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.
111 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
112 When redis_servers is set to a list of > 1 element, and the Redis servers
113 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
114 case of MOVED into a DEFER case instead, thus moving the query onto the
115 next server in the list. For a cluster of N elements, all N servers must
116 be defined in redis_servers.
118 GF/02 Catch and remove uninitialized value warning in exiqsumm
119 Check for existence of @ARGV before looking at $ARGV[0]
121 JH/01 Replace the store_release() internal interface with store_newblock(),
122 which internalises the check required to safely use the old one, plus
123 the allocate and data copy operations duplicated in both (!) of the
124 extant use locations.
126 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
127 modifier. This matches the restriction on the commandline.
129 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
130 Previously only the last row was returned.
132 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
133 we assumed that tags in the header were well-formed, and parsed the
134 element content after inspecting only the first char of the tag.
135 Assumptions at that stage could crash the receive process on malformed
138 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
139 While running the DKIM ACL we operate on the Permanent memory pool so that
140 variables created with "set" persist to the DATA ACL. Also (at any time)
141 DNS lookups that fail create cache records using the Permanent pool. But
142 expansions release any allocations made on the current pool - so a dnsdb
143 lookup expansion done in the DKIM ACL releases the memory used for the
144 DNS negative-cache, and bad things result. Solution is to switch to the
145 Main pool for expansions.
146 While we're in that code, add checks on the DNS cache during store_reset,
147 active in the testsuite.
148 Problem spotted, and debugging aided, by Wolfgang Breyha.
150 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
151 When none of the hosts presented to a transport match an already-open
152 connection, close it and proceed with the list. Previously we would
153 queue the message. Spotted by Lena with Yahoo, probably involving
156 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
157 Previously a spurious "250 OK id=" response was appended to the proper
160 JH/08 The "support for" informational output now, which built with Content
161 Scanning support, has a line for the malware scanner interfaces compiled
162 in. Interface can be individually included or not at build time.
164 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
165 by the template makefile "src/EDITME". The "STREAM" support for an older
166 ClamAV interface method is removed.
168 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
169 rows affected is given instead).
171 JH/11 The runtime Berkeley DB library version is now additionally output by
172 "exim -d -bV". Previously only the compile-time version was shown.
174 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
175 SMTP connection. Previously, when one had more receipients than the
176 first, an abortive onward connection was made. Move to full support for
177 multiple onward connections in sequence, handling cutthrough connection
178 for all multi-message initiating connections.
180 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
181 routers. Previously, a multi-recipient message would fail to match the
182 onward-connection opened for the first recipient, and cause its closure.
184 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
185 a timeout on read on a GnuTLS initiating connection, resulting in the
186 initiating connection being dropped. This mattered most when the callout
187 was marked defer_ok. Fix to keep the two timeout-detection methods
190 JH/15 Relax results from ACL control request to enable cutthrough, in
191 unsupported situations, from error to silently (except under debug)
192 ignoring. This covers use with PRDR, frozen messages, queue-only and
195 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
197 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
198 metadata, resulting in a crash in free().
200 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
201 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
202 Broken also in d185889f4, with init system revamp.
204 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
205 Previously we abruptly closed the connection after reading a malware-
206 found indication; now we go on to read the "scan ok" response line,
209 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
210 ACL. Previously, a crash would result.
212 JH/19 Speed up macro lookups during configuration file read, by skipping non-
213 macro text after a replacement (previously it was only once per line) and
214 by skipping builtin macros when searching for an uppercase lead character.
216 JH/20 DANE support moved from Experimental to mainline. The Makefile control
217 for the build is renamed.
219 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
220 was allocated for every new TLS startup, meaning one per message. Fix
221 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
223 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
224 reported the original. Fix to report (as far as possible) the ACL
225 result replacing the original.
227 JH/23 Fix memory leak during multi-message connections using STARTTLS under
228 OpenSSL. Certificate information is loaded for every new TLS startup,
229 and the resources needed to be freed.
231 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
233 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
236 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
237 DATA response info to the (existing) per-recipient response info for
238 the "C=" log element. It can have useful tracking info from the
239 destination system. Patch from Simon Arlott.
241 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
242 length value. Previously this would segfault.
244 HS/02 Support Avast multiline protoocol, this allows passing flags to
245 newer versions of the scanner.
247 JH/28 Ensure that variables possibly set during message acceptance are marked
248 dead before release of memory in the daemon loop. This stops complaints
249 about them when the debug_store option is enabled. Discovered specifically
250 for sender_rate_period, but applies to a whole set of variables.
251 Do the same for the queue-runner and queue-list loops, for variables set
252 from spool message files. Do the same for the SMTP per-message loop, for
253 certain variables indirectly set in ACL operations.
255 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
256 as a multi-recipient message from a mailinglist manager). The coding had
257 an arbitrary cutoff number of characters while checking for more input;
258 enforced by writing a NUL into the buffer. This corrupted long / fast
259 input. The problem was exposed more widely when more pipelineing of SMTP
260 responses was introduced, and one Exim system was feeding another.
261 The symptom is log complaints of SMTP syntax error (NUL chars) on the
262 receiving system, and refused recipients seen by the sending system
263 (propating to people being dropped from mailing lists).
264 Discovered and pinpointed by David Carter.
266 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
267 replaced by the ${authresults } expansion.
269 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
271 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
272 allows proper process termination in container environments.
274 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
275 Previously the "final dot" had a newline after it; ensure it is CR,LF.
277 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
278 and "err_perm", deprecated since 4.83 when the RFC-defined words
279 "temperror" and "permerror" were introduced.
281 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
282 transport-filters or DKIM-signing. The restriction was lost in the
283 consolidation of verify-callout and delivery SMTP handling.
284 Extend the restriction to also cover ARC-signing.
286 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
287 in defer=pass mode supply a 450 to the initiator. Previously the message
290 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
291 tls_require_ciphers is used as before.
293 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
294 "pass_unscanned". Only tmpfails from the scanner are written to
295 the paniclog, as they may require admin intervention (permission
296 denied, license issues). Other scanner errors (like decompression
297 bombs) do not cause a paniclog entry.
299 JH/36 Fix reinitialisation of DKIM logging variable between messages.
300 Previously it was possible to log spurious information in receive log
303 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
304 triggered odd behaviour from Outlook Express clients.
306 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
309 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
310 since the IETF WG has not yet settled on that versus the original
311 "bare" representation.
313 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
314 Previously the millisecond value corrupted the output.
315 Fix also for syslog_pid=no and log_selector +pid, for which the pid
316 corrupted the output.
322 JH/01 Rework error string handling in TLS interface so that the caller in
323 more cases is responsible for logging. This permits library-sourced
324 string to be attached to addresses during delivery, and collapses
325 pairs of long lines into single ones.
327 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
328 during configuration. Wildcards are allowed and expanded.
330 JH/02 Rework error string handling in DKIM to pass more info back to callers.
331 This permits better logging.
333 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
334 do not close it down and have the child transport start it up again on
335 the passed-on TCP connection. Instead, proxy the child (and any
336 subsequent ones) for TLS via a unix-domain socket channel. Logging is
337 affected: the continued delivery log lines do not have any DNSSEC, TLS
338 Certificate or OCSP information. TLS cipher information is still logged.
340 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
341 identical IP addresses on different listening ports. Will also affect
344 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
345 add noisy ifdef guards to special-case this sillyness.
346 Patch from Bernd Kuhls.
348 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
349 than 255 are no longer allowed.
351 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
352 Disable the session-cache too, which might reduce our load. Since we
353 currrectly use a new context for every connection, both as server and
354 client, there is no benefit for these.
355 GnuTLS appears to not support tickets server-side by default (we don't
356 call gnutls_session_ticket_enable_server()) but client side is enabled
357 by default on recent versions (3.1.3 +) unless the PFS priority string
360 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
361 <https://reproducible-builds.org/specs/source-date-epoch/>.
363 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
364 the check for any unsuccessful recipients did not notice the limit, and
365 erroneously found still-pending ones.
367 JH/08 Pipeline CHUNKING command and data together, on kernels that support
368 MSG_MORE. Only in-clear (not on TLS connections).
370 JH/09 Avoid using a temporary file during transport using dkim. Unless a
371 transport-filter is involved we can buffer the headers in memory for
372 creating the signature, and read the spool data file once for the
373 signature and again for transmission.
375 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
376 4.77 as the kernel support then wasn't solid, having issues in 64bit
377 mode. Now, it's been long enough. Add support for FreeBSD also.
379 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
380 case where the routing stage had gathered several addresses to send to
381 a host before calling the transport for the first, we previously failed
382 to close down TLS in the old transport process before passing the TCP
383 connection to the new process. The new one sent a STARTTLS command
384 which naturally failed, giving a failed delivery and bloating the retry
385 database. Investigation and fix prototype from Wolfgang Breyha.
387 JH/12 Fix check on SMTP command input synchronisation. Previously there were
388 false-negatives in the check that the sender had not preempted a response
389 or prompt from Exim (running as a server), due to that code's lack of
390 awareness of the SMTP input buffering.
392 PP/04 Add commandline_checks_require_admin option.
393 Exim drops privileges sanely, various checks such as -be aren't a
394 security problem, as long as you trust local users with access to their
395 own account. When invoked by services which pass untrusted data to
396 Exim, this might be an issue. Set this option in main configuration
397 AND make fixes to the calling application, such as using `--` to stop
400 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
401 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
402 responses to those, into a single TLS record each way (this usually means
403 a single packet). As a side issue, smtp_enforce_sync now works on TLS
406 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
407 affects you only if you're dancing at the edge of the param size limits.
408 If you are, and this message makes sense to you, then: raise the
409 configured limit or use OpenSSL 1.1. Nothing we can do for older
412 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
413 cmdline element to get the documented default one. Previously it was
416 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
417 are unsuable for non-message-specific reasons - eg. connection timeout,
418 banner-time rejection.
420 JH/16 Fix logging of delivery remote port, when specified by router, under
423 PP/06 Repair manualroute's ability to take options in any order, even if one
424 is the name of a transport.
427 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
429 JH/17 Change the list-building routines interface to use the expanding-string
430 triplet model, for better allocation and copying behaviour.
432 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
433 Previously it was constructed the first time a possibly-matching string
434 was met in the configuration file input during startup; now it is done
437 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
438 compatible one, to avoid the (poorly documented) possibility of a config
439 file in the working directory redirecting the DB files, possibly correpting
440 some existing file. CVE-2017-10140 assigned for BDB.
442 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
443 cache-hot. Previously, although the result was properly cached, the
444 initial verify call returned a defer.
446 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
447 the main verify for receipient in uncached-mode.
449 JH/22 Retire historical build files to an "unsupported" subdir. These are
450 defined as "ones for which we have no current evidence of testing".
452 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
453 if present. Previously it was ignored.
455 JH/24 Start using specified-initialisers in C structure init coding. This is
456 a C99 feature (it's 2017, so now considered safe).
458 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
459 if was a fixed-sized field and bitmask ops via macros; it is now more
462 PP/07 GitHub PR 56: Apply MariaDB build fix.
463 Patch provided by Jaroslav Škarvada.
465 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
466 during Coverity cleanups [4.87 JH/47]
467 Diagnosis and fix provided by Michael Fischer v. Mollard.
469 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
470 the right size to place the terminating semicolon on its own folded
471 line, the header hash was calculated to an incorrect value thanks to
472 the (relaxed) space the fold became.
474 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
475 and confused the parent.
477 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
478 which could crash as a result. This could lead to undeliverable messages.
480 JH/28 Logging: "next input sent too soon" now shows where input was truncated
483 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
484 matters on fast-turnover and PID-randomising systems, which were getting
485 out-of-order delivery.
487 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
488 a possibly-overlapping copy. The symptom was that "Remote host closed
489 connection in response to HELO" was logged instead of the actual 4xx
492 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
493 Previously only that bufferd was discarded, resulting in SYMTP command
496 JH/32 DKIM: when a message has multiple signatures matching an identity given
497 in dkim_verify_signers, run the dkim acl once for each. Previously only
498 one run was done. Bug 2189.
500 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
501 "panic the current process" to "deliberately defer". The panic log is
502 still written with the problem list name; the mail and reject logs now
503 get a temp-reject line for the message that was being handled, saying
504 something like "domains check lookup or other defer". The SMTP 451
505 message is still "Temporary local problem".
507 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
508 A crafted sequence of BDAT commands could result in in-use memory beeing
509 freed. CVE-2017-16943.
511 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
512 from SMTP input. Previously it was always done; now only done for DATA
513 and not BDAT commands. CVE-2017-16944.
515 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
516 to the message (such as an overlong header line). Previously this was
517 not done and we did not exit BDAT mode. Followon from the previous item
518 though a different problem.
524 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
525 than -2003 did; needs libidn2 in addition to libidn.
527 JH/02 The path option on a pipe transport is now expanded before use.
529 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
530 Patch provided by "Björn", documentation fix added too.
532 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
533 missing a wire-to-host endian conversion.
535 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
536 close after a BDAT command line could be taken as a following command,
537 giving a synch failure. Fix by only checking for synch immediately
538 before acknowledging the chunk.
540 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
541 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
542 macro. Patches provided by Josh Soref.
544 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
545 Previously we did not; the RFC seems ambiguous and VRFY is not listed
546 by IANA as a service extension. However, John Klensin suggests that we
549 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
550 the dkim code may be unix-mode line endings rather than smtp wire-format
551 CRLF, so prepend a CR to any bare LF.
553 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
554 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
556 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
557 the first were themselves being wrongly included in the feed into dkim
558 processing; with most chunk sizes in use this resulted in an incorrect
559 body hash calculated value.
561 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
562 DKIM signature block, for verification. Although advised against by
563 standards it is specifically not ruled illegal.
565 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
567 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
568 missing a body hash (the bh= tag).
570 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
571 It seems that HAProxy sends the Proxy Protocol information in clear and
572 only then does a TLS startup, so do the same.
574 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
575 TCP connections (such as for Spamd) unless the daemon successfully set
576 Fast Open mode on its listening sockets. This fixes breakage seen on
577 too-old kernels or those not configured for Fast Open, at the cost of
578 requiring both directions being enabled for TFO, and TFO never being used
579 by non-daemon-related Exim processes.
581 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
582 endings, at least on the first header line. Try to canonify any that get
583 past that check, despite the cost.
585 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
586 now limited to an arbitrary five deep, while parsing addresses with the
587 strip_excess_angle_brackets option enabled.
589 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
590 instead leave the unprompted TLS handshake in socket buffer for the
591 TLS library to consume.
593 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
595 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
597 JH/16 Drop variables when they go out of scope. Memory management drops a whole
598 region in one operation, for speed, and this leaves assigned pointers
599 dangling. Add checks run only under the testsuite which checks all
600 variables at a store-reset and panics on a dangling pointer; add code
601 explicitly nulling out all the variables discovered. Fixes one known
602 bug: a transport crash, where a dangling pointer for $sending_ip_address
603 originally assigned in a verify callout, is re-used.
605 PP/06 Drop '.' from @INC in various Perl scripts.
607 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
609 PP/08 Reduce a number of compilation warnings under clang; building with
610 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
611 should be warning-free.
613 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
615 HS/01 Fix portability problems introduced by PP/08 for platforms where
616 realloc(NULL) is not equivalent to malloc() [SunOS et al].
618 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
619 chunk. This allows us to accept broken chunked messages. We need a more
620 general solution here.
622 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
623 already-broken messages in the queue.
625 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
627 JH/19 Fix reference counting bug in routing-generated-address tracking.
633 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
634 supports it and a size is available (ie. the sending peer gave us one).
636 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
637 years of being deprecated). The replacements are the ACLs
638 acl_smtp_mime and acl_not_smtp_mime.
640 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
641 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
642 or even an in-clear connection were permitted. Now, if the host lookup
643 was dnssec and dane was requested then the host is only used if the
644 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
645 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
646 if one fails this test.
647 This means that a poorly-configured remote DNS will make it incommunicado;
648 but it protects against a DNS-interception attack on it.
650 JH/04 Bug 1810: make continued-use of an open smtp transport connection
651 non-noisy when a race steals the message being considered.
653 JH/05 If main configuration option tls_certificate is unset, generate a
654 self-signed certificate for inbound TLS connections.
656 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
657 in rewrites and routers.
659 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
660 and logged a warning sing 4.83; now they are a configuration file error.
662 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
663 (lacking @domain). Apply the same qualification processing as RCPT.
665 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
667 JH/10 Support ${sha256:} applied to a string (as well as the previous
670 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
671 a cutthrough deliver is pending, as we always want to make a connection.
672 This also avoids re-routing the message when later placing the cutthrough
673 connection after a verify cache hit.
674 Do not update it with the verify result either.
676 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
677 when routing results in more than one destination address.
679 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
680 signing (which inhibits the cutthrough capability). Previously only
681 the presence of an option was tested; now an expansion evaluating as
682 empty is permissible (obviously it should depend only on data available
683 when the cutthrough connection is made).
685 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
686 the relevant preceding SMTP command did not note the pipelining mode.
688 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
689 Previously they were not counted.
691 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
692 as one having no matching records. Previously we deferred the message
693 that needed the lookup.
695 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
696 distinguished as "(=".
698 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
699 for missing MX records. Previously it only worked for missing A records.
701 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
703 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
704 after the data-go-ahead and data-ack. Patch from Jason Betts.
706 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
707 even for a "none" policy. Patch from Tony Meyer.
709 JH/22 Fix continued use of a connection for further deliveries. If a port was
710 specified by a router, it must also match for the delivery to be
713 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
714 When one of the recipients of a message was unsuitable for the connection
715 (has no matching addresses), we lost track of needing to mark it
716 deferred. As a result mail would be lost.
718 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
720 JH/25 Decoding ACL controls is now done using a binary search; the source code
721 takes up less space and should be simpler to maintain. Merge the ACL
722 condition decode tables also, with similar effect.
724 JH/26 Fix problem with one_time used on a redirect router which returned the
725 parent address unchanged. A retry would see the parent address marked as
726 delivered, so not attempt the (identical) child. As a result mail would
729 JH/27 Fix a possible security hole, wherein a process operating with the Exim
730 UID can gain a root shell. Credit to http://www.halfdog.net/ for
731 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
734 JH/28 Enable {spool,log} filesystem space and inode checks as default.
735 Main config options check_{log,spool}_{inodes,space} are now
736 100 inodes, 10MB unless set otherwise in the configuration.
738 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
739 Previously it only applied to the main-section connection policy
742 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
744 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
745 by me. Added RFC7919 DH primes as an alternative.
747 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
748 are not in the system include path.
750 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
751 GnuTLS, when a session startup failed (eg because the client disconnected)
752 Exim did stdio operations after fclose. This was exposed by a recent
753 change which nulled out the file handle after the fclose.
755 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
756 signed directly by the cert-signing cert, rather than an intermediate
757 OCSP-signing cert. This is the model used by LetsEncrypt.
759 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
761 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
762 an incoming connection.
764 HS/02 Bug 1802: Do not half-close the connection after sending a request
767 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
768 fallback to "prime256v1".
770 JH/34 SECURITY: Use proper copy of DATA command in error message.
771 Could leak key material. Remotely exploitable. CVE-2016-9963.
777 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
778 and 3.4.4 - once the server is enabled to respond to an OCSP request
779 it does even when not requested, resulting in a stapling non-aware
780 client dropping the TLS connection.
782 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
783 support variable-length bit vectors. No functional change.
