6698 command line option is disabled.
6700 6. Macro names set by the -D option must start with an upper case letter, just
6701 like macro names defined in the configuration file.
6703 7. Added "dereference=" facility to LDAP.
6705 8. Two instances of the typo "uknown" in the source files are fixed.
6707 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6708 the Configure-Makefile script screwed up while processing it.
6710 10. Incorporated PCRE 4.4.
6712 11. The SMTP synchronization check was not operating right at the start of an
6713 SMTP session. For example, it could not catch a HELO sent before the client
6714 waited for the greeting. There is now a check for outstanding input at the
6715 point when the greeting is written. Because of the duplex, asynchronous
6716 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6717 way, but not yet received, when the check is performed.
6719 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6720 on TCP/IP sockets, because this apparently causes some broken clients to
6723 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6724 unchanged) from the Cygwin maintainer.
6726 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6727 was supported instead of testing for mbx. Effectively a typo.
6729 15. The spa authenticator server code was not checking that the input it
6730 received was valid base64.
6732 16. The debug output line for the "set" modifier in ACLs was not showing the
6733 name of the variable that was being set.
6735 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6737 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6738 lookup on its own. Something else was needed to trigger the lookup. For
6739 example, a match in host_lookup or the need for a host name in a host list.
6740 Now, if $sender_host_name is referenced and the host name has not yet been
6741 looked up, a lookup is performed. If the lookup fails, the variable remains
6742 empty, and $host_lookup_failed is set to "1".
6744 19. Added "eqi" as a case-independent comparison operator.
6746 20. The saslauthd authentication condition could segfault if neither service
6747 nor realm was specified.
6749 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6750 error message that was logged was misleading, and incoming SMTP
6751 connections were dropped. The message is now more accurate, and temporary
6752 errors are given to SMTP connections.
6754 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6755 (except RSET) with a 421 error, until QUIT is received. However, it was
6756 failing to send a response to QUIT.
6758 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6759 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6760 ACL is now run after the helo_try_verify_hosts code.
6762 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6763 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6764 case-independent, so other case variants are also recognized). Apparently
6765 some systems use these upper case variants.
6767 25. If more than two messages were waiting for the same host, and a transport
6768 filter was specified for the transport, Exim sent two messages over the
6769 same TCP/IP connection, and then failed with "socket operation on non-
6770 socket" when it tried to send the third.
6772 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6774 27. The extern definition of crypt16() in expand.c was not being excluded when
6775 the OS had its own crypt16() function.
6777 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6778 as a preferred synonym for return_size_limit, both as an option and as an
6781 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6783 30. Changed the default configuration ACL to relax the local part checking rule
6784 for addresses that are not in any local domains. For these addresses,
6785 slashes and pipe symbols are allowed within local parts, but the sequence
6786 /../ is explicitly forbidden.
6788 31. SPA server authentication was not clearing the challenge buffer before
6791 32. log_message in a "warn" ACL statement was writing to the reject log as
6792 well as to the main log, which contradicts the documentation and doesn't
6793 seem right (because no rejection is happening). So I have stopped it.
6795 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6796 However, I am unable to do any testing of this.
6798 34. Fixed an infelicity in the appendfile transport. When checking directories
6799 for a mailbox, to see if any needed to be created, it was accidentally
6800 using path names with one or more superfluous leading slashes; tracing
6801 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6803 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6804 contents are added to the log line that is written for every discarded
6805 recipient. (Previously a log_message setting was ignored.)
6807 36. The ${quote: operator now quotes the string if it is empty.
6809 37. The install script runs exim in order to find its version number. If for
6810 some reason other than non-existence or emptiness, which it checks, it
6811 could not run './exim', it was installing it with an empty version number,
6812 i.e. as "exim-". This error state is now caught, and the installation is
6815 38. An argument was missing from the function that creates an error message
6816 when Exim fails to connect to the socket for saslauthd authentication.
6817 This could cause Exim to crash, or give a corrupted message.
6819 39. Added isip, isip4, and isip6 to ${if conditions.
6821 40. The ACL variables $acl_xx are now saved with the message, and can be
6822 accessed later in routers, transports, and filters.
6824 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6825 strings in the file are not string-expanded.
6827 42. If a MAIL command specified a SIZE value that was too large to fit into an
6828 int variable, the check against message_size_limit failed. Such values are
6829 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6830 day this will have to be increased, but I don't think I want to be around
6831 when emails are that large.