785 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
786 The I= interface field on outgoing lines is now after the H= remote
787 host field, same as incoming lines. There is a separate
788 outgoing_interface log selector which allows you to disable the
791 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
792 If not running log_selector +smtp_connection the mainlog would be held
793 open indefinitely after a "too many connections" event, including to a
794 deleted file after a log rotate. Leave the per net connection logging
795 leaving it open for efficiency as that will be quickly detected by the
796 check on the next write.
798 HS/01 Bug 1671: Fix post transport crash.
799 Processing the wait-<transport> messages could crash the delivery
800 process if the message IDs didn't exist for some reason. When
801 using 'split_spool_directory=yes' the construction of the spool
802 file name failed already, exposing the same netto behaviour.
804 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
805 mime_regex ACL conditions.
807 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
808 to DSN fail messages (bounces): remote IP, remote greeting, remote response
809 to HELO, local diagnostic string.
811 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
812 log line to debug. Even when configured with a tls authenticator many
813 client connections are expected to not authenticate in this way, so
814 an authenticate fail is not an error.
816 HS/02 Add the Exim version string to the process info. This way exiwhat
817 gives some more detail about the running daemon.
819 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
820 matter for fast-change records such as DNSBLs.
822 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
823 retry record, even if constant. There may be multiple transports with
824 different interface settings and the retry behaviour needs to be kept
827 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
829 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
831 JH/10 Bug 840: fix log_defer_output option of pipe transport
833 JH/11 Bug 830: use same host for all RCPTS of a message, even under
834 hosts_randomize. This matters a lot when combined with mua_wrapper.
836 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
837 ${quote_pgsql:<string>} operator.
839 JH/13 Bug 1708: avoid misaligned access in cached lookup.
841 JH/14 Change header file name for freeradius-client. Relevant if compiling
842 with Radius support; from the Gentoo tree and checked under Fedora.
844 JH/15 Bug 1712: Introduce $prdr_requested flag variable
846 JH/16 Bug 1714: Permit an empty string as expansion result for transport
847 option transport_filter, meaning no filtering.
849 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
851 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
852 defaults to "*" (all hosts). The variable is now available when not built
853 with TLS, default unset, mainly to enable keeping the testsuite sane.
854 If a server certificate is not supplied (via tls_certificate) an error is
855 logged, and clients will find TLS connections fail on startup. Presumably
856 they will retry in-clear.
857 Packagers of Exim are strongly encouraged to create a server certificate
858 at installation time.
860 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
861 with the $config_file variable.
863 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
864 in transport context, after the attempt, and per-recipient. The latter type
865 is per host attempted. The event data is the error message, and the errno
866 information encodes the lookup type (A vs. MX) used for the (first) host,
867 and the trailing two digits of the smtp 4xx response.
869 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
870 to write to mainlog (or rejectlog, paniclog) in the window between file
871 creation and permissions/ownership being changed. Particularly affects
872 installations where exicyclog is run as root, rather than exim user;
873 result is that the running daemon panics and dies.
875 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
877 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
878 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
879 "pri" and "weight". Note that the previous implicit priority given by the
880 list order is no longer honoured.
882 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
885 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
886 by defining SUPPORT_SOCKS.
888 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
889 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
890 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
891 variables are renamed to proxy_{local,external}_{address,port}.
893 JH/27 Move Internationalisation support from Experimental to mainline, enabled
894 for a build by defining SUPPORT_I18N
896 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
897 of the query string, and make ${quote_redis:} do that quoting.
899 JH/29 Move Events support from Experimental to mainline, enabled by default
900 and removable for a build by defining DISABLE_EVENT.
902 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
904 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
905 cached by the daemon.
907 JH/32 Move Redis support from Experimental to mainline, enabled for a build
908 by defining LOOKUP_REDIS. The libhiredis library is required.
910 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
911 keys are given for lookup.
913 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
914 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
915 only supported when built with TLS support. The PolarSSL SHA routines
916 are still used when the TLS library is too old for convenient support.
918 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
919 openssl_options), for security. OpenSSL forces this from version 1.1.0
920 server-side so match that on older versions.
922 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
923 allocation for $value could be released as the expansion processing
924 concluded, but leaving the global pointer active for it.
926 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
927 and to use the domains and local_parts ACL conditions.
929 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
930 incorrectly not doubled on cutthrough transmission, hence seen as a
931 body-termination at the receiving system - resulting in truncated mails.
932 Commonly the sender saw a TCP-level error, and retransmitted the message
933 via the normal store-and-forward channel. This could result in duplicates
934 received - but deduplicating mailstores were liable to retain only the
935 initial truncated version.
937 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
939 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
941 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
942 we're in there, support oversigning also; bug 1309.
944 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
946 HS/04 Add support for keep_environment and add_environment options.
948 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
949 either intentional arithmetic overflow during PRNG, or testing config-
952 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
953 delivery resulted in actual delivery. Cancel cutthrough before DATA
956 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
957 rejects a recipient: pass the reject to the originator.
959 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
960 Many were false-positives and ignorable, but it's worth fixing the
963 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
964 for the new environment-manipulation done at startup. Move the routines
965 from being local to tls.c to being global via the os.c file.
967 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
968 an extract embedded as result-arg for a map, the first arg for extract
969 is unavailable so we cannot tell if this is a numbered or keyed
970 extraction. Accept either.
976 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
979 JH/02 The smtp transport option "multi_domain" is now expanded.
981 JH/03 The smtp transport now requests PRDR by default, if the server offers
984 JH/04 Certificate name checking on server certificates, when exim is a client,
985 is now done by default. The transport option tls_verify_cert_hostnames
986 can be used to disable this per-host. The build option
987 EXPERIMENTAL_CERTNAMES is withdrawn.
989 JH/05 The value of the tls_verify_certificates smtp transport and main options
990 default to the word "system" to access the system default CA bundle.
991 For GnuTLS, only version 3.0.20 or later.
993 JH/06 Verification of the server certificate for a TLS connection is now tried
994 (but not required) by default. The verification status is now logged by
995 default, for both outbound TLS and client-certificate supplying inbound
998 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1001 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1002 Status Notification (bounce) messages are now MIME format per RFC 3464.
1003 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1004 under the control of the dsn_advertise_hosts option, and routers may
1005 have a dsn_lasthop option.
1007 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1008 default, modifiable by a malware= option. The list separator for
1009 the options can now be changed in the usual way. Bug 68.
1011 JH/10 The smtp_receive_timeout main option is now expanded before use.
1013 JH/11 The incoming_interface log option now also enables logging of the
1014 local interface on delivery outgoing connections.
1016 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1017 if the interface and destination host and port all match.
1019 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1022 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1023 Patch from Andrew Lewis.
1025 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1026 now supports optional time-restrictions, weighting, and priority
1027 modifiers per server. Patch originally by <rommer@active.by>.
1029 JH/16 The spamd_address main option now supports a mixed list of local
1030 and remote servers. Remote servers can be IPv6 addresses, and
1031 specify a port-range.
1033 JH/17 Bug 68: The spamd_address main option now supports an optional
1034 timeout value per server.
1036 JH/18 Bug 1581: Router and transport options headers_add/remove can
1037 now have the list separator specified.
1039 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1042 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1045 JH/21 Support for the A6 type of dns record is withdrawn.
1047 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1048 rather than the verbs used.
1050 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1051 from 255 to 1024 chars.
1053 JH/24 Verification callouts now attempt to use TLS by default.
1055 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1056 are generic router options now. The defaults didn't change.
1058 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1059 Original patch from Alexander Shikoff, worked over by JH.
1061 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1062 files if ZCAT_COMMAND is not executable.
1064 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1066 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1068 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1069 Normally benign, it bites when the pair was led to by a CNAME;
1070 modern usage is to not canonicalize the domain to a CNAME target
1071 (and we were inconsistent anyway for A-only vs AAAA+A).
1073 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1075 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1076 when evaluating $sender_host_dnssec.
1078 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1079 $sender_helo_dnssec variable.
1081 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1083 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1085 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1087 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1088 documented as working, but never had. Support all but $spam_report.
1090 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1091 added for tls authenticator.
1093 HS/03 Add perl_taintmode main config option
1099 TL/01 When running the test suite, the README says that variables such as
1100 no_msglog_check are global and can be placed anywhere in a specific
1101 test's script, however it was observed that placement needed to be near
1102 the beginning for it to behave that way. Changed the runtest perl
1103 script to read through the entire script once to detect and set these
1104 variables, reset to the beginning of the script, and then run through
1105 the script parsing/test process like normal.
1107 TL/02 The BSD's have an arc4random API. One of the functions to induce
1108 adding randomness was arc4random_stir(), but it has been removed in
1109 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1110 function when detected.
1112 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1113 cause callback expansion.
1115 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1116 syntax errors in an expansion can be treated as a string instead of
1117 logging or causing an error, due to the internal use of bool_lax
1118 instead of bool when processing it.
1120 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1121 server certificates when making smtp deliveries.
1123 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1125 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1127 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1129 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1130 Merged patch from Sebastian Wiedenroth.
1132 JH/05 Fix results-pipe from transport process. Several recipients, combined
1133 with certificate use, exposed issues where response data items split
1134 over buffer boundaries were not parsed properly. This eventually
1135 resulted in duplicates being sent. This issue only became common enough
1136 to notice due to the introduction of connection certificate information,
1137 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1139 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1140 size buffer was used, resulting in syntax errors when an expansion
1143 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1144 version 3.3.6 or later.
1146 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1147 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1148 both become "event_action", the variables become $event_name, $event_data
1149 and $event_defer_errno. There is a new variable $verify_mode, usable in
1150 routers, transports and related events. The tls:cert event is now also
1151 raised for inbound connections, if the main configuration event_action
1154 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1155 early OCSP support, but no stapling (appears to be less than 1.0.0).
1157 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1158 server certificate names available under the smtp transport option
1159 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1162 JH/10 Time-related extraction expansions from certificates now use the main
1163 option "timezone" setting for output formatting, and are consistent
1164 between OpenSSL and GnuTLS compilations. Bug 1541.
1166 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1167 encoded parameter in the incoming message. Bug 1558.
1169 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1170 include certificate info, eximon was claiming there were spoolfile
1173 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1175 JH/14 Log delivery-related information more consistently, using the sequence
1176 "H=<name> [<ip>]" wherever possible.
1178 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1179 are problematic for Debian distribution, omit them from the release
1182 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1184 JH/16 Fix string representation of time values on 64bit time_t architectures.
1187 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1188 output list separator was used.
1193 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1194 checkers that were complaining about end of non-void function with no
1197 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1198 This was a regression introduced in 4.83 by another bugfix.
1200 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1202 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1203 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1209 TF/01 Correctly close the server side of TLS when forking for delivery.
1211 When a message was received over SMTP with TLS, Exim failed to clear up
1212 the incoming connection properly after forking off the child process to
1213 deliver the message. In some situations the subsequent outgoing
1214 delivery connection happened to have the same fd number as the incoming
1215 connection previously had. Exim would try to use TLS and fail, logging
1216 a "Bad file descriptor" error.
1218 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1219 utilities have not been installed.
1221 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1222 temporary space as the ACL may create new global variables.
1224 TL/01 LDAP support uses per connection or global context settings, depending
1225 upon the detected version of the libraries at build time.
1227 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1228 to extract and use the src ip:port in logging and expansions as if it
1229 were a direct connection from the outside internet. PPv2 support was
1230 updated based on HAProxy spec change in May 2014.
1232 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1234 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1235 Properly escape header and check for NULL return.
1237 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1240 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1242 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1243 characters in header names, implemented as a verify condition.
1244 Contributed by Michael Fischer v. Mollard.
1246 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1247 results permerror and temperror. Previous values are deprecated but
1248 still accepted. In a future release, err_perm and err_temp will be
1249 completely removed, which will be a backward incompatibility if the
1250 ACL tests for either of these two old results. Patch contributed by
1251 user bes-internal on the mailing list.
1253 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1255 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1256 selectors, in both main and reject logs.
1258 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1261 JH/07 Add malware type "sock" for talking to simple daemon.
1263 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1265 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1266 routers/transports under cutthrough routing.
1268 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1269 numbers. Touch up "bool" conditional to keep the same definition.
1271 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1273 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1275 JH/12 Expand items in router/transport headers_add or headers_remove lists
1276 individually rather than the list as a whole. Bug 1452.
1278 Required for reasonable handling of multiple headers_ options when
1279 they may be empty; requires that headers_remove items with embedded
1280 colons must have them doubled (or the list-separator changed).
1282 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1283 view the policy declared in the DMARC record. Currently, $dmarc_status
1284 is a combined value of both the record presence and the result of the
1287 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1289 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1290 dnslookup router and the smtp transport (applying to the forward
1293 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1294 of ldap servers used for a specific lookup. Patch provided by Heiko
1297 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1298 New variable $lookup_dnssec_authenticated for observability.
1300 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1301 Patch submitted by Lars Timman.
1303 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1305 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1306 Requires trusted mode and valid format message id, aborts otherwise.
1307 Patch contributed by Heiko Schlichting.
1309 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1310 certextract with support for various fields. Bug 1358.
1312 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1313 is requested by default, modifiable by smtp transport option
1316 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1317 operate on certificate variables to give certificate fingerprints
1318 Also new ${sha256:cert_variable}.
1320 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1322 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1323 Christian Aistleitner.
1325 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1327 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1328 file. Patch from Wolfgang Breyha.
1330 JH/25 Expand the coverage of the delivery $host and $host_address to
1331 client authenticators run in verify callout. Bug 1476.
1333 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1334 align with daemon_smtp_ports. Bug 72.
1336 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1337 support and error reporting did not work properly.
1339 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1340 and is readable. Patch from Andrew Colin Kissa.
1342 TL/14 Enhance documentation of ${run expansion and how it parses the
1343 commandline after expansion, particularly in the case when an
1344 unquoted variable expansion results in an empty value.
1346 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1348 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1349 by renaming away from it. Addresses GH issue 15, reported by
1352 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1354 TL/15 SECURITY: prevent double expansion in math comparison functions
1355 (can expand unsanitized data). Not remotely exploitable.
1362 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1364 PP/02 Make -n do something, by making it not do something.
1365 When combined with -bP, the name of an option is not output.
1367 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1370 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1371 $sender_host_name and config options to manage this, and basic check
1374 PP/05 DSCP support for outbound connections and control modifier for inbound.
1376 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1377 (Only plugin which currently uses this is kerberos4, which nobody should
1378 be using, but we should make it available and other future plugins might
1379 conceivably use it, even though it would break NAT; stuff *should* be
1380 using channel bindings instead).
1382 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1383 name; added for Sendmail compatibility; requires admin caller.
1384 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1385 just ignore it); requires trusted caller.
1386 Also parse but ignore: -Ac -Am -X<logfile>
1389 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1391 TL/02 Add +smtp_confirmation as a default logging option.
1393 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1394 Patch by Magnus Holmgren from 2007-02-20.
1396 TL/04 Bugzilla 1281 - Spec typo.
1397 Bugzilla 1283 - Spec typo.
1398 Bugzilla 1290 - Spec grammar fixes.
1400 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1402 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1404 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1405 mailing list by by Dmitry Isaikin.
1407 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1409 JH/02 Support "G" suffix to numbers in ${if comparisons.
1411 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1413 NM/01 Bugzilla 1197 - Spec typo
1414 Bugzilla 1196 - Spec examples corrections
1416 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1418 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1419 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1422 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1423 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1425 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1426 "acl {{name}{arg}...}", and optional args on acl condition
1429 JH/05 Permit multiple router/transport headers_add/remove lines.
1431 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1433 JH/07 Avoid using a waiting database for a single-message-only transport.
1434 Performance patch from Paul Fisher. Bugzilla 1262.
1436 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1439 JH/09 Add $headers_added variable, with content from use of ACL modifier
1440 add_header (but not yet added to the message). Bugzilla 199.
1442 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1443 Pulled from Bugzilla 817 by Wolfgang Breyha.
1445 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1447 (nb: this is the same fix as in Exim 4.80.1)
1449 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1452 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1455 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1457 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1458 advertise SMTP AUTH mechanism to us, instead of a generic
1459 protocol violation error. Also, make Exim more robust to bad
1460 data from the Dovecot auth socket.
1462 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1464 When a queue runner is handling a message, Exim first routes the
1465 recipient addresses, during which it prunes them based on the retry
1466 hints database. After that it attempts to deliver the message to
1467 any remaining recipients. It then updates the hints database using
1470 So if a recipient address works intermittently, it can get repeatedly
1471 deferred at routing time. The retry hints record remains fresh so the
1472 address never reaches the final cutoff time.
1474 This is a fairly common occurrence when a user is bumping up against
1475 their storage quota. Exim had some logic in its local delivery code
1476 to deal with this. However it did not apply to per-recipient defers
1477 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1479 This change adds a proper retry rule check during routing so that the
1480 final cutoff time is checked against the message's age. We only do
1481 this check if there is an address retry record and there is not a
1482 domain retry record; this implies that previous attempts to handle
1483 the address had the retry_use_local_parts option turned on. We use
1484 this as an approximation for the destination being like a local
1485 delivery, as in LMTP.
1487 I suspect this new check makes the old local delivery cutoff check
1488 redundant, but I have not verified this so I left the code in place.
1490 TF/02 Correct gecos expansion when From: is a prefix of the username.
1492 Test 0254 submits a message to Exim with the header
1496 When I ran the test suite under the user fanf2, Exim expanded
1497 the header to contain my full name, whereas it should have added
1498 a Resent-Sender: header. It erroneously treats any prefix of the
1499 username as equal to the username.
1501 This change corrects that bug.
1503 GF/01 DCC debug and logging tidyup
1504 Error conditions log to paniclog rather than rejectlog.
1505 Debug lines prefixed by "DCC: " to remove any ambiguity.
1507 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1509 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1510 Bug spotted by Jeremy Harris; was flawed since initial commit.
1511 Would have resulted in OCSP responses post-SNI triggering an Exim
1512 NULL dereference and crash.
1514 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1516 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1517 Bug detection, analysis and fix by Samuel Thibault.
1518 Bugzilla 1331, Debian bug #698092.
1520 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1522 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1523 Server implementation by Todd Lyons, client by JH.
1524 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1525 config variable "prdr_enable" controls whether the server
1526 advertises the facility. If the client requests PRDR a new
1527 acl_data_smtp_prdr ACL is called once for each recipient, after
1528 the body content is received and before the acl_smtp_data ACL.
1529 The client is controlled by both of: a hosts_try_prdr option
1530 on the smtp transport, and the server advertisement.
1531 Default client logging of deliveries and rejections involving
1532 PRDR are flagged with the string "PRDR".
1534 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1535 fclose(). Diagnosis by Todd Lyons.
1537 PP/17 Update configure.default to handle IPv6 localhost better.
1538 Patch by Alain Williams (plus minor tweaks).
1541 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1542 This is now consistent with GnuTLS, and is now documented: the
1543 previous undocumented portable approach to treating the option as
1544 unset was to force an expansion failure. That still works, and
1545 an empty string is now equivalent.
1547 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1548 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1549 not performing validation itself.
1551 PP/20 Added force_command boolean option to pipe transport.
1552 Patch from Nick Koston, of cPanel Inc.
1554 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1557 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1559 PP/21 Fix eximon continuous updating with timestamped log-files.
1560 Broken in a format-string cleanup in 4.80, missed when I repaired the
1561 other false fix of the same issue.
1562 Report and fix from Heiko Schlichting.
1565 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1566 Report from Prashanth Katuri.
1568 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1569 It's SecureTransport, so affects any MacOS clients which use the
1570 system-integrated TLS libraries, including email clients.
1572 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1573 using a MIME ACL for non-SMTP local injection.
1574 Report and assistance in diagnosis by Warren Baker.
1576 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1578 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1580 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1581 last id that failed so it may be referenced in subsequent ACL's.
1583 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1586 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1587 ldap library initialization, allowing self-signed CA's to be
1588 used. Also properly sets require_cert option later in code by
1589 using NULL (global ldap config) instead of ldap handle (per
1590 session). Bug diagnosis and testing by alxgomz.
1592 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1593 the src/util/ subdirectory.
1595 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1596 renamed to Transport Post Delivery Action by Jeremy Harris, as
1599 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1600 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1601 redis_servers = needs to be configured which will be used by the redis
1602 lookup. Patch from Warren Baker, of The Packet Hub.
1604 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1606 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1607 hostname or reverse DNS when processing a host list. Used suggestions
1608 from multiple comments on this bug.
1610 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1612 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1613 Missed a few lines, added it to make the runtest require no keyboard
1616 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1617 contains upper case chars. Make router use caseful_local_part.
1619 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1620 support when GnuTLS has been built with p11-kit.
1626 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1628 This, or similar/improved, will also be change PP/11 of 4.82.
1634 PP/01 Handle short writes when writing local log-files.
1635 In practice, only affects FreeBSD (8 onwards).
1636 Bugzilla 1053, with thanks to Dmitry Isaikin.
1638 NM/01 Bugzilla 949 - Documentation tweak
1640 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1643 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1645 PP/02 Implemented gsasl authenticator.
1647 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1649 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1650 `pkg-config foo` for cflags/libs.
1652 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1653 with rest of GSASL and with heimdal_gssapi.
1655 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1656 `pkg-config foo` for cflags/libs for the TLS implementation.
1658 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1659 properties get this fed in as external SSF. A number of robustness
1660 and debugging improvements to the cyrus_sasl authenticator.
1662 PP/08 cyrus_sasl server now expands the server_realm option.
1664 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1665 Patch by Jeremy Harris.
1667 PP/10 Added dbmjz lookup type.
1669 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1671 PP/12 MAIL args handles TAB as well as SP, for better interop with
1672 non-compliant senders.
1673 Analysis and variant patch by Todd Lyons.
1675 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1676 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1677 Patch from Dirk Mueller <dmueller@suse.com>
1679 PP/13 tls_peerdn now print-escaped for spool files.
1680 Observed some $tls_peerdn in wild which contained \n, which resulted
1681 in spool file corruption.
1683 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1684 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1685 or write after TLS renegotiation, which otherwise led to messages
1688 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1689 as a tracking header (ie: a signed header comes before the signature).
1690 Patch from Wolfgang Breyha.
1692 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1693 comma-sep list; embedded commas doubled.
1695 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1697 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1699 Report and patch from Dmitry Banschikov.
1701 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1702 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1703 failures. We appear to now support TLS1.1+ with Exim.
1705 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1706 lets Exim select keys and certificates based upon TLS SNI from client.
1707 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1708 before an outbound SMTP session. New log_selector, +tls_sni.
1710 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1711 NULL dereference. Report and patch from Alun Jones.
1713 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1714 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1715 Not seeing resolver debug output on NetBSD, but suspect this is a
1716 resolver implementation change.
1718 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1719 Left warnings. Added "eximon gdb" invocation mode.
1721 PP/21 Defaulting "accept_8bitmime" to true, not false.
1723 PP/22 Added -bw for inetd wait mode support.
1725 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1726 locate the relevant includes and libraries. Made this the default.
1728 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1729 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1731 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1732 This may cause build issues on older platforms.
1734 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1735 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1736 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1737 Added SNI support via GnuTLS too.
1738 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1740 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1742 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1743 Applied second patch from Janne, implementing suggestion to default
1744 multiple-strings-in-record handling to match SPF spec.
1746 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1748 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1749 read-only, out of scope).
1750 Patch from Wolfgang Breyha, report from Stuart Northfield.
1752 PP/29 Fix three issues highlighted by clang analyser static analysis.
1753 Only crash-plausible issue would require the Cambridge-specific
1754 iplookup router and a misconfiguration.
1755 Report from Marcin Mirosław.
1757 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1759 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1760 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1761 As part of this, removing so much warning spew let me fix some minor
1762 real issues in debug logging.
1764 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1765 assignment on my part. Fixed.
1767 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1768 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1769 Janne Snabb (who went above and beyond: thank you).
1771 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1772 string otherwise requires a connection and a bunch more work and it's
1773 relatively easy to get wrong. Should also expose TLS library linkage
1776 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1777 64-bit ${eval} (JH/03).
1779 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1780 GNU libc to support some of the 64-bit stuff, should not lead to
1781 conflicts. Defined before os.h is pulled in, so if a given platform
1782 needs to override this, it can.
1784 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1785 protection layer was required, which is not implemented.
1786 Bugzilla 1254, patch from Wolfgang Breyha.
1788 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1789 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1790 tls_dhparam take prime identifiers. Also unbreak combination of
1791 OpenSSL+DH_params+TLSSNI.
1793 PP/39 Disable SSLv2 by default in OpenSSL support.
1799 PP/01 Solaris build fix for Oracle's LDAP libraries.
1800 Bugzilla 1109, patch from Stephen Usher.
1802 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1804 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1807 TF/02 Fix a couple more cases where we did not log the error message
1808 when unlink() failed. See also change 4.74-TF/03.
1810 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1811 lock up or crash if it happened to be inside a call to libc when it
1812 got a SIGUSR1 from exiwhat.
1814 The SIGUSR1 handler appends the current process status to the process
1815 log which is later printed by exiwhat. It used to use the general
1816 purpose logging code to do this, but several functions it calls are
1817 not safe for signals.
1819 The new output code in the SIGUSR1 handler is specific to the process
1820 log, and simple enough that it's easy to inspect for signal safety.
1821 Removing some special cases also simplifies the general logging code.
1822 Removing the spurious timestamps from the process log simplifies
1825 TF/04 Improved ratelimit ACL condition.
1827 The /noupdate option has been deprecated in favour of /readonly which
1828 has clearer semantics. The /leaky, /strict, and /readonly update modes
1829 are mutually exclusive. The update mode is no longer included in the
1830 database key; it just determines when the database is updated. (This
1831 means that when you upgrade Exim will forget old rate measurements.)
1833 Exim now checks that the per_* options are used with an update mode that
1834 makes sense for the current ACL. For example, when Exim is processing a
1835 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1836 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1837 must specify per_mail/readonly. If you omit the update mode it defaults to
1838 /leaky where that makes sense (as before) or /readonly where required.
1840 The /noupdate option is now undocumented but still supported for
1841 backwards compatibility. It is equivalent to /readonly except that in
1842 ACLs where /readonly is required you may specify /leaky/noupdate or
1843 /strict/noupdate which are treated the same as /readonly.
1845 A useful new feature is the /count= option. This is a generalization
1846 of the per_byte option, so that you can measure the throughput of other
1847 aggregate values. For example, the per_byte option is now equivalent
1848 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1850 The per_rcpt option has been generalized using the /count= mechanism
1851 (though it's more complicated than the per_byte equivalence). When it is
1852 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1853 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1854 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1855 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1856 Note that using per_rcpt with a non-readonly update mode in more than
1857 one ACL will cause the recipients to be double-counted. (The per_mail
1858 and per_byte options don't have this problem.)
1860 The handling of very low rates has changed slightly. If the computed rate
1861 is less than the event's count (usually one) then this event is the first
1862 after a long gap. In this case the rate is set to the same as this event's
1863 count, so that the first message of a spam run is counted properly.
1865 The major new feature is a mechanism for counting the rate of unique
1866 events. The new per_addr option counts the number of different
1867 recipients that someone has sent messages to in the last time period. It
1868 behaves like per_rcpt if all the recipient addresses are different, but
1869 duplicate recipient addresses do not increase the measured rate. Like
1870 the /count= option this is a general mechanism, so the per_addr option
1871 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1872 example, measure the rate that a client uses different sender addresses
1873 with the options per_mail/unique=$sender_address. There are further
1874 details in the main documentation.
1876 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1878 TF/06 Removed a few PCRE remnants.
1880 TF/07 Automatically extract Exim's version number from tags in the git
1881 repository when doing development or release builds.
1883 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1884 Bugzilla 879. Patch from Paul Fisher.
1886 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1887 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1890 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1892 PP/05 Variable $av_failed, true if the AV scanner deferred.
1893 Bugzilla 1078. Patch from John Horne.
1895 PP/06 Stop make process more reliably on build failure.
1896 Bugzilla 1087. Patch from Heiko Schlittermann.
1898 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1899 Bugzilla 1089. Patch from Heiko Schlittermann.
1901 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1902 Bugzilla 1131. Patch from Holger Weiß.
1904 PP/09 Handle IPv6 addresses with SPF.
1905 Bugzilla 860. Patch from Wolfgang Breyha.
1907 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1909 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1912 PP/11 match_* no longer expand right-hand-side by default.
1913 New compile-time build option, EXPAND_LISTMATCH_RHS.
1914 New expansion conditions, "inlist", "inlisti".
1916 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1918 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1920 PP/14 fix log_write() format string regression from TF/03.
1921 Bugzilla 1152. Patch from Dmitry Isaikin.
1927 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1929 PP/02 Harmonised TLS library version reporting; only show if debugging.
1930 Layout now matches that introduced for other libraries in 4.74 PP/03.
1932 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1934 PP/04 New "dns_use_edns0" global option.
1936 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1939 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1940 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1942 TK/01 Updated PolarSSL code to 0.14.2.
1943 Bugzilla 1097. Patch from Andreas Metzler.
1945 PP/07 Catch divide-by-zero in ${eval:...}.
1946 Fixes bugzilla 1102.
1948 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1951 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1952 format-string attack -- SECURITY: remote arbitrary code execution.
1954 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1955 time unintentionally subject to list matching rules, letting the header
1956 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1957 arbitrary string expansion). This allowed for information disclosure.
1959 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1960 INT_MIN/-1 -- value coerced to INT_MAX.
1966 NM/01 Workaround for PCRE version dependency in version reporting
1969 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1970 This fixes portability to compilers other than gcc, notably
1971 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1973 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1974 makefiles for portability to HP-UX and POSIX correctness.
1976 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1977 Also via indented variable definition in the Makefile.
1978 (Debugging by Oliver Heesakkers).
1980 PP/02 Restore caching of spamd results with expanded spamd_address.
1981 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1983 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1984 Improves build reliability. Fix from: Frank Elsner
1986 NM/02 Fix wide character breakage in the rfc2047 coding
1987 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1989 NM/03 Allow underscore in dnslist lookups
1990 Fixes bug 1026. Patch from Graeme Fowler
1992 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1993 Code patches from Adam Ciarcinski of NetBSD.
1995 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1998 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1999 is logged, to avoid truncation. Patch from John Horne.
2001 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2002 Patch from Jakob Hirsch.
2004 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2005 SQL string expansion failure details.
2006 Patch from Andrey Oktyabrski.
2008 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2009 Patch from Simon Arlott.
2011 PP/09 New lookups functionality failed to compile on old gcc which rejects
2012 extern declarations in function scope.
2013 Patch from Oliver Fleischmann
2015 PP/10 Use sig_atomic_t for flags set from signal handlers.
2016 Check getgroups() return and improve debugging.
2017 Fixed developed for diagnosis in bug 927 (which turned out to be
2020 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2021 Patch from Mark Zealey.
2023 PP/12 Bugzilla 1056: Improved spamd server selection.
2024 Patch from Mark Zealey.
2026 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2027 Based on patch from Heiko Schlittermann.
2029 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2030 Patch from Uwe Doering, sign-off by Michael Haardt.
2032 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2033 variable declaration deep within a block. Bug and patch from
2036 PP/15 lookups-Makefile IRIX compatibility coercion.
2038 PP/16 Make DISABLE_DKIM build knob functional.
2040 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2041 Patch by Simon Arlott
2043 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2044 variable argument macros. Our copy now differs from upstream.
2050 TF/01 Failure to get a lock on a hints database can have serious
2051 consequences so log it to the panic log.
2053 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2054 controlled using the smtp_confirmation log selector.
2056 TF/03 Include the error message when we fail to unlink a spool file.
2058 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2059 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2060 for maintaining out-of-tree patches for some time.
2062 PP/01 Bugzilla 139: Documentation and portability issues.
2063 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2064 Handle per-OS dynamic-module compilation flags.
2066 PP/02 Let /dev/null have normal permissions.
2067 The 4.73 fixes were a little too stringent and complained about the
2068 permissions on /dev/null. Exempt it from some checks.
2069 Reported by Andreas M. Kirchwitz.
2071 PP/03 Report version information for many libraries, including
2072 Exim version information for dynamically loaded libraries. Created
2073 version.h, now support a version extension string for distributors
2074 who patch heavily. Dynamic module ABI change.
2076 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2077 privilege escalation vulnerability whereby the Exim run-time user
2078 can cause root to append content of the attacker's choosing to
2081 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2084 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2085 If dropping privileges for untrusted macros, we disabled normal logging
2086 on the basis that it would fail; for the Exim run-time user, this is not
2087 the case, and it resulted in successful deliveries going unlogged.
2088 Fixed. Reported by Andreas Metzler.
2094 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2095 only prepend for the Resent-* case. Fixes regression introduced in
2096 Exim 4.70 by NM/22 for Bugzilla 607.
2098 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2099 increasing numbers of administrators be bitten by this.
2101 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2103 PP/03 Bugzilla 994: added openssl_options main configuration option.
2105 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2107 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2109 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2111 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2112 without a peer certificate, leading to a segfault because of an
2113 assumption that peers always have certificates. Be a little more
2114 paranoid. Problem reported by Martin Tscholak.
2116 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2117 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2118 NB: ClamAV planning to remove STREAM in "middle of 2010".
2119 CL also introduces -bmalware, various -d+acl logging additions and
2120 more caution in buffer sizes.
2122 PP/09 Implemented reverse_ip expansion operator.
2124 PP/10 Bugzilla 937: provide a "debug" ACL control.
2126 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2128 PP/12 Bugzilla 973: Implement --version.
2130 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2132 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2134 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2136 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2137 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2138 ignore trailing whitespace.
2140 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2142 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2145 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2146 Notification from Dr Andrew Aitchison.
2148 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2149 ExtendedDetectionInfo response format.
2150 Notification from John Horne.
2152 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2155 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2156 XSL and documented dependency on system catalogs, with examples of how
2159 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2162 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2163 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2164 configuration file which is writeable by the Exim user or group.
2166 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2167 of configuration files to cover files specified with the -C option if
2168 they are going to be used with root privileges, not just the default
2171 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2172 option (effectively making it always true).
2174 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2175 files to be used while preserving root privileges.
2177 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2178 that rogue child processes cannot use them.
2180 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2181 run-time user, instead of root.
2183 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2184 Exim run-time user without dropping privileges.
2186 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2187 result string, instead of calling string_vformat() twice with the same
2190 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2191 for other users. Others should always drop root privileges if they use
2192 -C on the command line, even for a whitelisted configure file.
2194 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2196 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2202 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2203 $data_path, and $header_path variables; fixed documentation bugs and
2206 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2207 exipick to access non-standard spools, including the "frozen" queue
2210 NM/01 Bugzilla 965: Support mysql stored procedures.
2211 Patch from Alain Williams
2213 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2215 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2216 Patch from Andreas Metzler
2218 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2219 Patch from Kirill Miazine
2221 NM/05 Bugzilla 671: Added umask to procmail example.
2223 JJ/03 installed exipick 20100323.0, fixing doc bug
2225 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2226 directory. Notification and patch from Dan Rosenberg.
2228 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2230 TK/02 Improve log output when DKIM signing operation fails.
2232 MH/01 Treat the transport option dkim_domain as a colon separated
2233 list, not as a single string, and sign the message with each element,
2234 omitting multiple occurences of the same signer.
2236 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2237 Bugzilla 985, 986. Patch by Simon Arlott
2239 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2240 Patch by Simon Arlott
2242 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2243 MBX locking. Notification from Dan Rosenberg.
2249 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2251 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2253 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2255 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2257 NM/04 Bugzilla 915: Flag broken perl installation during build.
2263 TK/01 Added patch by Johannes Berg that expands the main option
2264 "spamd_address" if it starts with a dollar sign.
2266 TK/02 Write list of recipients to X-Envelope-Sender header when building
2267 the mbox-format spool file for content scanning (suggested by Jakob
2270 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2271 (http://www.dcc-servers.net/) support via dccifd. Activated by
2272 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2274 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2275 by Mark Daniel Reidel <mr@df.eu>.
2277 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2278 When building exim an external PCRE library is now needed -
2279 PCRE is a system library on the majority of modern systems.
2280 See entry on PCRE_LIBS in EDITME file.
2282 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2283 conversation. Added nologin parameter to request.
2284 Patch contributed by Kirill Miazine.
2286 TF/01 Do not log submission mode rewrites if they do not change the address.
2288 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2290 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2291 log files in place. Contributed by Roberto Lima.
2293 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2295 TF/03 Bugzilla 615: When checking the local_parts router precondition
2296 after a local_part_suffix or local_part_prefix option, Exim now
2297 does not use the address's named list lookup cache, since this
2298 contains cached lookups for the whole local part.
2300 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2301 Robert Millan. Documentation is in experimental-spec.txt.
2303 TF/04 Bugzilla 668: Fix parallel build (make -j).
2305 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2307 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2308 Patch provided by Jan Srzednicki.
2310 TF/05 Leading white space used to be stripped from $spam_report which
2311 wrecked the formatting. Now it is preserved.
2313 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2314 that they are available at delivery time.
2316 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2318 TF/08 TLS error reporting now respects the incoming_interface and
2319 incoming_port log selectors.
2321 TF/09 Produce a more useful error message if an SMTP transport's hosts
2322 setting expands to an empty string.
2324 NM/06 Bugzilla 744: EXPN did not work under TLS.
2325 Patch provided by Phil Pennock.
2327 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2328 Patch provided by Richard Godbee.
2330 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2331 acl_smtp_notquit, added index entry.
2333 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2334 Patch provided by Eugene Bujak.
2336 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2337 accept(). Patch provided by Maxim Dounin.
2339 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2340 Patch provided by Phil Pennock.
2342 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2344 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2345 Patch provided by Brad "anomie" Jorsch.
2347 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2348 Patch provided by Dean Brooks.
2350 TK/05 Add native DKIM support (does not depend on external libraries).
2352 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2353 Patch provided by Graeme Fowler.
2355 NM/16 Bugzilla 851: Documentation example syntax fix.
2357 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2359 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2362 NM/19 Bugzilla 745: TLS version reporting.
2363 Patch provided by Phil Pennock.
2365 NM/20 Bugzilla 167: bool: condition support.
2366 Patch provided by Phil Pennock.
2368 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2369 clients. Patch provided by Phil Pennock.
2371 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2372 Patch provided by Brad "anomie" Jorsch.
2374 NM/23 Bugzilla 687: Fix misparses in eximstats.
2375 Patch provided by Heiko Schlittermann.
2377 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2378 Patch provided by Heiko Schlittermann.
2380 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2381 plus update to original patch.
2383 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2385 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2386 Patch provided by David Brownlee.
2388 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2390 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2392 NM/30 Bugzilla 888: TLS documentation bugfixes.