6838 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6839 iconv() is not standard in FreeBSD.
6841 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6842 IPv6 enabled. The observed symptom was a segmentation fault on return from
6843 the function os_common_find_running_interfaces() in src/os.c.
6845 3. In the check_special_case() function in daemon.c I had used "errno" as an
6846 argument name, which causes warnings on some systems. This was basically a
6847 typo, since it was named "eno" in the comments!
6849 4. The code that waits for the clock to tick (at a resolution of some fraction
6850 of a second) so as to ensure message-id uniqueness was always waiting for
6851 at least one whole tick, when it could have waited for less. [This is
6852 almost certainly not relevant at current processor speeds, where it is
6853 unlikely to ever wait at all. But we try to future-proof.]
6855 5. The function that sleeps for a time interval that includes fractions of a
6856 second contained a race. It did not block SIGALRM between setting the
6857 timer, and suspending (a couple of lines later). If the interval was short
6858 and the sigsuspend() was delayed until after it had expired, the suspension
6859 never ended. On busy systems this could lead to processes getting stuck for
6862 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6863 process, before it forks any delivery processes. The open lookup caching
6864 mechanism meant that the open file or database connection was passed into
6865 the delivery process. The problem was that delivery processes always tidy
6866 up cached lookup data. This could cause a problem for the next delivery
6867 process started by the queue runner, because the external queue runner
6868 process does not know about the closure. So the next delivery process
6869 still has data in the lookup cache. In the case of a file lookup, there was
6870 no problem because closing a file descriptor in a subprocess doesn't affect
6871 the parent. However, if the lookup was caching a connection to a database,
6872 the connection was closed, and the second delivery process was likely to
6873 see errors such as "PGSQL: query failed: server closed the connection
6874 unexpectedly". The problem has been fixed by closing all cached lookups
6875 in a queue runner before running a delivery process.
6877 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6878 seem to have the "const" qualifier which it has on other OS. I've
6881 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6882 *of the same type* (client or server) with the same public name that an
6883 error should be diagnosed.
6885 9. When Exim looked up a host name for an IP address, but failed to find the
6886 original IP address when looking up the host name (a safety check), it
6887 output the message "<ip address> does not match any IP for NULL", which was
6888 confusing, to say the least. The bug was that the host name should have
6889 appeared instead of "NULL".
6891 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6892 user that is built into the binary, and the -C or -D options is used, root
6893 privilege is dropped before the configuration file is read. In addition,
6894 logging is switched to stderr instead of the normal log files. If the
6895 configuration then re-defines the Exim user, the unprivileged environment
6896 is probably not what is expected, so Exim logs a panic warning message (but
6899 However, if deliver_drop_privilege is set, the unprivileged state may well
6900 be exactly what is intended, so the warning has been cut out in that case,
6901 and Exim is allowed to try to write to its normal log files.
6907 1. smtp_return_error_details was not giving details for temporary sender
6908 or receiver verification errors.
6910 2. Diagnose a configuration error if two authenticators have the same public
6913 3. Exim used not to create the message log file for a message until the first
6914 delivery attempt. This could be confusing when incoming messages were held
6915 for policy or load reasons. The message log file is now created at the time
6916 the message is received, and an initial "Received" line is written to it.
6918 4. The automatically generated man page for command line options had a minor
6919 bug that caused no ill effects; however, a more serious problem was that
6920 the procedure for building the man page automatically didn't always
6921 operate. Consequently, release 4.20 contains an out-of-date version. This
6922 shouldn't happen again.
6924 5. When building Exim with embedded Perl support, the script that builds the
6925 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6926 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6928 6. The freeze_tell option was not being used for messages that were frozen on
6929 arrival, either by an ACL or by local_scan().
6931 7. Added the smtp_incomplete_transaction log selector.
6933 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6934 was accepting AUTH without a new EHLO.
6936 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6937 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6940 10. Logging of TCP/IP connections (when configured) now happens in the main
6941 daemon process instead of the child process, so that the TCP/IP connection
6942 count is more accurate (but it can never be perfect).
6944 11. The use of "drop" in a nested ACL was not being handled correctly in the
6945 outer ACL. Now, if condition failure induced by the nested "drop" causes
6946 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6947 or "require"), the connection is dropped.
6949 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6950 that yield "discard" can now be used with an "accept" or a "discard" verb,
6951 but an error is generated for any others (because I can't see a useful way
6952 to define what should happen).
6954 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6955 are now processed in the same way as lines in the Exim configuration file.