2394 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2396 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2397 Unlike the original bugzilla I have changed all shell scripts in src tree.
2399 NM/33 Bugzilla 898: Transport filter timeout fix.
2400 Patch by Todd Rinaldo.
2402 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2403 Patch by Serge Demonchaux.
2405 NM/35 Bugzilla 39: Base64 decode bug fixes.
2406 Patch by Jakob Hirsch.
2408 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2410 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2412 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2414 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2420 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2421 ALT-N's libdkim that I have put here:
2422 http://duncanthrax.net/exim-experimental/
2424 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2425 were called 'true' and it seems that C99 defines that as a reserved
2426 keyword to be used with 'bool' variable types. That means you could
2427 not include C99-style headers which use bools without triggering
2428 build errors in sieve.c.
2430 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2431 as mailq or other aliases. Changed the --help handling significantly
2432 to do whats expected. exim_usage() emits usage/help information.
2434 SC/01 Added the -bylocaldomain option to eximstats.
2436 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2438 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2440 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2446 PH/01 Another patch from the Sieve maintainer.
2448 PH/02 When an IPv6 address is converted to a string for single-key lookup
2449 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2450 dots are used instead of colons so that keys in lsearch files need not
2451 contain colons. This was done some time before quoting was made available
2452 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2453 (notated using the quote facility) so as to distinguish them from IPv4
2454 keys. This meant that lookups for IP addresses in host lists did not work
2455 for iplsearch lookups.
2457 This has been fixed by arranging for IPv6 addresses to be expressed with
2458 colons if the lookup type is iplsearch. This is not incompatible, because
2459 previously such lookups could never work.
2461 The situation is now rather anomalous, since one *can* have colons in
2462 ordinary lsearch keys. However, making the change in all cases is
2463 incompatible and would probably break a number of configurations.
2465 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2468 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2469 conversion specification without a maximum field width, thereby enabling
2470 a rogue spamd server to cause a buffer overflow. While nobody in their
2471 right mind would setup Exim to query an untrusted spamd server, an
2472 attacker that gains access to a server running spamd could potentially
2473 exploit this vulnerability to run arbitrary code as the Exim user.
2475 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2476 $primary_hostname instead of what libspf2 thinks the hosts name is.
2478 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2479 a directory entry by the name of the lookup key. Previously, if a
2480 symlink pointed to a non-existing file or a file in a directory that
2481 Exim lacked permissions to read, a lookup for a key matching that
2482 symlink would fail. Now it is enough that a matching directory entry
2483 exists, symlink or not. (Bugzilla 503.)
2485 PH/03 The body_linecount and body_zerocount variables are now exported in the
2488 PH/04 Added the $dnslist_matched variable.
2490 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2491 This means they are set thereafter only if the connection becomes
2494 PH/06 Added the client_condition to authenticators so that some can be skipped
2495 by clients under certain conditions.
2497 PH/07 The error message for a badly-placed control=no_multiline_responses left
2498 "_responses" off the end of the name.
2500 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2502 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2503 (without spaces) instead of just copying the configuration text.
2505 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2507 PH/11 Added $max_received_linelength.
2509 PH/12 Added +ignore_defer and +include_defer to host lists.
2511 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2512 way in which PCRE > 7.0 is built.
2514 PH/14 Implemented queue_only_load_latch.
2516 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2517 MAIL command. The effect was to mangle the value on 64-bit systems.
2519 PH/16 Another patch from the Sieve maintainer.
2521 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2523 PH/18 If a system quota error occurred while trying to create the file for
2524 a maildir delivery, the message "Mailbox is full" was not appended to the
2525 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2526 only to a quota excession during the actual writing of the file.
2528 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2529 characters?) which causes problems in log lines. The DN values are now
2530 passed through string_printing() before being added to log lines.
2532 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2533 and InterBase are left for another time.)
2535 PH/21 Added message_body_newlines option.
2537 PH/22 Guard against possible overflow in moan_check_errorcopy().
2539 PH/23 POSIX allows open() to be a macro; guard against that.
2541 PH/24 If the recipient of an error message contained an @ in the local part
2542 (suitably quoted, of course), incorrect values were put in $domain and
2543 $local_part during the evaluation of errors_copy.
2549 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2550 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2553 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2554 issue a MAIL command.
2556 PH/02 In an ACL statement such as
2558 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2560 if a client was not listed at all, or was listed with a value other than
2561 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2562 the condition was not true (as it should be), so access was not denied.
2563 The bug was that the ! inversion was incorrectly passed on to the second
2564 item. This has been fixed.
2566 PH/03 Added additional dnslists conditions == and =& which are different from
2567 = and & when the dns lookup returns more than one IP address.
2569 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2570 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2572 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2573 FSYNC, which compiles an option called disable_fsync that allows for
2574 bypassing fsync(). The documentation is heavily laced with warnings.
2576 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2578 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2579 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2580 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2581 including adding "make clean"; (3) Added -fPIC when compiling the test
2582 dynamically loaded module, to get rid of a warning.
2584 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2585 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2586 = 0s. The bug is otherwise harmless.
2588 PH/07 There was a bug in the dovecot authenticator such that the value of
2589 $auth1 could be overwritten, and so not correctly preserved, after a
2590 successful authentication. This usually meant that the value preserved by
2591 the server_setid option was incorrect.
2593 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2595 PH/09 Installed PCRE release 7.0.
2597 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2598 run for batched SMTP input. It is now run at the start of every message
2599 in the batch. While fixing this I discovered that the process information
2600 (output by running exiwhat) was not always getting set for -bs and -bS
2601 input. This is fixed, and it now also says "batched" for BSMTP.
2603 PH/11 Added control=no_pipelining.
2605 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2606 patch, slightly modified), and move the expansion of helo_data till after
2607 the connection is made in the smtp transport (so it can use these
2610 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2612 PH/14 Added log_selector = +pid.
2614 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2616 PH/16 Add ${if forany and ${if forall.
2618 PH/17 Added dsn_from option to vary the From: line in DSNs.
2620 PH/18 Flush SMTP output before performing a callout, unless control =
2621 no_callout_flush is set.
2623 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2624 was true (the default) a successful delivery failed to delete the retry
2625 item, thus causing premature timeout of the address. The bug is now
2628 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2630 PH/21 Long custom messages for fakedefer and fakereject are now split up
2631 into multiline responses in the same way that messages for "deny" and
2632 other ACL rejections are.
2634 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2635 with slight modification.
2637 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2638 draft, changing the syntax and factoring some duplicate code".
2640 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2641 for deliveries of the second and subsequent messages over the same SMTP
2644 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2645 ${reduce, with only minor "tidies".
2647 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2649 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2650 expansion side effects.
2652 PH/27 When a message times out after an over-quota error from an Exim-imposed
2653 quota, the bounce message says "mailbox is full". This message was not
2654 being given when it was a system quota that was exceeded. It now should
2657 MH/03 Made $recipients available in local_scan(). local_scan() already has
2658 better access to the recipient list through recipients_list[], but
2659 $recipients can be useful in postmaster-provided expansion strings.
2661 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2662 in the case of a MAIL command with additional options following the
2663 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2664 were accidentally chopped off.
2666 PH/29 SMTP synchronization checks are implemented when a command is read -
2667 there is a check that no more input is waiting when there shouldn't be
2668 any. However, for some commands, a delay in an ACL can mean that it is
2669 some time before the response is written. In this time, more input might
2670 arrive, invalidly. So now there are extra checks after an ACL has run for
2671 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2672 pipelining has not been advertised.
2674 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2676 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2677 returned in the SMTP response when smtp_return_error_details was false.
2678 This has been fixed.
2680 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2681 instead of the C I/O that was originally supplied, because problems were
2682 reported on Solaris.
2684 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2685 Exim which did not show up earlier: it was assuming that a call to
2686 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2687 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2688 was a macro that became an assignment, so it seemed to work. This has
2689 changed to a proper function call with a void return, hence the compile
2690 error. Exim's code has been fixed.
2692 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2695 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2697 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2699 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2700 because it turns out to be too big or there is a timeout) there is no
2701 "Completed" line in the log. When some messages of this type were
2702 selected by exigrep, they were listed as "not completed". Others were
2703 picked up by some special patterns. I have improved the selection
2704 criteria to be more general.
2706 PH/38 The host_find_failed option in the manualroute router can now be set
2707 to "ignore", to completely ignore a host whose IP address cannot be
2708 found. If all hosts are ignored, the behaviour is controlled by the new
2709 host_all_ignored option.
2711 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2712 homing or because of multiple MX records with /mx) generated more than
2713 one IP address, and the following item turned out to be the local host,
2714 all the secondary addresses of the first item were incorrectly removed
2715 from the list, along with the local host and any following hosts (which
2716 is what is supposed to happen).
2718 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2719 whoever called Exim into the -H file. In the case of the daemon it was
2720 behaving confusingly. When first started, it used values for whoever
2721 started the daemon, but after a SIGHUP it used the Exim user (because it
2722 calls itself on a restart). I have changed the code so that it now always
2725 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2726 message are rejected with the same error (e.g. no authentication or bad
2727 sender address), and a DATA command is nevertheless sent (as can happen
2728 with PIPELINING or a stupid MUA), the error message that was given to the
2729 RCPT commands is included in the rejection of the DATA command. This is
2730 intended to be helpful for MUAs that show only the final error to their
2733 PH/42 Another patch from the Sieve maintainer.
2735 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2736 Eximstats - Fixed some broken HTML links and added missing column headers
2738 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2739 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2741 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2743 PH/43 Yet another patch from the Sieve maintainer.
2745 PH/44 I found a way to check for a TCP/IP connection going away before sending
2746 the response to the final '.' that terminates a message, but only in the
2747 case where the client has not sent further data following the '.'
2748 (unfortunately, this is allowed). However, in many cases there won't be
2749 any further data because there won't be any more messages to send. A call
2750 to select() can be used: if it shows that the input is "ready", there is
2751 either input waiting, or the socket has been closed. An attempt to read
2752 the next input character can distinguish the two cases. Previously, Exim
2753 would have sent an OK response which the client would never have see.
2754 This could lead to message repetition. This fix should cure that, at
2755 least in a lot of common cases.
2757 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2758 advertised in response to EHLO.
2764 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2765 fixed by 4.65/MH/01 (is this a record?) are fixed:
2767 (i) An empty string was always treated as zero by the numeric comparison
2768 operators. This behaviour has been restored.
2770 (ii) It is documented that the numeric comparison operators always treat
2771 their arguments as decimal numbers. This was broken in that numbers
2772 starting with 0 were being interpreted as octal.
2774 While fixing these problems I realized that there was another issue that
2775 hadn't been noticed. Values of message_size_limit (both the global option
2776 and the transport option) were treated as octal if they started with 0.
2777 The documentation was vague. These values are now always treated as
2778 decimal, and I will make that clear in the documentation.
2784 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2785 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2788 MH/01 Don't check that the operands of numeric comparison operators are
2789 integers when their expansion is in "skipping" mode (fixes bug
2790 introduced by 4.64-PH/07).
2792 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2793 child addresses, Exim now panics and dies. Previously, because the count
2794 is held in a short int, deliveries were likely to be lost. As such a
2795 large number of recipients for a single message is ridiculous
2796 (performance will be very, very poor), I have chosen to impose a limit
2797 rather than extend the field.
2803 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2804 leftover -K file (the existence of which was triggered by #402).
2805 While we were at it, introduced process PID as part of the -K
2806 filename. This should rule out race conditions when creating
2809 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2810 processing considerably. Previous code took too long for large mails,
2811 triggering a timeout which in turn triggers #401.
2813 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2814 in the DK code in transports.c. sendfile() is not really portable,
2815 hence the _LINUX specificness.
2817 TF/01 In the add_headers option to the mail command in an Exim filter,
2818 there was a bug that Exim would claim a syntax error in any
2819 header after the first one which had an odd number of characters
2822 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2823 callout verification, Exim cached a "reject" for the entire domain. This
2824 is correct for most verifications, but it is not correct for a recipient
2825 verification with use_sender or use_postmaster set, because in that case
2826 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2827 case of MAIL FROM:<> rejection from other early rejections (e.g.
2828 rejection of HELO). When verifying a recipient using a non-null MAIL
2829 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2830 Whatever the result of the callout, the value of the domain cache is
2831 left unchanged (for any other kind of callout, getting as far as trying
2832 RCPT means that the domain itself is ok).
2834 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2837 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2838 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2839 session, the daemon ignores ECONNECT errors and logs others; it now
2840 ignores EPIPE as well.
2842 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2843 (quoted-printable decoding).
2845 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2846 later the small subsequent patch to fix an introduced bug.
2848 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2850 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2852 PH/08 An error is now given if message_size_limit is specified negative.
2854 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2855 to be given (somewhat) arbitrary names.
2857 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2860 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2861 miscellaneous code fixes
2863 PH/10 Added the log_reject_target ACL modifier to specify where to log
2866 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2867 hostname. This is wrong, because it relates to the incoming message (and
2868 probably the interface on which it is arriving) and not to the outgoing
2869 callout (which could be using a different interface). This has been
2870 changed to use the value of the helo_data option from the smtp transport
2871 instead - this is what is used when a message is actually being sent. If
2872 there is no remote transport (possible with a router that sets up host
2873 addresses), $smtp_active_hostname is used.
2875 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2876 tweaks were necessary in order to get it to work (see also 21 below):
2877 (a) The code assumed that strncpy() returns a negative number on buffer
2878 overflow, which isn't the case. Replaced with Exim's string_format()
2880 (b) There were several signed/unsigned issues. I just did the minimum
2881 hacking in of casts. There is scope for a larger refactoring.
2882 (c) The code used strcasecmp() which is not a standard C function.
2883 Replaced with Exim's strcmpic() function.
2884 (d) The code set only $1; it now sets $auth1 as well.
2885 (e) A simple test gave the error "authentication client didn't specify
2886 service in request". It would seem that Dovecot has changed its
2887 interface. Fortunately there's a specification; I followed it and
2888 changed what the client sends and it appears to be working now.
2890 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2893 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2894 address A is aliased to B and C, where B exists and C does not. Without
2895 -v the output is "A verified" because verification stops after a
2896 successful redirection if more than one address is generated. However,
2897 with -v the child addresses are also verified. Exim was outputting "A
2898 failed to verify" and then showing the successful verification for C,
2899 with its parentage. It now outputs "B failed to verify", showing B's
2900 parentage before showing the successful verification of C.
2902 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2903 look up a TXT record in a specific list after matching in a combined
2906 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2907 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2908 they consult the DNS. I had assumed they would set it the way they
2909 wanted; and indeed my experiments on Linux seem to show that in some
2910 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2911 To be on the safe side, however, I have now made the interface to
2912 host_find_byname() similar to host_find_bydns(), with an argument
2913 containing the DNS resolver options. The host_find_byname() function now
2914 sets these options at its start, just as host_find_bydns() does. The smtp
2915 transport options dns_qualify_single and dns_search_parents are passed to
2916 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2917 of host_find_byname() use the default settings of RES_DEFNAMES
2918 (qualify_single) but not RES_DNSRCH (search_parents).
2920 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2921 spool_read_header() do less string testing, by means of a preliminary
2922 switch on the second character of optional "-foo" lines. (This is
2923 overdue, caused by the large number of possibilities that now exist.
2924 Originally there were few.) While I was there, I also converted the
2925 str(n)cmp tests so they don't re-test the leading "-" and the first
2926 character, in the hope this might squeeze out yet more improvement.
2928 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2929 flag allowing group syntax was set by the header_syntax check but not
2930 turned off, possible causing trouble later; (2) The flag was not being
2931 set at all for the header_verify test, causing "group"-style headers to
2932 be rejected. I have now set it in this case, and also caused header_
2933 verify to ignore an empty address taken from a group. While doing this, I
2934 came across some other cases where the code for allowing group syntax
2935 while scanning a header line wasn't quite right (mostly, not resetting
2936 the flag correctly in the right place). These bugs could have caused
2937 trouble for malformed header lines. I hope it is now all correct.
2939 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2940 with the "reply" argument non-NULL. The code, however (which originally
2941 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2942 but it didn't always do it. This confused somebody who was copying the
2943 code for some other use. I have removed all the tests.
2945 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2946 feature that was used to support insecure browsers during the U.S. crypto
2947 embargo. It requires special client support, and Exim is probably the
2948 only MTA that supported it -- and would never use it because real RSA is
2949 always available. This code has been removed, because it had the bad
2950 effect of slowing Exim down by computing (never used) parameters for the
2951 RSA_EXPORT functionality.
2953 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2954 authenticator to fail if there's a tab character in the incoming data
2955 (there should never be unless someone is messing about, as it's supposed
2956 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2957 if the connection is using TLS or if the remote IP is the same as the
2958 local IP, and the "valid-client-cert option" if a client certificate has
2961 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2962 authenticators. This can be used for authorization after authentication
2963 succeeds. (In the case of plaintext, it servers for both authentication
2966 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2967 if any retry times were supplied.
2969 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2970 connection, where there is no client IP address to check. In this
2971 situation, the verify now always succeeds.
2973 PH/25 Applied John Jetmore's -Mset patch.
2975 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2977 PH/27 In a string expansion for a processed (not raw) header when multiple
2978 headers of the same name were present, leading whitespace was being
2979 removed from all of them, but trailing whitespace was being removed only
2980 from the last one. Now trailing whitespace is removed from each header
2981 before concatenation. Completely empty headers in a concatenation (as
2982 before) are ignored.
2984 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2985 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2987 PH/29 [Removed. This was a change that I later backed out, and forgot to
2988 correct the ChangeLog entry (that I had efficiently created) before
2989 committing the later change.]
2991 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2992 address errors (4xx response to RCPT) over the same connection as other
2993 messages routed to the same hosts. Such deliveries are always "forced",
2994 so retry times are not inspected. This resulted in far too many retries
2995 for the affected addresses. The effect occurred only when there were more
2996 hosts than the hosts_max_try setting in the smtp transport when it had
2997 the 4xx errors. Those hosts that it had tried were not added to the list
2998 of hosts for which the message was waiting, so if all were tried, there
2999 was no problem. Two fixes have been applied:
3001 (i) If there are any address or message errors in an SMTP delivery, none
3002 of the hosts (tried or untried) are now added to the list of hosts
3003 for which the message is waiting, so the message should not be a
3004 candidate for sending over the same connection that was used for a
3005 successful delivery of some other message. This seems entirely
3006 reasonable: after all the message is NOT "waiting for some host".
3007 This is so "obvious" that I'm not sure why it wasn't done
3008 previously. Hope I haven't missed anything, but it can't do any
3009 harm, as the worst effect is to miss an optimization.
3011 (ii) If, despite (i), such a delivery is accidentally attempted, the
3012 routing retry time is respected, so at least it doesn't keep
3013 hammering the server.
3015 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3016 in ${readsocket because some servers need this prod.
3018 PH/32 Added some extra debug output when updating a wait-xxx database.
3020 PH/33 The hint "could be header name not terminated by colon", which has been
3021 given for certain expansion errors for a long time, was not being given
3022 for the ${if def:h_colon_omitted{... case.
3024 PH/34 The spec says: "With one important exception, whenever a domain list is
3025 being scanned, $domain contains the subject domain." There was at least
3026 one case where this was not true.