6956 In particular, continuation lines are supported.
6958 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6960 15. Added -ti meaning -t -i.
6962 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6963 domains, and warn by logging if others are found.
6965 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6966 returned from GIFCONF when Exim is trying to find the list of interfaces on
6967 a host. The code in os.c has been modified to copy each ifreq to an aligned
6968 structure in all cases.
6970 Also, in some cases, the returned ifreq's were being copied to a 'struct
6971 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6972 means the last couple of bytes of an IPv6 address could be chopped if the
6973 ifreq contained only a normal sockaddr (14 bytes storage).
6975 18. Named domain lists were not supported in the hosts_treat_as_local option.
6976 An entry such as +xxxx was not recognized, and was treated as a literal
6979 19. Ensure that header lines added by a DATA ACL are included in the reject log
6980 if the ACL subsequently rejects the message.
6982 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6983 MD5 (which is deprecated).
6985 21. When testing a filter file using -bf, Exim was writing a message when it
6986 took the sender from a "From " line in the message, but it was not doing so
6987 when it took $return_path from a Return-Path: header line. It now does.
6989 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6990 with a valid header line field name (a series of printing characters
6991 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6993 23. Changed "disc" in the source to "disk" to conform to the documentation and
6994 the book and for uniformity.
6996 24. Ignore Sendmail's -Ooption=value command line item.
6998 25. When execve() failed while trying to run a command in a pipe transport,
6999 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7000 could be confused with a return value of 69 from the command itself. This
7001 has been changed to 127, the value the shell returns if it is asked to run
7002 a non-existent command. The wording for the related log line suggests a
7003 non-existent command as the problem.
7005 26. If received_header_text expands to an empty string, do not add a Received:
7006 header line to the message. (Well, it adds a token one on the spool, but
7007 marks it "old" so that it doesn't get used or transmitted.)
7009 27. Installed eximstats 1.28 (addition of -nt option).
7011 28. There was no check for failure on the call to getsockname() in the daemon
7012 code. This can fail if there is a shortage of resources on the system, with
7013 ENOMEM, for example. A temporary error is now given on failure.
7015 29. Contrary to the C standard, it seems that in some environments, the
7016 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7017 program. Exim now does this explicitly; it affects the formatting of
7018 timestamps using strftime().
7020 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7021 complaints. I've now initialized all the variables, to avoid this.
7023 32. Header lines added by a system filter were not being "seen" during
7024 transport-time rewrites.
7026 33. The info_callback() function passed to OpenSSL is set up with type void
7027 (*)(SSL *, int, int), as described somewhere. However, when calling the
7028 function (actually a macro) that sets it up, the type void(*)() is
7029 expected. I've put in a cast to prevent warnings from picky compilers.
7031 34. If a DNS black list lookup found a CNAME record, but there were no A
7032 records associated with the domain it pointed at, Exim crashed.
7034 35. If a DNS black list lookup returned more than one A record, Exim ignored
7035 all but the first. It now scans all returned addresses if a particular IP
7036 value is being sought. In this situation, the contents of the
7037 $dnslist_value variable are a list of all the addresses, separated by a
7040 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7041 to accept a host name and all its aliases if the forward lookup for any of
7042 them yielded the IP address of the incoming connection. Now it accepts only
7043 those names whose forward lookup yields the correct IP address. Any other
7044 names are discarded. This closes a loophole whereby a rogue DNS
7045 administrator could create reverse DNS records to break through a
7046 wildcarded host restriction in an ACL.
7048 37. If a user filter or a system filter that ran in a subprocess used any of
7049 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7050 the wrong values were passed to the pipe command ($thisaddress had the
7051 value of $0, $0 had the value of $1, etc). This bug was introduced by
7052 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7054 38. Improved the line breaking for long SMTP error messages from ACLs.
7055 Previously, if there was no break point between 40 and 75 characters, Exim
7056 left the rest of the message alone. Two changes have been made: (a) I've
7057 reduced the minimum length to 35 characters; (b) if it can't find a break
7058 point between 35 and 75 characters, it looks ahead and uses the first one
7059 that it finds. This may give the occasional overlong line, but at least the
7060 remaining text gets split now.
7062 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7063 file descriptors might be low, and that setting 1000 would always raise it.
7064 It turns out that in some environments, the limit is already over 1000 and
7065 that lowering it causes trouble. So now Exim takes care not to decrease it.