3028 PH/35 The error "getsockname() failed: connection reset by peer" was being
3029 written to the panic log as well as the main log, but it isn't really
3030 panic-worthy as it just means the connection died rather early on. I have
3031 removed the panic log writing for the ECONNRESET error when getsockname()
3034 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3035 runs only) independently of the message's sender address. This meant
3036 that, if the 4xx error was in fact related to the sender, a different
3037 message to the same recipient with a different sender could confuse
3038 things. In particular, this can happen when sending to a greylisting
3039 server, but other circumstances could also provoke similar problems.
3040 I have changed the default so that the retry time for these errors is now
3041 based a combination of the sender and recipient addresses. This change
3042 can be overridden by setting address_retry_include_sender=false in the
3045 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3046 remote server are returned as part of bounce messages. This was not
3047 happening for LMTP over a pipe (the lmtp transport), but now it is the
3048 same for both kinds of LMTP.
3050 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3051 in header lines that were in fact CNAMEs. This is no longer the case.
3053 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3054 and queue runs started by the daemon processed all messages. This has
3055 been fixed so that -R and -S can now usefully be given with -q<time>.
3057 PH/40 Import PCRE release 6.7 (fixes some bugs).
3059 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3061 PH/42 Give an error if -q is specified more than once.
3063 PH/43 Renamed the variables $interface_address and $interface_port as
3064 $received_ip_address and $received_port, to make it clear that these
3065 values apply to message reception, and not to the outgoing interface when
3066 a message is delivered. (The old names remain recognized, of course.)
3068 PH/44 There was no timeout on the connect() call when using a Unix domain
3069 socket in the ${readsocket expansion. There now is.
3071 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3072 be meaningful with "accept".
3074 SC/01 Eximstats V1.43
3075 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3077 SC/02 Eximstats V1.44
3078 Use a glob alias rather than an array ref in the generated
3079 parser. This improves both readability and performance.
3081 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3082 Collect SpamAssassin and rejection statistics.
3083 Don't display local sender or destination tables unless
3084 there is data to show.
3085 Added average volumes into the top table text output.
3087 SC/04 Eximstats V1.46
3088 Collect data on the number of addresses (recipients)
3089 as well as the number of messages.
3091 SC/05 Eximstats V1.47
3092 Added 'Message too big' to the list of mail rejection
3093 reasons (thanks to Marco Gaiarin).
3095 SC/06 Eximstats V1.48
3096 Mainlog lines which have GMT offsets and are too short to
3097 have a flag are now skipped.
3099 SC/07 Eximstats V1.49 (Alain Williams)
3100 Added the -emptyok flag.
3102 SC/08 Eximstats V1.50
3103 Fixes for obtaining the IP address from reject messages.
3105 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3106 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3107 whitespace changes from 4.64-PH/27
3109 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3112 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3113 are found, allow negative numbers in numeric criteria)
3115 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3117 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3118 to match changes made in 4.64-PH/43
3120 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3122 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3123 "domain" is actually an IP address, and force a failure. This locks out
3124 those revolvers/nameservers that support "A-for-A" lookups, in
3125 contravention of the specifications.
3127 PH/48 When a host name was looked up from an IP address, and the subsequent
3128 forward lookup of the name timed out, the host name was left in
3129 $sender_host_name, contrary to the specification.
3131 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3132 restricted to single-key lookups, Exim was not diagnosing an error if
3133 * or *@ was used with a query-style lookup.
3135 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3137 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3138 long ago, but noone interested enough thought of it. Let's just say that
3139 the "1.1" means that there are some new functions that weren't there at
3140 some point in the past.
3142 PH/51 Error processing for expansion failure of helo_data from an smtp
3143 transport during callout processing was broken.
3145 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3146 tested/used via the -bh/-bhc/-bs options.
3148 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3149 bug, fixed in subsequent PCRE releases).
3151 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3152 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3154 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3160 SC/01 Use a glob alias rather than an array ref in eximstats generated
3161 parser. This improves both readability and performance.
3163 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3164 Don't display local sender or destination tables in eximstats unless
3165 there is data to show.
3166 Added average volumes into the eximstats top table text output.
3168 SC/03 Collect data on the number of addresses (recipients) as well
3169 as the number of messages in eximstats.
3171 TF/01 Correct an error in the documentation for the redirect router. Exim
3172 does (usually) call initgroups() when daemonizing.
3174 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3175 with consistent privilege compared to when running as a daemon.
3177 TF/03 Note in the spec that $authenticated_id is not set for local
3178 submissions from trusted users.
3180 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3181 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3183 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3184 by adding some example configuration directives to the default
3185 configuration file. A little bit of work is required to uncomment the
3186 directives and define how usernames and passwords are checked, but
3187 there is now a framework to start from.
3189 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3190 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3191 without this. I don't know how relevant this is to other LDAP libraries.
3193 PH/02 Add the verb name to the "unknown ACL verb" error.
3195 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3197 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3199 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3200 directory not expanded when it should be if an expanded home directory
3201 was set for the address (which is overridden by the transport).
3203 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3206 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3207 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3208 because it is too late at that time, and has no effect.
3210 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3211 security issue with \' (bugzilla #107). I could not use the
3212 PQescapeStringConn() function, because it needs a PGconn value as one of
3215 PH/08 When testing addresses using -bt, indicate those final addresses that
3216 are duplicates that would not cause an additional delivery. At least one
3217 person was confused, thinking that -bt output corresponded to deliveries.
3218 (Suppressing duplicates isn't a good idea as you lose the information
3219 about possibly different redirections that led to the duplicates.)
3221 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3222 systems where poll() doesn't work, in particular OS X.
3224 PH/10 Added more information to debugging output for retry time not reached.
3226 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3227 operations in malware.c.
3229 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3232 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3233 syslog_duplication set false, log lines that would normally be written
3234 both the the main log and to the reject log were not written to syslog at
3237 PH/14 In the default configuration, change the use of "message" in ACL warn
3238 statements to "add_header".
3240 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3241 not followed by a command (e.g. "seen endif").
3243 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3244 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3247 PH/17 Added extra conditions to the default value of delay_warning_condition
3251 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3252 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3253 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3256 The Auto-Submitted: and various List- headers are standardised, whereas I
3257 don't think Precedence: ever was.
3259 PH/18 Refactored debugging code in route_finduser() to show more information,
3260 in particular, the error code if getpwnam() issues one.
3262 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3263 This is apparently needed in addition to the PH/07 change above to avoid
3264 any possible encoding problems.
3266 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3267 but not after initializing Perl.
3269 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3270 output them only if debugging. By default they are written stderr,
3271 apparently, which is not desirable.
3273 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3276 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3279 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3281 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3282 authenticated or an ident call has been made. Suppress the default
3283 values for $authenticated_id and $authenticated_sender (but permit -oMai
3284 and -oMas) when testing with -bh.
3286 PH/24 Re-jigged the order of the tests in the default configuration so that the
3287 tests for valid domains and recipients precede the DNS black list and CSA
3288 tests, on the grounds that those ones are more expensive.
3290 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3291 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3292 This bug exists in every version of Exim that I still have, right back to
3295 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3296 However, an attempt to turn on case-sensitivity in a regex key by
3297 including (?-i) didn't work because the subject string was already
3298 lowercased, and the effects were non-intuitive. It turns out that a
3299 one-line patch can be used to allow (?-i) to work as expected.
3305 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3306 other effects) broke the use of negated acl sub-conditions.
3308 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3311 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3312 "Deny" causes Exim to reject the incoming connection with a 554 error.
3313 Unfortunately, if there is a major crisis, such as a disk failure,
3314 tcp-wrappers gives "deny", whereas what one would like would be some
3315 kind of temporary error. A kludge has been added to help with this.
3316 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3317 554 error is used if errno is still zero or contains ENOENT (which occurs
3318 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3321 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3323 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3324 errors. Otherwise a message that provokes a temporary error (when other
3325 messages do not) can cause a whole host to time out.
3327 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3328 addresses were routed directly to files or pipes from a redirect router.
3329 File deliveries just didn't batch; pipe deliveries might have suffered
3332 PH/06 A failure to get a lock for a hints database would erroneously always say
3333 "Failed to get write lock", even when it was really a read lock.
3335 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3336 of 0600. This has been changed to use the value of the lockfile_mode
3337 option (which defaults to 0600).
3339 PH/08 Applied small patch from the Sieve maintainer.
3341 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3342 folder from quota calculations, a direct delivery into this folder messed
3343 up the contents of the maildirsize file. This was because the regex was
3344 used only to exclude .Trash (or whatever) when the size of the mailbox
3345 was calculated. There was no check that a delivery was happening into an
3346 excluded directory. This bug has been fixed by ignoring all quota
3347 processing for deliveries into excluded directories.
3349 PH/10 Added the maildirfolder_create_regex option to appendfile.
3355 PH/01 The code for finding all the local interface addresses on a FreeBSD
3356 system running IPv6 was broken. This may well have applied to all BSD
3357 systems, as well as to others that have similar system calls. The broken
3358 code found IPv4 interfaces correctly, but gave incorrect values for the
3359 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3360 that it would not match correctly against @[] and not recognize the IPv6
3363 PH/02 The ipliteral router was not recognizing addresses of the form user@
3364 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3366 PH/03 Added disable_ipv6.
3368 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3369 decoded form, because it is most often used to construct To: headers
3370 lines in autoreplies, and the decoded form may well be syntactically
3371 invalid. However, $reply_address has leading white space removed, and all
3372 newlines turned into spaces so that the autoreply transport does not
3375 PH/05 If group was specified without a user on a router, and no group or user
3376 was specified on a transport, the group from the router was ignored.
3378 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3379 for visible compile-time settings that can be used to change these
3380 numbers, for those that want even more. Backwards compatibility with old
3381 spool files has been maintained. However, going back to a previous Exim
3382 release will lost any variables that are in spool files.
3384 PH/07 Two small changes when running in the test harness: increase delay when
3385 passing a TCP/IP connection to a new process, in case the original
3386 process has to generate a bounce, and remove special handling of
3387 127.0.0.2 (sic), which is no longer necessary.
3389 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3390 be the same on different OS.
3392 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3395 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3396 whether --show-vars was specified or not
3398 JJ/02 exipick: Added support for new ACL variable spool format introduced
3401 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3402 syntactically invalid From: or Reply-to: line, and a filter used this to
3403 generate an autoreply, and therefore failed to obtain an address for the
3404 autoreply, Exim could try to deliver to a non-existent relative file
3405 name, causing unrelated and misleading errors. What now happens is that
3406 it logs this as a hard delivery error, but does not attempt to create a
3409 PH/11 The exinext utility has a -C option for testing purposes, but although
3410 the given file was scanned by exinext itself; it wasn't being passed on
3411 when Exim was called.
3413 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3414 an end-of-file indication when reading a command response.
3416 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3417 compiled. In many other places in Exim, IPv6 addresses are always
3418 recognized, so I have changed this. It also means that IPv4 domain
3419 literals of the form [IPV4:n.n.n.n] are now always recognized.
3421 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3422 used if the router is not running as root, for example, when verifying at
3423 ACL time, or when using -bh. The debugging output from this situation was
3424 non-existent - all you got was a failure to exec. I have made two
3427 (a) Failures to set uid/gid, the current directory, or a process leader
3428 in a subprocess such as that created by queryprogram now generate
3429 suitable debugging output when -d is set.
3431 (b) The queryprogram router detects when it is not running as root,
3432 outputs suitable debugging information if -d is set, and then runs
3433 the subprocess without attempting to change uid/gid.
3435 PH/15 Minor change to Makefile for building test_host (undocumented testing
3438 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3439 additional section of a DNS packet that returns MX or SRV records.
3440 Instead, it always explicitly searches for A/AAAA records. This avoids
3441 major problems that occur when a DNS server includes only records of one
3442 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3443 fixed another bug: if SRV records were looked up and the corresponding
3444 address records were *not* found in the additional section, the port
3445 values from the SRV records were lost.
3447 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3448 using the correct key (the original address) when searching the retry
3449 rules in order to find which one to use for generating the retry hint.
3451 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3452 adding the default one. Similarly, if it contains a Reply-To: header, the
3453 errors_reply_to option, if set, is not used.
3455 PH/19 When calculating a retry time, Exim used to measure the "time since
3456 failure" by looking at the "first failed" field in the retry record. Now
3457 it does not use this if it is later than than the arrival time of the
3458 message. Instead it uses the arrival time. This makes for better
3459 behaviour in cases where some deliveries succeed, thus re-setting the
3460 "first failed" field. An example is a quota failure for a huge message
3461 when small messages continue to be delivered. Without this change, the
3462 "time since failure" will always be short, possible causing more frequent
3463 delivery attempts for the huge message than are intended.
3464 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3466 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3467 $1, $2, $3) because the numerical variables can be reset during some
3468 expansion items (e.g. "match"), thereby losing the authentication data.
3470 PH/21 Make -bV show the size of off_t variables so that the test suite can
3471 decide whether to run tests for quotas > 2G.
3473 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3474 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3475 filecount value is greater than 2G or if a quota value is greater than 2G
3476 on a system where the size of off_t is not greater than 4, a panic error
3479 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3480 never match. The debug and -bh output now contains an explicit error
3481 message indicating a malformed IPv4 address or mask.
3483 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3484 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3485 PH/23 above applies.
3487 PH/25 Do not write to syslog when running in the test harness. The only
3488 occasion when this arises is a failure to open the main or panic logs
3489 (for which there is an explicit test).
3491 PH/26 Added the /no_tell option to "control=freeze".
3493 PH/27 If a host name lookup failed very early in a connection, for example, if
3494 the IP address matched host_lookup and the reverse lookup yielded a name
3495 that did not have a forward lookup, an error message of the form "no IP
3496 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3497 could be logged. Now it outputs the IP address instead of "NULL".
3499 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3500 allows the sender and the authenticated sender to be set when
3501 submitting a message from within Exim. Since child_open_exim() is
3502 documented for local_scan(), the new function should be too.
3504 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3505 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3506 results in an empty string is now treated as unset.
3508 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3510 PH/31 Added sender_verify_fail logging option.
3512 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3513 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3514 tidied the source and removed it altogether.
3516 PH/33 When a queue run was abandoned because the load average was too high, a
3517 log line was always written; now it is written only if the queue_run log
3518 selector is set. In addition, the log line for abandonment now contains
3519 information about the queue run such as the pid. This is always present
3520 in "start" and "stop" lines but was omitted from the "abandon" line.
3522 PH/34 Omit spaces between a header name and the colon in the error message that
3523 is given when verify = headers_syntax fails (if there are lots of them,
3524 the message gets confusing).
3526 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3527 names, as there are now some PTR records that contain slashes. This check
3528 is only to protect against broken name servers that fall over on strange
3529 characters, so the fact that it applies to all lookups doesn't matter.
3531 PH/36 Now that the new test suite is complete, we can remove some of the
3532 special code in Exim that was needed for the old test suite. For example,
3533 sorting DNS records because real resolvers return them in an arbitrary
3534 order. The new test suite's fake resolver always returns records in the
3537 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3538 bounces) except when queue_only is set, to avoid logging races between
3539 the different processes.
3541 PH/38 Panic-die if .include specifies a non-absolute path.
3543 PH/39 A tweak to the "H" retry rule from its user.
3545 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3546 a label. They prevented compilation on older perls.
3548 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3549 a warning to be raised on newish perls.
3551 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3552 on queue. Changes to match documented behaviour of showing count of
3553 messages matching specified criteria.
3555 PH/40 Changed the default ident timeout from 30s to 5s.
3557 PH/41 Added support for the use of login_cap features, on those BSD systems
3558 that have them, for controlling the resources used by pipe deliveries.
3560 PH/42 The content-scanning code uses fopen() to create files in which to put
3561 message data. Previously it was not paying any attention to the mode of
3562 the files. Exim runs with umask(0) because the rest of the code creates
3563 files with open(), and sets the required mode explicitly. Thus, these
3564 files were ending up world-writeable. This was not a big issue, because,
3565 being within the spool directory, they were not world-accessible. I have
3566 created a function called modefopen, which takes an additional mode
3567 argument. It sets umask(777), creates the file, chmods it to the required
3568 mode, then resets the umask. All the relevant calls to fopen() in the
3569 content scanning code have been changed to use this function.
3571 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3572 to 24 hours. This avoids potential overflow problems when processing G
3573 and H retry rules. I suspect nobody ever tinkers with this value.
3575 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3577 PH/45 When the plaintext authenticator is running as a client, the server's
3578 challenges are checked to ensure they are valid base64 strings. By
3579 default, the authentication attempt is cancelled if an invalid string is
3580 received. Setting client_ignore_invalid_base64 true ignores these errors.
3581 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3582 they are received. Thus, the responses can be made to depend on the
3583 challenges. If an invalid string is ignored, an empty string is placed in
3586 PH/46 Messages that are created by the autoreply transport now contains a
3587 References: header, in accordance with RFCs 2822 and 3834.
3589 PH/47 Added authenticated_sender_force to the smtp transport.
3591 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3593 PH/49 Installed latest patch from the Sieve maintainer.
3595 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3596 was also set, the appendfile transport was unnecessarily scanning a
3597 directory of message files (e.g. for maildir delivery) to find the count
3598 of files (along with the size), even though it did not need this
3599 information. It now does the scan only if it needs to find either the
3600 size of the count of files.
3602 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3604 PH/52 Two bugs concerned with error handling when the smtp transport is
3607 (i) Exim was not creating retry information for temporary errors given
3608 for individual recipients after the DATA command when the smtp transport
3609 was used in LMTP mode. This meant that they could be retried too
3610 frequently, and not timed out correctly.
3612 (ii) Exim was setting the flag that allows error details to be returned
3613 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3614 recipients that were returned after the DATA command.
3616 PH/53 This is related to PH/52, but is more general: for any failing address,
3617 when detailed error information was permitted to be returned to the
3618 sender, but the error was temporary, then after the final timeout, only
3619 "retry timeout exceeded" was returned. Now it returns the full error as
3620 well as "retry timeout exceeded".
3622 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3623 do this, and (what is worse) MTAs that accept it.
3625 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3626 will now be deprecated.
3628 PH/56 New os.c-cygwin from the Cygwin maintainer.
3630 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3631 formats (previously only available in exim formats via -bpr, -bpru,
3632 and -bpra. Now also available in native and exiqgrep formats)
3634 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3635 with very large, slow to parse queues
3637 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3639 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3641 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3642 responses to authentication challenges, though it was showing the
3643 challenges; (ii) I've removed the CR characters from the debug output for
3646 PH/58 Allow for the insertion of a newline as well as a space when a string
3647 is turned into more than one encoded-word during RFC 2047 encoding. The
3648 Sieve code now uses this.
3650 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3651 data_4xx, lost_connection, tls_required.
3653 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3654 message was being sent as an SMTP response.
3656 PH/61 Add -l and -k options to exicyclog.
3658 PH/62 When verifying, if an address was redirected to one new address, so that
3659 verification continued, and the new address failed or deferred after
3660 having set something in $address_data, the value of $address_data was not
3661 passed back to the ACL. This was different to the case when no
3662 redirection occurred. The value is now passed back in both cases.
3664 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3665 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3666 use login_cap.h, so on its own it isn't the distinguishing feature. The
3667 new name refers directly to the setclassresources() function.
3669 PH/65 Added configuration files for NetBSD3.
3671 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3673 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3674 is preferred over IPv4.
3676 PH/68 The bounce_return_message and bounce_return_body options were not being
3677 honoured for bounces generated during the reception of non-SMTP messages.