7067 40. When delivering a message, the value of $return_path is set to $sender_
7068 address at the start of routing (routers may change the value). By an
7069 oversight, this default was not being set up when an address was tested by
7070 -bt or -bv, which affected the outcome if any router or filter referred to
7073 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7074 log only when "message" or "log_message" are set. However, if one of the
7075 conditions was an address verification, or a call to a nested ACL, the
7076 messages generated by the underlying test were being passed through. This
7077 no longer happens. The underlying message is available in $acl_verify_
7078 message for both "message" and "log_message" expansions, so it can be
7079 passed through if needed.
7081 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7082 new expansion $bh_ to give the encoded byte string without charset
7083 translation. Translation happens only if iconv() is available; HAVE_ICONV
7084 indicates this at build time. HEADERS_CHARSET gives the charset to
7085 translate to; headers_charset can change it in the configuration, and
7086 "headers charset" can change it in an individual filter file.
7088 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7089 that creates RFC 2047 encoded "words" labels them as that charset instead
7090 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7091 expansion operator; (ii) when Exim creates a From: line for a local
7092 message; (iii) when a header line is rewritten to include a "phrase" part.
7094 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7095 buggy, causing it to skip the first lines of messages whose message ID
7096 ended in 'D'. This would not have bitten before Exim release 4.14, because
7097 message IDs were unlikely to end in 'D' before then. The effect was to have
7098 incorrect size information for certain domains.
7100 45. #include "config.h" was missing at the start of the crypt16.c module. This
7101 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7104 46. If there was a timeout during a "random" callout check, Exim treated it as
7105 a failure of the random address, and carried on sending RSET and the real
7106 address. If the delay was just some slowness somewhere, the response to the
7107 original RCPT would be taken as a response to RSET and so on, causing
7108 mayhem of various kinds.
7110 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7111 when I implemented it. It didn't allow for the fact that some option values
7112 may legitimately be negative (e.g. size_addition), and it didn't even do
7113 the right test for positive values.
7115 48. Domain names in DNS records are case-independent. Exim always looks them up
7116 in lower case. Some resolvers return domain names in exactly the case they
7117 appear in the zone file, that is, they may contain uppercase letters. Not
7118 all resolvers do this - some return always lower case. Exim was treating a
7119 change of case by a resolver as a change of domain, similar to a widening
7120 of a domain abbreviation. This triggered its re-routing code and so it was
7121 trying to route what was effectively the same domain again. This normally
7122 caused routing to fail (because the router wouldn't handle the domain
7123 twice). Now Exim checks for this case specially, and just changes the
7124 casing of the domain that it ultimately uses when it transmits the message
7127 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7130 50. If a filter generated a file delivery with a non-absolute name (possible if
7131 no home directory exists for the router), the forbid_file option was not
7134 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7135 the existing equality matching.
7137 52. Exim was using ints instead of ino_t variables in some places where it was
7138 dealing with inode numbers.
7140 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7141 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7142 called TMPDIR, and if it finds it is different, it changes its value.
7144 54. The smtp_printf() function is now made available to local_scan() so
7145 additional output lines can be written before returning. There is also an
7146 smtp_fflush() function to enable the detection of a dropped connection.
7147 The variables smtp_input and smtp_batched_input are exported to
7150 55. Changed the default runtime configuration: the message "Unknown user"
7151 has been removed from the ACL, and instead placed on the localuser router,
7152 using the cannot_route_message feature. This means that any verification
7153 failures that generate their own messages won't get overridden. Similarly,
7154 the "Unrouteable address" message that was in the ACL for unverifiable
7155 relay addresses has also been removed.
7157 56. Added hosts_avoid_esmtp to the smtp transport.
7159 57. The exicyclog script was not checking for the esoteric option
7160 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7161 will work only if exicyclog is run under the appropriate euid.
7163 58. Following a discussion on the list, the rules by which Exim recognises line
7164 endings on incoming messages have been changed. The -dropcr and drop_cr
7165 options are now no-ops, retained only for backwards compatibility. The
7166 following line terminators are recognized: LF CRLF CR. However, special
7167 processing applies to CR:
7169 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7170 nor a local message in the state where . is a terminator.
7172 (ii) If a bare CR is encountered in a header line, an extra space is added
7173 after the line terminator so as not to end the header. The reasoning
7174 behind this is that bare CRs in header lines are most likely either
7175 to be mistakes, or people trying to play silly games.
7177 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7178 was being incorrectly given as 18 bytes larger than it should have been.