3678 In particular, this applied to messages rejected by the ACL. This bug has
3679 been fixed. However, if bounce_return_message is true and bounce_return_
3680 body is false, the headers that are returned for a non-SMTP message
3681 include only those that have been read before the error was detected.
3682 (In the case of an ACL rejection, they have all been read.)
3684 PH/69 The HTML version of the specification is now built in a directory called
3685 spec_html instead of spec.html, because the latter looks like a path with
3686 a MIME-type, and this confuses some software.
3688 PH/70 Catch two compiler warnings in sieve.c.
3690 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3691 function verify_get_ident() calls ip_connect() to connect a socket, but
3692 if the "connect()" function timed out, ip_connect() used to close the
3693 socket. However, verify_get_ident() also closes the socket later, and in
3694 between Exim writes to the log, which may get opened at this point. When
3695 the socket was closed in ip_connect(), the log could get the same file
3696 descriptor number as the socket. This naturally causes chaos. The fix is
3697 not to close the socket in ip_connect(); the socket should be closed by
3698 the function that creates it. There was only one place in the code where
3699 this was missing, in the iplookup router, which I don't think anybody now
3700 uses, but I've fixed it anyway.
3702 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3703 well as to direct DNS lookups. Otherwise the handling of names in host
3704 lists is inconsistent and therefore confusing.
3710 PH/01 Two changes to the default runtime configuration:
3712 (1) Move the checks for relay_from_hosts and authenticated clients from
3713 after to before the (commented out) DNS black list checks.
3715 (2) Add control=submission to the relay_from_hosts and authenticated
3716 clients checks, on the grounds that messages accepted by these
3717 statements are most likely to be submissions.
3719 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3721 (1) Generate an error if the third argument for the ${prvs expansion is
3724 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3727 (3) Reset the variables that are obtained from the first argument of
3728 ${prvscheck and used in the second argument before leaving the code,
3729 because their memory is reclaimed, so using them afterwards may do
3732 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3733 one (it's much easier than Tom thought :-).
3735 (5) Because of (4), we can now allow for the use of $prvscheck_result
3736 inside the third argument.
3738 PH/03 For some reason, the default setting of PATH when running a command from
3739 a pipe transport was just "/usr/bin". I have changed it to
3742 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3743 anything to be listed in the output from -bV.
3745 PH/05 When a filter generated an autoreply, the entire To: header line was
3746 quoted in the delivery log line, like this:
3748 => >A.N.Other <ano@some.domain> <original@ddress> ...
3750 This has been changed so that it extracts the operative address. There
3751 may be more than one such address. If so, they are comma-separated, like
3754 => >ano@some.domain,ona@other.domain <original@ddress> ...
3756 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3757 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3758 not being looked up in the rDNS to get a host name, Exim was showing the
3759 IP address twice in Received: lines, even though the IP addresses were
3760 identical. For example:
3762 Received: from [1.2.3.4] (helo=[1.2.3.4])
3764 However, if the real host name was known, it was omitting the HELO data
3765 if it matched the actual IP address. This has been tidied up so that it
3766 doesn't show the same IP address twice.
3768 PH/07 When both +timestamp and +memory debugging was on, the value given by
3769 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3770 called by the debug printing, thereby overwriting the timestamp buffer.
3771 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3773 PH/08 When the original message was included in an autoreply transport, it
3774 always said "this is a copy of the message, including all the headers",
3775 even if body_only or headers_only was set. It now gives an appropriate
3778 PH/09 Applied a patch from the Sieve maintainer which:
3780 o fixes some comments
3781 o adds the (disabled) notify extension core
3782 o adds some debug output for the result of if/elsif tests
3783 o points to the current vacation draft in the documentation
3784 and documents the missing references header update
3788 o fixes a bug in processing the envelope test (when testing
3789 multiple envelope elements, the last element determined the
3792 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3793 Electronic Mail") by including:
3795 Auto-submitted: auto-generated
3797 in the messages that it generates (bounce messages and others, such as
3798 warnings). In the case of bounce messages for non-SMTP messages, there was
3799 also a typo: it was using "Auto_submitted" (underscore instead of
3800 hyphen). Since every message generated by Exim is necessarily in response
3801 to another message, thes have all been changed to:
3803 Auto-Submitted: auto-replied
3805 in accordance with these statements in the RFC:
3807 The auto-replied keyword:
3809 - SHOULD be used on messages sent in direct response to another
3810 message by an automatic process,
3812 - MUST NOT be used on manually-generated messages,
3814 - MAY be used on Delivery Status Notifications (DSNs) and Message
3815 Disposition Notifications (MDNs),
3817 - MUST NOT be used on messages generated by automatic or periodic
3818 processes, except for messages which are automatic responses to
3821 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3822 to the default Received: header definition.
3824 PH/12 Added log selector acl_warn_skipped (default on).
3826 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3827 variables because (a) they are in the wrong storage pool and (b) even if
3828 they were copied, it wouldn't work properly because of the caching.
3830 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3831 checking when decoding. Apparently there are clients that generate
3832 overlong encoded strings. Why am I not surprised?
3834 PH/15 If the first argument of "${if match_address" was not empty, but did not
3835 contain an "@" character, Exim crashed. Now it writes a panic log message
3836 and treats the condition as false.
3838 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3840 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3841 extension "envelope-auth". The code is finished and in agreement with
3842 other implementations, but there is no documentation so far and in fact,
3843 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3844 not changing the active code.
3846 Print executed "if" and "elsif" statements when debugging is used. This
3847 helps a great deal to understand what a filter does.
3849 Document more things not specified clearly in RFC3028. I had all this
3850 sorted out, when out of a sudden new issues came to my mind. Oops."
3852 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3855 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3856 canonical form (as documented). However, after a host name lookup from
3857 the IP address, check_host() was doing a simple string comparison with
3858 addresses acquired from the DNS when checking that the found name did
3859 have the original IP as one of its addresses. Since any found IPv6
3860 addresses are likely to be in abbreviated form, the comparison could
3861 fail. Luckily, there already exists a function for doing the comparison
3862 by converting both addresses to binary, so now that is used instead of
3863 the text comparison.
3865 PH/20 There was another similar case to PH/19, when a complete host name was
3866 given in a host list; looking up its IP address could give an abbreviated
3867 form, whereas the current host's name might or might not be abbreviated.
3868 The same fix has been applied.
3874 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3875 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3878 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3879 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3881 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3883 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3884 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3885 Also, for "Auto-Submitted", treat anything other than "no" as
3886 non-personal, in accordance with RFC 3834. (Previously it treated
3887 anything starting "auto-" as non-personal.)
3889 TF/01 The control=submission/name=... option had a problem with syntax
3890 errors if the name included a slash character. The /name= option
3891 now slurps the rest of the string, so it can include any characters
3892 but it must come last in the list of options (after /sender_retain
3895 PH/05 Some modifications to the interface to the fake nameserver for the new
3903 TK/01 Added the "success_on_redirect" address verification option. See
3904 NewStuff for rationale and an example.
3906 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3908 PH/02 Patch to exigrep to allow it to work on syslog lines.
3910 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3911 fread() to skip over the body file's header line, because in Cygwin the
3912 header line is locked and is inaccessible.
3914 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3915 co-exist for some time) to make it clear that it is the Exim ID that is
3916 referenced, not the Message-ID: header line.
3918 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3919 string_format() function, because snprintf() does not exist on all
3922 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3923 ${stat: expansion item.
3925 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3926 protocol synchronization error", to keep the pedants happy.
3928 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3929 well as for IRIX systems, when gcc is being used. See the host.c source
3932 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3934 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3937 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3938 transport and to the smtp transport in LMTP mode.
3940 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3942 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3943 run a filter in a subprocess. This could lead to confusion in subsequent
3944 lookups in the parent process. There should also be a search_tidyup() at
3945 the end of the subprocess.
3947 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3948 only if the host matched helo_try_verify_hosts, which caused the
3949 verification to occur when the EHLO/HELO command was issued. The ACL just
3950 tested the remembered result. Now, if a previous verification attempt has
3951 not happened, "verify = helo" does it there and then.
3953 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3955 TK/03 Fix log output including CR from clamd.
3957 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3958 exist provoked a memory error which could cause a segfault.
3960 PH/15 Installed PCRE 6.2
3962 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3964 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3965 of the problem. Specifically, suggested +O2 rather than +O1 for the
3968 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3970 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3971 record, the port was not correctly logged when the outgoing_port log
3972 selector was set (it logged the transort's default port).
3974 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3975 fallback_hosts, and "hosts" in the smtp transport.
3977 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3978 host errors such as "Connection refused".
3980 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3981 authentication with radiusclient 0.4.9:
3983 - Error returned from rc_read_config was caught wrongly
3984 - Username/password not passed on to radius server due to wrong length.
3986 The presumption is that some radiusclient API changes for 4.51/PH/17
3987 were not taken care of correctly. The code is still untested by me (my
3988 Linux distribution still has 0.3.2 of radiusclient), but it was
3989 contributed by a Radius user.
3991 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3992 expanding the "port" option of the smtp transport.
3994 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3995 while reading a MIME header. Thanks to Tom Hughes for a patch.
3997 PH/24 Include config.h inside local_scan.h so that configuration settings are
4000 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4001 that in an ACL for RCPT (for example), you can examine exactly what was
4004 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4005 commands, but it was not correctly comparing the address with the actual
4006 client host address. Thus, it would show the EHLO address in Received:
4007 header lines when this was not necessary.
4009 PH/27 Added the % operator to ${eval:}.
4011 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4012 it should be ignoring failures (because with -C, for example, it has lost
4013 privilege). It wasn't ignoring creation failures other than "already
4016 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4019 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4020 because an input line was too long, either on its own, or by virtue of
4021 too many continuations, the temporary file was not being removed, and the
4022 return code was incorrect.
4024 PH/31 Missing "BOOL" in function definition in filtertest.c.
4026 PH/32 Applied Sieve patches from the maintainer.
4028 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4030 PH/33 Added "verify = not_blind".
4032 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4033 Local/Makefile (with some defaults set). These are used in built scripts
4034 such as exicyclog, but they have never been used in the exim_install
4035 script (though there are many overriding facilities there). I have
4036 arranged that the exim_install script now takes note of these two
4039 PH/35 Installed configuration files for Dragonfly.
4041 PH/36 When a locally submitted message by a trusted user did not contain a
4042 From: header, and the sender address was obtained from -f or from an SMTP
4043 MAIL command, and the trusted user did not use -F to supply a sender
4044 name, $originator_name was incorrectly used when constructing a From:
4045 header. Furthermore, $originator_name was used for submission mode
4046 messages from external hosts without From: headers in a similar way,
4047 which is clearly wrong.
4049 PH/37 Added control=suppress_local_fixups.
4051 PH/38 When log_selector = +received_sender was set, and the addition of the
4052 sender made the log line's construction buffer exactly full, or one byte
4053 less than full, an overflow happened when the terminating "\n" was
4056 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4057 when the result of a list match is failure because a DNS lookup failed.
4059 PH/40 RM_COMMAND is now used in the building process.
4061 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4062 the "build-* directories that it finds.
4064 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4065 address in a domain literal was a prefix of an interface address.
4067 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4068 when verifying a sender address, unless rewrite_headers is false.
4070 PH/44 Wrote a long comment about why errors_to addresses are verified as
4071 recipients, not senders.
4073 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4074 the ratelimit ACL was added.
4076 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4078 PH/46 Added extra information about PostgreSQL errors to the error string.
4080 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4081 avoiding the need to install special zones in a real server. This is
4082 backwards compatible; if it can't find the fake resolver, it drops back.
4083 Thus, both old and new test suites can be run.
4085 TF/02 Added util/ratelimit.pl
4087 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4088 clock is set back in time.
4090 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4091 Candler <B.Candler@pobox.com>.
4093 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4094 OK for addresses that are the result of redirections.
4096 PH/48 A number of further additions for the benefit of the new test suite,
4097 including a fake gethostbyname() that interfaces to the fake DNS resolver
4100 TF/06 The fix for widen_domains has also been applied to qualify_single and
4101 search_parents which are the other dnslookup options that can cause
4104 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4107 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4109 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4110 service" if it encountered an MX record with an empty target hostname.
4111 The message is now "an MX or SRV record indicated no SMTP service".
4113 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4114 if the DNS of the sending site is misconfigured. This is quite a
4115 common situation. This change restores the behaviour of treating a
4116 helo verification defer as a failure.
4118 PH/51 If self=fail was set on a router, the bounce message did not include the
4119 actual error message.
4125 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4127 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4128 command itself ended in error, the underlying message about the transport
4129 filter timeout was being overwritten with the pipe command error. Now the
4130 underlying error message should be appended to the second error message.
4132 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4134 PH/02 Reduce the amount of output that "make" produces by default. Full output
4135 can still be requested.
4137 PH/03 The warning log line about a condition test deferring for a "warn" verb
4138 was being output only once per connection, rather than after each
4139 occurrence (because it was using the same function as for successful
4140 "warn" verbs). This seems wrong, so I have changed it.
4142 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4143 it should not have, which might have caused a crash in the right
4144 circumstances, but probably never did.
4146 PH/04 Installed a modified version of Tony Finch's patch to make submission
4147 mode fix the return path as well as the Sender: header line, and to
4148 add a /name= option so that you can make the user's friendly name appear
4151 TF/03 Added the control = fakedefer ACL modifier.
4153 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4154 Mark Lowes for thorough testing.
4156 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4158 TK/03 Merged latest SRS patch from Miles Wilton.
4160 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4161 with the definition in sysexits.h (which is #included earlier).
4162 Fortunately, Exim does not actually use EX_OK. The code used to try to
4163 preserve the sysexits.h value, by assuming that macro definitions were
4164 scanned for macro replacements. I have been disabused of this notion,
4165 so now the code just undefines EX_OK before #including unistd.h.
4167 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4168 in the smtp transport. When a block could not be written in a single
4169 write() function, the timeout was being re-applied to each part-write.
4170 This seems wrong - if the receiver was accepting one byte at a time it
4171 would take for ever. The timeout is now adjusted when this happens. It
4172 doesn't have to be particularly precise.
4174 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4175 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4177 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4178 without a domain if the check to <postmaster@domain> fails.
4180 SC/01 Eximstats: added -xls and the ability to specify output files
4181 (patch written by Frank Heydlauf).
4183 SC/02 Eximstats: use FileHandles for outputting results.
4185 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4187 SC/04 Eximstats: fixed display of large numbers with -nvr option
4189 SC/05 Eximstats: fixed merging of reports with empty tables.
4191 SC/06 Eximstats: added the -include_original_destination flag
4193 SC/07 Eximstats: removed tabs and trailing whitespace.
4195 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4197 TK/06 MBOX spool code: Add real "From " MBOX separator line
4198 so the .eml file is really in mbox format (even though
4199 most programs do not really care). Patch from Alex Miller.
4201 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4202 The latter is generated from $received_to and is only set if the
4203 message has one envelope recipient. SA can use these headers,
4204 obviously out-of-the-box. Patch from Alex Miller.
4206 PH/08 The ${def test on a variable was returning false if the variable's
4207 value was "0", contrary to what the specification has always said!
4208 The result should be true unless the variable is empty.
4210 PH/09 The syntax error of a character other than { following "${if
4211 def:variable_name" (after optional whitespace) was not being diagnosed.
4212 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4213 accidental colon was present, for example, could give incorrect results.
4215 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4216 result is used (not including appendfile, where other changes are about
4219 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4220 This involved changing a lot of size variables from int to off_t. It
4221 should work with maildirs and everything.
4223 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4224 spamd dying while we are connected to it.
4226 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4229 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4230 function for BDB 4.3.
4232 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4234 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4235 the sources for PCRE have been re-arranged and more files are now
4238 PH/15 The code I had for printing potentially long long variables in PH/11
4239 above was not the best (it lost precision). The length of off_t variables
4240 is now inspected at build time, and an appropriate printing format (%ld
4241 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4242 to be "long long int" or "long int". This is needed for the internal
4243 formatting function string_vformat().
4245 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4246 the configuration file to be ":syslog", then the script "guesses" where
4247 the logs files are, rather than using the compiled in default. In our
4248 case the guess is not the same as the compiled default, so the script
4249 suddenly stopped working when I started to use syslog. The patch checks
4250 to see if log_file_path is "". If so, it attempts to read it from exim
4251 with no configuration file to get the compiled in version, before it
4252 falls back to the previous guessing code."
4254 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4255 implementing BATV in an Exim configuration. See NewStuff for the gory
4258 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4259 Makefile that are specific to HP-UX.
4261 PH/18 If the "use_postmaster" option was set for a recipient callout together
4262 with the "random" option, the postmaster address was used as the MAIL
4263 FROM address for the random test, but not for the subsequent recipient
4264 test. It is now used for both.
4266 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4267 patch removes a few documentation additions to RFC 3028, because the
4268 latest draft now contains them. It adds the new en;ascii-case comparator
4269 and a new error check for 8bit text in MIME parts. Comparator and
4270 require names are now matched exactly. I enabled the subaddress
4271 extension, but it is not well tested yet (read: it works for me)."
4273 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4274 rework some of the code of TK/09 above to avoid the hardwired use of
4275 "%lld" and "long long". Replaced the call to snprintf() with a call to
4278 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4279 records point to non-existent hosts", "retry timeout exceeded", and
4280 "retry time not reached for any host after a long failure period".
4282 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4283 experimental DomainKeys support:
4285 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4286 (2) On an error such as an illegally used "control", the wrong name for
4287 the control was given.
4289 These problems did NOT occur unless DomainKeys support was compiled.
4291 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4293 PH/24 Added ${if match_ip condition.
4295 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4296 databases so that it will be absolutely obvious if a crash occurs in the
4297 DB library. This is a regular occurrence (often caused by mis-matched
4300 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4301 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4302 picked up on a user's system that detects such things. There doesn't seem
4303 to be a gcc warning option for this - only an attribute that has to be
4304 put on the function's prototype. It seems that in Fedora Core 4 they have
4305 set this on a number of new functions. No doubt there will be more in due
4308 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4309 specify a transport. However, if an address that was verified by such a
4310 router was the subject of a callout, Exim crashed because it tried to
4311 read the rcpt_include_affixes from the non-existent transport. Now it
4312 just assumes that the setting of that option is false. This bug was
4313 introduced by 4.51/PH/31.
4315 PH/28 Changed -d+all to exclude +memory, because that information is very
4316 rarely of interest, but it makes the output a lot bigger. People tend to
4317 do -d+all out of habit.
4319 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4320 code in os-type was giving problems when libc.so lives in lib64, like on
4323 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4324 aren't the modern standard, and it seems that some systems' include files
4325 don't always have them. Exim was already checking for some of the newer
4326 ones like T_AAAA, and defining it itself. I've added checks for all the
4327 record types that Exim uses.
4329 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4330 not automatically generating a new one, as it is supposed to. This
4331 prevented TLS from working. If the file did exist, but contained invalid
4332 data, a new version was generated, as expected. It was only the case of a
4333 non-existent file that was broken.
4335 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4336 with a change in libdomainkeys > 0.64.
4338 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4339 from DNS. If the selector record carries the flag, it now has
4340 precedence over the domain-wide flag.
4342 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4344 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4345 the use of a port name that isn't defined in /etc/services) occurred, the
4346 message was deferred as in a normal delivery, and thus remained on the
4347 spool, instead of being failed because of the mua_wrapper setting. This
4348 is now fixed, and I tidied up some of the mua_wrapper messages at the
4351 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4352 the messages in a hash of arrays rather than using individual hashes.
4353 This is a bit cleaner and results in dramatic memory savings, albeit
4354 at a slight CPU cost.
4356 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4357 as requested by Marc Sherman.
4359 SC/10 Eximstats: added histograms for user specified patterns as requested
4362 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4364 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4365 fopen() in the content-scanning modules that did not already have it.
4371 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4372 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4374 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4376 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4378 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4379 to test Sieve filters that use "vacation".
4381 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4382 that changes the way the GnuTLS parameters are stored in the cache file.
4383 The new format can be generated externally. For backward compatibility,
4384 if the data in the cache doesn't make sense, Exim assumes it has read an
4385 old-format file, and it generates new data and writes a new file. This
4386 means that you can't go back to an older release without removing the
4389 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4390 work if there are any delivery delays because "one_time" forces the
4391 parent to be marked "delivered", so its unseen clone is never tried
4392 again. For this reason, Exim now forbids the simultaneous setting of
4395 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4396 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4397 entry if you want to know the details. The fix, however, neglected to
4398 consider the case when local delivery batching is involved. The test for
4399 "previously delivered" was not happening when checking to see if an
4400 address could be batched with a previous (undelivered) one; under
4401 certain circumstances this could lead to multiple deliveries to the same
4404 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4405 in its include files, and this causes problems building Exim.
4407 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4408 header_syntax) but Exim was just ignoring anything given after a slash.
4409 In particular, this caused confusion with an attempt to use "verify =
4410 reverse_host_lookup/defer_ok". An error is now given when options are
4411 supplied for verify items that do not have them. (Maybe reverse_host_
4412 lookup should have a defer_ok option, but that's a different point.)
4414 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4415 defined by RFC 821) to 2048, because there were problems with some AUTH
4416 commands, and RFC 1869 says the size should be increased for extended
4417 SMTP commands that take arguments.
4419 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4422 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4423 "unknown" error; now it says that the functionality isn't in the binary.
4425 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4426 an address' error message when a string expansion fails (syntax or
4427 whatever). Otherwise the password may appear in the log. Following change
4428 PH/42 below, there is no longer a chance of it appearing in a bounce
4431 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4433 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4434 of its addresses was ever tried. (Bugzilla bug #2.)
4436 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4437 the result incorrectly in the debug output. (It correctly added a newline
4438 to what was transported.)
4440 TF/01 Added $received_time.
4442 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4443 commented out examples of how to interface to a virus scanner and to
4444 SpamAssassin. Also added commented examples of av_scanner and
4445 spamd_address settings.
4447 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4448 and controls are allowed in which ACLs. There were a couple of minor
4449 errors. Some of the entries in the conditions table (which is a table of
4450 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4451 negation of where the condition IS allowed.
4453 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4455 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4456 header file does not have a version number, so I've had to invent a new
4457 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4458 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4459 radiusclient), but it was contributed by a Radius user.
4461 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4462 files or directories, for OpenSSL.
4464 PH/19 When an Exim process that is running as root has to create an Exim log
4465 file, it does so in a subprocess that runs as exim:exim so as to get the
4466 ownership right at creation (otherwise, other Exim processes might see
4467 the file with the wrong ownership). There was no test for failure of this
4468 fork() call, which would lead to the process getting stuck as it waited
4469 for a non-existent subprocess. Forks do occasionally fail when resources
4470 run out. I reviewed all the other calls to fork(); they all seem to check
4473 PH/20 When checking for unexpected SMTP input at connect time (before writing
4474 the banner), Exim was not dealing correctly with a non-positive return
4475 from the read() function. If the client had disconnected by this time,
4476 the result was a log entry for a synchronization error with an empty
4477 string after "input=" when read() returned zero. If read() returned -1
4478 (an event I could not check), uninitialized data bytes were printed.
4479 There were reports of junk text (parts of files, etc) appearing after
4482 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4484 PH/22 Added support for macro redefinition, and (re)definition in between
4485 driver and ACL definitions.
4487 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4488 forgetting to use the resulting value; it was using the unexpanded value.
4490 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4491 hadn't been configured. The fix is from Juergen Kreileder, who
4492 understands it better than I do:
4494 "Here's what I see happening with three configured cyrus_sasl
4495 authenticators configured (plain, login, cram-md5):
4497 On startup auth_cyrus_sasl_init() gets called for each of these.
4498 This means three calls to sasl_listmech() without a specified mech_list.
4499 => SASL tests which mechs of all available mechs actually work
4500 => three warnings about OTP not working
4501 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4503 With the patch, sasl_listmech() also gets called three times. But now
4504 SASL's mech_list option is set to the server_mech specified in the the
4505 authenticator. Or in other words, the answer from sasl_listmech()
4506 gets limited to just the mech you're testing for (which is different
4508 => the return list contains just 'plain' or 'login', 'cram-md5' or
4509 nothing depending on the value of ob->server_mech.
4511 I've just tested the patch: Authentication still works fine,
4512 unavailable mechs specified in the exim configuration are still
4513 caught, and the auth.log warnings about OTP are gone."
4515 PH/25 When debugging is enabled, the contents of the command line are added
4516 to the debugging output, even when log_selector=+arguments is not
4519 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4520 answer is "GNU", and only if the return is "GNU/something" is the answer
4523 PH/27 $acl_verify_message is now set immediately after the failure of a
4524 verification in an ACL, and so is available in subsequent modifiers. In
4525 particular, the message can be preserved by coding like this:
4527 warn !verify = sender
4528 set acl_m0 = $acl_verify_message
4530 Previously, $acl_verify_message was set only while expanding "message"
4531 and "log_message" when a very denied access.
4533 PH/28 Modified OS/os.c-Linux with
4535 -#ifndef OS_LOAD_AVERAGE
4536 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4538 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4539 nomenclature these days.)
4541 PH/29 Installed patch from the Sieve maintainer that adds the options
4542 sieve_useraddress and sieve_subaddress to the redirect router.
4544 PH/30 In these circumstances:
4545 . Two addresses routed to the same list of hosts;
4546 . First host does not offer TLS;
4547 . First host accepts first address;
4548 . First host gives temporary error to second address;
4549 . Second host offers TLS and a TLS session is established;
4550 . Second host accepts second address.
4551 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4552 and peerdn, if requested) that were in fact used only for the second
4555 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4556 attention to any local part prefix or suffix that was matched by the
4557 router that accepted the address. It now behaves in the same way as it
4558 does for delivery: the affixes are removed from the local part unless
4559 rcpt_include_affixes is set on the transport.
4561 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4562 timeout during the DATA phase of an incoming message.
4564 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4565 applied a patch sanctioned by the Sieve maintainer.
4567 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4568 the uid or gid is negative. A case of a negative gid caused this to be
4569 noticed. The fix allows for either to be negative.
4571 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4572 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4575 PH/36 The change PH/12 above was broken. Fixed it.
4577 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4578 the grounds that routing the same address twice would always produce the
4579 same answer. This might have been true once, but it is certainly no
4580 longer true now. Routing a child address may depend on the previous
4581 routing that produced that child. Some complicated redirection strategies
4582 went wrong when messages had multiple recipients, and made Exim's
4583 behaviour dependent on the order in which the addresses were given.
4585 I have moved the duplicate checking until after the routing is complete.
4586 Exim scans the addresses that are assigned to local and remote
4587 transports, and removes any duplicates. This means that more work will be
4588 done, as duplicates will always all be routed, but duplicates are
4589 presumably rare, so I don't expect this is of any significance.
4591 For deliveries to pipes, files, and autoreplies, the duplicate checking
4592 still happens during the routing process, since they are not going to be
4595 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4596 It corrects a timeout issue with spamd. This is Ian's comment: "The
4597 background is that sometimes spamd either never reads data from a
4598 connection it has accepted, or it never writes response data. The exiscan
4599 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4600 blindly assumes that writes won't block so it may never time out."
4602 PH/39 Allow G after quota size as well as K and M.
4604 PH/40 The value set for $authenticated_id in an authenticator may not contain
4605 binary zeroes or newlines because the value is written to log lines and
4606 to spool files. There was no check on this. Now the value is run through
4607 the string_printing() function so that such characters are converted to
4608 printable escape sequences.
4610 PH/41 $message_linecount is a new variable that contains the total number of
4611 lines in the message. Compare $body_linecount, which is the count for the
4614 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4615 bounce and delay warning messages, except in certain special cases, which
4618 (a) An SMTP error message from a remote host;
4619 (b) A message specified in a :fail: redirection;
4620 (c) A message specified in a "fail" command in a system filter;
4621 (d) A message specified in a FAIL return from the queryprogram router;
4622 (e) A message specified by the cannot_route_message router option.
4624 In these cases only, Exim does include the error details in bounce and
4625 warning messages. There are also a few cases where bland messages such
4626 as "unrouteable address" or "local delivery error" are given.
4628 PH/43 $value is now also set for the "else" part of a ${run expansion.
4630 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4631 being worked on, but at least Exim now implements the latest version to
4634 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4635 process to complete was treated as a delivery failure, a timeout while
4636 writing the message to the pipe was logged, but erroneously treated as a
4637 successful delivery. Such timeouts include transport filter timeouts. For
4638 consistency with the overall process timeout, these timeouts are now
4639 treated as errors, giving rise to delivery failures by default. However,
4640 there is now a new Boolean option for the pipe transport called
4641 timeout_defer, which, if set TRUE, converts the failures into defers for
4642 both kinds of timeout. A transport filter timeout is now identified in
4645 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4646 systems where "make" and "gmake" are different, calling "gmake" at top
4647 level broke things. I've arranged for the value of $(MAKE) to be passed
4648 from the Makefile to this script so that it can call the same version of
4652 A note about Exim versions 4.44 and 4.50
4653 ----------------------------------------
4655 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4656 changes of various kinds. As a consequence, a big documentation update was
4657 needed. This delayed the release for rather longer than seemed good, especially
4658 in the light of a couple of (minor) security issues. Therefore, the changes
4659 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4660 release. So 4.44 and 4.50 are in effect two different branches that both start
4663 I have left the 4.50 change log unchanged; it contains all the changes since
4664 4.43. The change log for 4.44 is below; many of its items are identical to
4665 those for 4.50. This seems to be the most sensible way to preserve the
4666 historical information.
4672 1. Minor wording change to the doc/README.SIEVE file.
4674 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4675 computation of the current number of files was incorrect.
4677 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4678 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4681 4. Give more explanation in the error message when the command for a transport
4682 filter fails to execute.
4684 5. There are several places where Exim runs a non-Exim command in a
4685 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4686 was being done only for the command run by the queryprogram router. It is
4687 now done for all such subprocesses. The other cases are: ${run, transport
4688 filters, and the commands run by the lmtp and pipe transports.
4690 6. Added CONFIGURE_GROUP build-time option.
4692 7. Some older OS have a limit of 256 on the maximum number of file
4693 descriptors. Exim was using setrlimit() to set 1000 as a large value
4694 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4695 systems. I've change it so that if it can't get 1000, it tries for 256.
4697 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4698 was an oversight, and furthermore, ever since the addition of extra
4699 controls (e.g. 4.43/32), the checks on when to allow different forms of
4700 "control" were broken. There should now be diagnostics for all cases when a
4701 control that does not make sense is encountered.
4703 9. Added the /retain_sender option to "control=submission".
4705 10. $recipients is now available in the predata ACL (oversight).
4707 11. Tidy the search cache before the fork to do a delivery from a message
4708 received from the command line. Otherwise the child will trigger a lookup
4709 failure and thereby defer the delivery if it tries to use (for example) a
4710 cached ldap connection that the parent has called unbind on.
4712 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4713 of $address_data from the recipient verification was clobbered by the
4714 sender verification.
4716 13. The value of address_data from a sender verification is now available in
4717 $sender_address_data in subsequent conditions in the ACL statement.
4719 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4721 15. Added a new option "connect=<time>" to callout options, to set a different
4724 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4725 was its contents. (It was OK if the option was not defined at all.)
4727 17. A "Completed" log line is now written for messages that are removed from
4728 the spool by the -Mrm option.
4730 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4731 information about exactly what failed.
4733 19. Added -dd to debug only the daemon process.
4735 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4736 handles timeouts, both on the server side and network timeouts. Renamed the
4737 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4739 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4740 It is now set to "smtps".
4742 22. $host_address is now set to the target address during the checking of
4743 ignore_target_hosts.
4745 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4746 being passed; this would have caused $sender_host_name to have been used if
4747 matching the list had actually called for a host name (not very likely,
4748 since this list is usually IP addresses). A host name is now passed as
4751 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4752 code that specifies a non-restarting handler (typically sigaction() in
4753 modern systems) in an attempt to fix a rare and obscure crash bug.
4755 25. Narrowed the window for a race in the daemon that could cause it to ignore
4756 SIGCHLD signals. This is not a major problem, because they are used only to
4757 wake it up if nothing else does.
4759 26. A malformed maildirsize file could cause Exim to calculate negative values
4760 for the mailbox size or file count. Odd effects could occur as a result.
4761 The maildirsize information is now recalculated if the size or filecount
4764 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4765 support for a long time. Removed HAVE_SYS_VFS_H.
4767 28. Installed the latest version of exipick from John Jetmore.
4769 29. In an address list, if the pattern was not a regular expression, an empty
4770 subject address (from a bounce message) matched only if the pattern was an
4771 empty string. Non-empty patterns were not even tested. This was the wrong
4772 because it is perfectly reasonable to use an empty address as part of a
4773 database query. An empty address is now tested by patterns that are
4774 lookups. However, all the other forms of pattern expect the subject to
4775 contain a local part and a domain, and therefore, for them, an empty
4776 address still always fails if the pattern is not itself empty.
4778 30. Exim went into a mad DNS loop when attempting to do a callout where the
4779 host was specified on an smtp transport, and looking it up yielded more
4780 than one IP address.
4782 31. Re-factored the code for checking spool and log partition space into a
4783 function that finds that data and another that does the check. The former
4784 is then used to implement four new variables: $spool_space, $log_space,
4785 $spool_inodes, and $log_inodes.
4787 32. The RFC2047 encoding function was originally intended for short strings
4788 such as real names; it was not keeping to the 75-character limit for
4789 encoded words that the RFC imposes. It now respects the limit, and
4790 generates multiple encoded words if necessary. To be on the safe side, I
4791 have increased the buffer size for the ${rfc2047: expansion operator from
4794 33. It is now permitted to omit both strings after an "if" condition; if the
4795 condition is true, the result is "true". As before, when the second string
4796 is omitted, a false condition yields an empty string. This makes it less
4797 cumbersome to write custom ACL and router conditions.
4799 34. Failure to deliver a bounce message always caused it to be frozen, even if
4800 there was an errors_to setting on the router. The errors_to setting is now
4803 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4804 canonical form (fully expanded) before being placed in
4805 $sender_host_address.
4807 36. The table in the code that translates DNS record types into text (T_A to
4808 "A" for instance) was missing entries for NS and CNAME. It is just possible
4809 that this could have caused confusion if both these types were looked up
4810 for the same domain, because the text type is used as part of Exim's
4811 per-process caching. But the chance of anyone hitting this buglet seems
4814 37. The dnsdb lookup has been extended in a number of ways.
4816 (1) There is a new type, "zns", which walks up the domain tree until it
4817 finds some nameserver records. It should be used with care.
4819 (2) There is a new type, "mxh", which is like "mx" except that it returns
4820 just the host names, not the priorities.
4822 (3) It is now possible to give a list of domains (or IP addresses) to be
4823 looked up. The behaviour when one of the lookups defers can be
4824 controlled by a keyword.
4826 (4) It is now possible to specify the separator character for use when
4827 multiple records are returned.
4829 38. The dnslists ACL condition has been extended: it is now possible to supply
4830 a list of IP addresses and/or domains to be looked up in a particular DNS
4833 39. Added log_selector=+queue_time_overall.
4835 40. When running the queue in the test harness, wait just a tad after forking a
4836 delivery process, to get repeatability of debugging output.
4838 41. Include certificate and key file names in error message when GnuTLS fails
4839 to set them up, because the GnuTLS error message doesn't include the name
4840 of the failing file when there is a problem reading it.
4842 42. Allow both -bf and -bF in the same test run.
4844 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4846 44. The "Exiscan patch" is now merged into the mainline Exim source.
4848 45. Sometimes the final signoff response after QUIT could fail to get
4849 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4850 < 0 before doing a fflush(). This bug looks as though it goes back to the
4851 introduction of TLS in release 3.20, but "sometimes" must have been rare
4852 because the tests only now provoked it.
4854 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4855 (this can affect the format of dates).
4857 47. exim_tidydb, when checking for the continued existence of a message for
4858 which it has found a message-specific retry record, was not finding
4859 messages that were in split spool directories. Consequently, it was
4860 deleting retry records that should have stayed in existence.
4862 48. Steve fixed some bugs in eximstats.
4864 49. The SPA authentication driver was not abandoning authentication and moving
4865 on to the next authenticator when an expansion was forced to fail,
4866 contradicting the general specification for all authenticators. Instead it
4867 was generating a temporary error. It now behaves as specified.
4869 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4870 (the order specifies the preference for clients). The order is now AES256,
4871 AES128, 3DES, ARCFOUR128.
4873 51. Small patch to Sieve code - explicitly set From: when generating an
4876 52. Exim crashed if a remote delivery caused a very long error message to be
4877 recorded - for instance if somebody sent an entire SpamAssassin report back
4878 as a large number of 550 error lines. This bug was coincidentally fixed by
4879 increasing the size of one of Exim's internal buffers (big_buffer) that
4880 happened as part of the Exiscan merge. However, to be on the safe side, I
4881 have made the code more robust (and fixed the comments that describe what
4884 53. Now that there can be additional text after "Completed" in log lines (if
4885 the queue_time_overall log selector is set), a one-byte patch to exigrep
4886 was needed to allow it to recognize "Completed" as not the last thing in
4889 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4890 patch that reportedly fixes this has been added. I am not expert enough to
4891 create a test for it. This is what the patch creator wrote:
4893 "I found a little strange behaviour of ldap code when working with
4894 Windows 2003 AD Domain, where users was placed in more than one
4895 Organization Units. When I tried to give exim partial DN, the exit code
4896 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4897 But simultaneously result of request was absolutely normal ldap result,
4898 so I produce this patch..."
4900 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4901 REFERENCE, so I have modified the code to exclude the patch when that macro
4904 55. Some experimental protocols are using DNS PTR records for new purposes. The
4905 keys for these records are domain names, not reversed IP addresses. The
4906 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4907 leaves it alone. Component reversal etc. now happens only for IP addresses.
4910 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4912 57. Double the size of the debug message buffer (to 2048) so that more of very
4913 long debug lines gets shown.
4915 58. The exicyclog utility now does better if the number of log files to keep
4916 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4918 59. Two changes related to the smtp_active_hostname option:
4920 (1) $smtp_active_hostname is now available as a variable.
4921 (2) The default for smtp_banner uses $smtp_active_hostname instead
4922 of $primary_hostname.
4924 60. The host_aton() function is supposed to be passed a string that is known
4925 to be a valid IP address. However, in the case of IPv6 addresses, it was
4926 not checking this. This is a hostage to fortune. Exim now panics and dies
4927 if the condition is not met. A case was found where this could be provoked
4928 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4929 components; fortuitously, this particular loophole had already been fixed
4930 by change 4.50/55 above.
4932 If there are any other similar loopholes, the new check in host_aton()
4933 itself should stop them being exploited. The report I received stated that
4934 data on the command line could provoke the exploit when Exim was running as
4935 exim, but did not say which command line option was involved. All I could
4936 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4937 running as the user.
4940 61. There was a buffer overflow vulnerability in the SPA authentication code
4941 (which came originally from the Samba project). I have added a test to the
4942 spa_base64_to_bits() function which I hope fixes it.
4945 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4946 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4948 63. The daemon start-up calls getloadavg() while still root for those OS that
4949 need the first call to be done as root, but it missed one case: when
4950 deliver_queue_load_max is set with deliver_drop_privilege. This is
4951 necessary for the benefit of the queue runner, because there is no re-exec
4952 when deliver_drop_privilege is set.
4954 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4955 This has been fixed.
4957 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4958 was in use, was not putting the data itself into the right store pool;
4959 consequently, it could be overwritten for a subsequent message in the same
4960 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4963 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4965 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4966 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4967 the function were treating the return as a boolean value, which happened to
4968 work because 0=false and not-0=true, but is not correct code.
4970 68. The host_aton() function was not handling scoped IPv6 addresses (those
4971 with, for example, "%eth0" on the end) correctly.
4973 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4974 negated items (that is, ~something) in unsigned ints. Some compilers
4975 apparently mutter when there is no cast.
4977 70. If an address verification called from an ACL failed, and did not produce a
4978 user-specific message (i.e. there was only a "system" message), nothing was
4979 put in $acl_verify_message. In this situation, it now puts the system
4982 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4983 session; change 4.31/43 added the unwanted input to the log line - except
4984 that it did not do this in the start of session case. It now does.
4986 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4987 This is wrong and can cause the other end to generate a synchronization
4988 error if it is another Exim or anything else that does the synchronization
4989 check. A QUIT command is no longer sent after a timeout.
4991 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4992 during host lookups.
4994 74. The defer_ok option of callout verification was not working if it was used
4995 when verifying addresses in header lines, that is, for this case:
4997 verify = header_sender/callout=defer_ok
4999 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5000 those file descriptors could be used for SMTP connections. If anything
5001 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5002 could be sent to the SMTP client, causing chaos. The daemon now opens
5003 stdin, stdout, and stderr to /dev/null when it puts itself into the
5006 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5007 log by default. The user can override this with suitable Perl magic.
5009 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5010 the log message when discard triggers, was not working for the DATA ACL or
5011 for the non-SMTP ACL.
5013 78. Error message wording change in sieve.c.
5015 79. If smtp_accept_max_per_host was set, the number of connections could be
5016 restricted to fewer than expected, because the daemon was trying to set up
5017 a new connection before checking whether the processes handling previous
5018 connections had finished. The check for completed processes is now done
5019 earlier. On busy systems, this bug wouldn't be noticed because something
5020 else would have woken the daemon, and it would have reaped the completed
5023 80. If a message was submitted locally by a user whose login name contained one
5024 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5025 It caused a spool format error. I have fixed the spool reading code. A
5026 related problem was that the "from" clause in the Received: line became
5027 illegal because of the space(s). It is now covered by ${quote_local_part.
5029 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5032 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5034 83. Patch from Sieve maintainer to fix unterminated string problem in
5035 "vacation" handling.
5037 84. Some minor changes to the Linux configuration files to help with other
5038 OS variants using glibc.
5040 85. One more patch for Sieve to update vacation handling to latest spec.
5043 ----------------------------------------------------
5044 See the note above about the 4.44 and 4.50 releases.
5045 ----------------------------------------------------
5051 1. Change 4.43/35 introduced a bug that caused file counts to be
5052 incorrectly computed when quota_filecount was set in an appendfile
5055 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5056 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5059 3. Give more explanation in the error message when the command for a transport
5060 filter fails to execute.
5062 4. There are several places where Exim runs a non-Exim command in a
5063 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5064 was being done only for the command run by the queryprogram router. It is
5065 now done for all such subprocesses. The other cases are: ${run, transport
5066 filters, and the commands run by the lmtp and pipe transports.
5068 5. Some older OS have a limit of 256 on the maximum number of file
5069 descriptors. Exim was using setrlimit() to set 1000 as a large value
5070 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5071 systems. I've change it so that if it can't get 1000, it tries for 256.
5073 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5074 was an oversight, and furthermore, ever since the addition of extra
5075 controls (e.g. 4.43/32), the checks on when to allow different forms of
5076 "control" were broken. There should now be diagnostics for all cases when a
5077 control that does not make sense is encountered.
5079 7. $recipients is now available in the predata ACL (oversight).
5081 8. Tidy the search cache before the fork to do a delivery from a message
5082 received from the command line. Otherwise the child will trigger a lookup
5083 failure and thereby defer the delivery if it tries to use (for example) a
5084 cached ldap connection that the parent has called unbind on.
5086 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5087 of $address_data from the recipient verification was clobbered by the
5088 sender verification.
5090 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5091 was its contents. (It was OK if the option was not defined at all.)
5093 11. A "Completed" log line is now written for messages that are removed from
5094 the spool by the -Mrm option.
5096 12. $host_address is now set to the target address during the checking of
5097 ignore_target_hosts.
5099 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5100 being passed; this would have caused $sender_host_name to have been used if
5101 matching the list had actually called for a host name (not very likely,
5102 since this list is usually IP addresses). A host name is now passed as
5105 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5106 code that specifies a non-restarting handler (typically sigaction() in
5107 modern systems) in an attempt to fix a rare and obscure crash bug.
5109 15. Narrowed the window for a race in the daemon that could cause it to ignore
5110 SIGCHLD signals. This is not a major problem, because they are used only to
5111 wake it up if nothing else does.
5113 16. A malformed maildirsize file could cause Exim to calculate negative values
5114 for the mailbox size or file count. Odd effects could occur as a result.
5115 The maildirsize information is now recalculated if the size or filecount
5118 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5119 support for a long time. Removed HAVE_SYS_VFS_H.
5121 18. Updated exipick to current release from John Jetmore.
5123 19. Allow an empty sender to be matched against a lookup in an address list.
5124 Previously the only cases considered were a regular expression, or an
5127 20. Exim went into a mad DNS lookup loop when doing a callout where the
5128 host was specified on the transport, if the DNS lookup yielded more than
5131 21. The RFC2047 encoding function was originally intended for short strings
5132 such as real names; it was not keeping to the 75-character limit for
5133 encoded words that the RFC imposes. It now respects the limit, and
5134 generates multiple encoded words if necessary. To be on the safe side, I
5135 have increased the buffer size for the ${rfc2047: expansion operator from
5138 22. Failure to deliver a bounce message always caused it to be frozen, even if
5139 there was an errors_to setting on the router. The errors_to setting is now
5142 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5143 canonical form (fully expanded) before being placed in
5144 $sender_host_address.
5146 24. Updated eximstats to version 1.33
5148 25. Include certificate and key file names in error message when GnuTLS fails
5149 to set them up, because the GnuTLS error message doesn't include the name
5150 of the failing file when there is a problem reading it.
5152 26. Expand error message when OpenSSL has problems setting up cert/key files.
5155 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5156 (this can affect the format of dates).
5158 28. exim_tidydb, when checking for the continued existence of a message for
5159 which it has found a message-specific retry record, was not finding
5160 messages that were in split spool directories. Consequently, it was
5161 deleting retry records that should have stayed in existence.
5163 29. eximstats updated to version 1.35
5164 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5165 1.35 - bugfix such that pie charts by volume are generated correctly
5167 30. The SPA authentication driver was not abandoning authentication and moving
5168 on to the next authenticator when an expansion was forced to fail,
5169 contradicting the general specification for all authenticators. Instead it
5170 was generating a temporary error. It now behaves as specified.
5172 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5173 (the order specifies the preference for clients). The order is now AES256,
5174 AES128, 3DES, ARCFOUR128.
5176 31. Small patch to Sieve code - explicitly set From: when generating an
5179 32. Exim crashed if a remote delivery caused a very long error message to be
5180 recorded - for instance if somebody sent an entire SpamAssassin report back
5181 as a large number of 550 error lines. This bug was coincidentally fixed by
5182 increasing the size of one of Exim's internal buffers (big_buffer) that
5183 happened as part of the Exiscan merge. However, to be on the safe side, I
5184 have made the code more robust (and fixed the comments that describe what
5187 33. Some experimental protocols are using DNS PTR records for new purposes. The
5188 keys for these records are domain names, not reversed IP addresses. The
5189 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5190 leaves it alone. Component reversal etc. now happens only for IP addresses.
5193 34. The host_aton() function is supposed to be passed a string that is known
5194 to be a valid IP address. However, in the case of IPv6 addresses, it was
5195 not checking this. This is a hostage to fortune. Exim now panics and dies
5196 if the condition is not met. A case was found where this could be provoked
5197 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5198 components; fortuitously, this particular loophole had already been fixed
5199 by change 4.50/55 or 4.44/33 above.
5201 If there are any other similar loopholes, the new check in host_aton()
5202 itself should stop them being exploited. The report I received stated that
5203 data on the command line could provoke the exploit when Exim was running as
5204 exim, but did not say which command line option was involved. All I could
5205 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5206 running as the user.
5209 35. There was a buffer overflow vulnerability in the SPA authentication code
5210 (which came originally from the Samba project). I have added a test to the
5211 spa_base64_to_bits() function which I hope fixes it.
5214 36. The daemon start-up calls getloadavg() while still root for those OS that
5215 need the first call to be done as root, but it missed one case: when
5216 deliver_queue_load_max is set with deliver_drop_privilege. This is
5217 necessary for the benefit of the queue runner, because there is no re-exec
5218 when deliver_drop_privilege is set.
5220 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5221 was in use, was not putting the data itself into the right store pool;
5222 consequently, it could be overwritten for a subsequent message in the same
5223 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5226 38. Sometimes the final signoff response after QUIT could fail to get
5227 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5228 < 0 before doing a fflush(). This bug looks as though it goes back to the
5229 introduction of TLS in release 3.20, but "sometimes" must have been rare
5230 because the tests only now provoked it.
5236 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5237 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5238 Now it returns an integer. A number of calls were still expecting a T/F
5239 return. Fortuitously, in all cases, the tests worked in OK situations,
5240 which is the norm. However, things would have gone wrong on any write
5241 failures on the smtp file descriptor. This function is used when sending
5242 messages over SMTP and also when doing verify callouts.
5244 2. When Exim is called to do synchronous delivery of a locally submitted
5245 message (the -odf or -odi options), it no longer closes stderr before doing
5248 3. Implemented the mua_wrapper option.
5250 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5252 5. Implemented the functions header_remove(), header_testname(),
5253 header_add_at_position(), and receive_remove_recipient(), and exported them
5256 6. If an ACL "warn" statement specified the addition of headers, Exim already
5257 inserted X-ACL-Warn: at the start if there was no header name. However, it
5258 was not making this test for the second and subsequent header lines if
5259 there were newlines in the string. This meant that an invalid header could
5260 be inserted if Exim was badly configured.
5262 7. Allow an ACL "warn" statement to add header lines at the start or after all
5263 the Received: headers, as well as at the end.
5265 8. Added the rcpt_4xx retry error code.
5267 9. Added postmaster_mailfrom=xxx to callout verification option.
5269 10. Added mailfrom=xxxx to the callout verification option, for verify=
5272 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5273 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5275 12. Inserted some casts to stop certain compilers warning when using pointer
5276 differences as field lengths or precisions in printf-type calls (mostly
5277 affecting debugging statements).
5279 13. Added optional readline() support for -be (dynamically loaded).
5281 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5282 same clock tick as a message's arrival, so that its received time was the
5283 same as the "first fail" time on the retry record, and that message
5284 remained on the queue past the ultimate address timeout, every queue runner
5285 would try a delivery (because it was past the ultimate address timeout) but
5286 after another failure, the ultimate address timeout, which should have then
5287 bounced the address, did not kick in. This was a "< instead of <=" error;
5288 in most cases the first failure would have been in the next clock tick
5289 after the received time, and all would be well.
5291 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5292 being recognized when the domain list was tested by the match_domain
5293 condition in an expansion string.
5295 16. Added the ${str2b64: operator.
5297 17. Exim was always calling setrlimit() to set a large limit for the number of
5298 processes, without checking whether the existing limit was already
5299 adequate. (It did check for the limit on file descriptors.) Furthermore,
5300 errors from getrlimit() and setrlimit() were being ignored. Now they are
5301 logged to the main and panic logs, but Exim does carry on, to try to do its
5302 job under whatever limits there are.
5304 18. Imported PCRE 5.0.
5306 19. Trivial typo in log message " temporarily refused connection" (the leading
5309 20. If the log selector return_path_on_delivery was set and an address was
5310 redirected to /dev/null, the delivery process crashed because it assumed
5311 that a return path would always be set for a "successful" delivery. In this
5312 case, the whole delivery is bypassed as an optimization, and therefore no
5315 21. Internal re-arrangement: the function for sending a challenge and reading
5316 a response while authentication was assuming a zero-terminated challenge
5317 string. It's now changed to take a pointer and a length, to allow for
5318 binary data in such strings.
5320 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5322 23. Exim was not respecting finduser_retries when seeking the login of the
5323 uid under which it was called; it was always trying 10 times. (The default
5324 setting of finduser_retries is zero.) Also, it was sleeping after the final
5325 failure, which is pointless.
5327 24. Implemented tls_on_connect_ports.
5329 25. Implemented acl_smtp_predata.
5331 26. If the domain in control=submission is set empty, Exim assumes that the
5332 authenticated id is a complete email address when it generates From: or
5333 Sender: header lines.
5335 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5336 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5337 chown and chgrp in /bin and hostname in /usr/bin.
5339 28. Exim was keeping the "process log" file open after each use, just as it
5340 does for the main log. This opens the possibility of it remaining open for
5341 long periods when the USR1 signal hits a daemon. Occasional processlog
5342 errors were reported, that could have been caused by this. Anyway, it seems
5343 much more sensible not to leave this file open at all, so that is what now
5346 29. The long-running daemon process does not normally write to the log once it
5347 has entered its main loop, and it closes the log before doing so. This is
5348 so that log files can straightforwardly be renamed and moved. However,
5349 there are a couple of unusual error situations where the daemon does write
5350 log entries, and I had neglected to close the log afterwards.
5352 30. The text of an SMTP error response that was received during a remote
5353 delivery was being truncated at 512 bytes. This is too short for some of
5354 the long messages that one sometimes sees. I've increased the limit to
5357 31. It is now possible to make retry rules that apply only when a message has a
5358 specific sender, in particular, an empty sender.
5360 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5361 it possible to be selective about when SMTP synchronization is enforced.
5363 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5365 32. Implemented hosts_connection_nolog.
5367 33. Added an ACL for QUIT.
5369 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5372 35. Added mailbox_size and mailbox_filecount to appendfile.
5374 36. Added control = no_multiline_responses to ACLs.
5376 37. There was a bug in the logic of the code that waits for the clock to tick
5377 in the case where the clock went backwards by a substantial amount such
5378 that the microsecond fraction of "now" was more than the microsecond
5379 fraction of "then" (but the whole seconds number was less).
5381 38. Added support for the libradius Radius client library this is found on
5382 FreeBSD (previously only the radiusclient library was supported).
5388 1. When certain lookups returned multiple values in the form name=value, the
5389 quoting of the values was not always being done properly. Specifically:
5390 (a) If the value started with a double quote, but contained no whitespace,
5392 (b) If the value contained whitespace other than a space character (i.e.
5393 tabs or newlines or carriage returns) it was not quoted.
5394 This fix has been applied to the mysql and pgsql lookups by writing a
5395 separate quoting function and calling it from the lookup code. The fix
5396 should probably also be applied to nisplus, ibase and oracle lookups, but
5397 since I cannot test any of those, I have not disturbed their existing code.
5399 2. A hit in the callout cache for a specific address caused a log line with no
5400 reason for rejecting RCPT. Now it says "Previous (cached) callout
5401 verification failure".
5403 3. There was an off-by-one bug in the queryprogram router. An over-long
5404 return line was truncated at 256 instead of 255 characters, thereby
5405 overflowing its buffer with the terminating zero. As well as fixing this, I
5406 have increased the buffer size to 1024 (and made a note to document this).
5408 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5409 when Exim is waiting for an SMTP response from a remote server, Exim
5410 restarts its select() call on the socket, thereby resetting its timeout.
5411 This is not a problem when such interrupts are rare. Somebody set up a cron
5412 job to run exiwhat every 2 minutes, which is less than the normal select()
5413 timeout (5 or 10 minutes). This meant that the select() timeout never
5414 kicked in because it was always reset. I have fixed this by comparing the
5415 time when an interrupt arrives with the time at the start of the first call
5416 to select(). If more time than the timeout has elapsed, the interrupt is
5417 treated as a timeout.
5419 5. Some internal re-factoring in preparation for the addition of Sieve
5420 extensions (by MH). In particular, the "personal" test is moved to a
5421 separate function, and given an option for scanning Cc: and Bcc: (which is
5422 not set for Exim filters).
5424 6. When Exim created an email address using the login of the caller as the
5425 local part (e.g. when creating a From: or Sender: header line), it was not
5426 quoting the local part when it contained special characters such as @.
5428 7. Installed new OpenBSD configuration files.
5430 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5431 try to make them clearer.
5433 9. Callout options, other than the timeout value, were being ignored when
5434 verifying sender addresses in header lines. For example, when using
5436 verify = header_sender/callout=no_cache
5438 the cache was (incorrectly) being used.
5440 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5441 only the Cygwin environment.
5443 11. When return_path_on_delivery was set as a log selector, if different remote
5444 addresses in the same message used different return paths and parallel
5445 remote delivery occurred, the wrong values would sometimes be logged.
5446 (Whenever a remote delivery process finished, the return path value from
5447 the most recently started remote delivery process was logged.)
5449 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5450 lines when AUTH and/or TLS are in use. This is the "received protocol"
5451 field. Exim used to use "asmtp" for authenticated SMTP, without any
5452 indication (in the protocol name) for TLS use. Now it follows the RFC and
5453 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5454 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5455 names appear in log lines as well as in Received: header lines.
5457 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5458 extensions, and comparison tests, and to fix some bugs.
5460 14. Changes to the "personal" filter test:
5462 (1) The test was buggy in that it was just doing the equivalent of
5463 "contains" tests on header lines. For example, if a user's address was
5464 anne@some.where, the "personal" test would incorrectly be true for
5466 To: susanne@some.where
5468 This test is now done by extracting each address from the header in turn,
5469 and checking the entire address. Other tests that are part of "personal"
5470 are now done using regular expressions (for example, to check local parts
5471 of addresses in From: header lines).
5473 (2) The list of non-personal local parts in From: addresses has been
5474 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5475 taken from the Sieve specification recommendations.
5477 (3) If the message contains any header line starting with "List-" it is
5478 treated as non-personal.
5480 (4) The test for "circular" in the Subject: header line has been removed
5481 because it now seems ill-conceived.
5483 15. Minor typos in src/EDITME comments corrected.
5485 16. Installed latest exipick from John Jetmore.
5487 17. If headers_add on a router specified a text string that was too long for
5488 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5489 of string_sprintf() is now avoided.
5491 18. $message_body_size was not set (it was always zero) when running the DATA
5492 ACL and the local_scan() function.
5494 19. For the "mail" command in an Exim filter, no default was being set for
5495 the once_repeat time, causing a random time value to be used if "once" was
5496 specified. (If the value happened to be <= 0, no repeat happened.) The
5497 default is now 0s, meaning