7179 This is a VOB (very old bug).
7181 60. This may never have affected anything current, but just in case it has:
7182 When the local host is found other than at the start of a list of hosts,
7183 the local host, those with the same MX, and any that follow, are discarded.
7184 When the list in question was part of a longer list of hosts, the following
7185 hosts (not currently being processed) were also being discarded. This no
7186 longer happens. I'm not sure if this situation could ever has previously
7189 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7192 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7193 header. This is something that is recommended in a new Internet Draft, and
7194 is something that is documented as being done by Sendmail. There are two
7195 possible values. For messages generated by the autoreply transport, Exim
7198 Auto-Submitted: auto-replied
7200 whereas for all other generated messages (e.g. bounces) it adds
7202 Auto-Submitted: auto-generated
7204 63. The "personal" condition in filters now includes a test for the
7205 Auto-Submitted: header. If it contains the string "auto-" the message it
7206 not considered personal.
7208 64. Added rcpt_include_affixes as a generic transport option.
7210 65. Added queue_only_override (default true).
7212 66. Added the syslog_duplication option.
7214 67. If what should have been the first header line of a message consisted of
7215 a space followed by a colon, Exim was mis-interpreting it as a header line.
7216 It isn't of course - it is syntactically invalid and should therefore be
7217 treated as the start of the message body. The misbehaviour could have
7218 caused a number of strange effects, including loss of data in subsequent
7219 header lines, and spool format errors.
7221 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7222 client host had authenticated. This control can now be exercised by an ACL
7223 for more flexibility.
7225 69. By default, callouts do not happen when testing with -bh. There is now a
7226 variant, -bhc, which does actually run the callout code, including
7227 consulting and updating the callout cache.
7229 70. Added support for saslauthd authentication, courtesy of Alexander
7232 71. If statvfs() failed on the spool or log directories while checking their
7233 size for availability, Exim confusingly gave the error "space shortage".
7234 Furthermore, in debugging mode it crashed with a floating point exception.
7235 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7236 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7237 serious problem, Exim now writes to the main and panic logs when this
7238 happens, with details of the failure. It then refuses to accept the
7239 incoming message, giving the message "spool directory problem" or "log
7240 directory problem" with a 421 code for SMTP messages.
7242 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7243 0, 1, and 2 exist, because some OS complain for execs without them (see
7244 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7245 descriptors. However, the code omitted to check that the open succeeded,
7246 causing mysterious errors if for some reason the permissions on /dev/null
7247 got screwed. Now Exim writes a message to the main and panic logs, and
7248 bombs out if it can't open /dev/null.
7250 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7251 interact so that it is all more flexible. It is supposed to remain
7252 backwards compatible. Also added extra_local_interfaces.
7254 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7255 to bomb out with an assertion failure - to the client this appears as a
7256 connection drop. This problem occurs in the part of the code that was taken
7257 from the Samba project. Fortunately, the assertion is in a very simple
7258 function, so I have fixed this by reproducing the function inline in the
7259 one place where it is called, and arranging for authentication to fail
7260 instead of killing the process with assert().
7262 75. The SPA client code was not working when the server requested OEM rather
7263 than Unicode encoding.
7265 76. Added code to make require_files with a specific uid setting more usable in
7266 the case where statting the file as root fails - usually a non-root-mounted
7267 NFS file system. When this happens and the failure is EACCES, Exim now
7268 forks a subprocess and does the per-uid checking as the relevant uid.
7270 77. Added process_log_path.
7272 78. If log_file_path was not explicitly set, a setting of check_log_space or
7273 check_log_inodes was ignored.
7275 79. If a space check for the spool or log partitions fails, the incident is now
7276 logged. Of course, in the latter case the data may get lost...
7278 80. Added the %p formatting code to string_format() so that it can be used to
7279 print addresses in debug_print(). Adjusted all the address printing in the
7280 debugging in store.c to use %p rather than %d.
7282 81. There was a concern that a line of code in smtp_in.c could overflow a
7283 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7284 initially expressed, the concern was not well-founded, because trailing
7285 spaces are removed early. However, if the trailing spaces were followed by
7286 a NULL, they did not get removed, so the overflow was possible. Two fixes
7289 (a) I re-wrote the offending code in a cleaner fashion.
7290 (b) If an incoming SMTP command contains a NULL character, it is rejected
7293 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7294 runs initgroups(), so that if the Exim user is in any additional groups,
7295 they will be used during message reception.
7301 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob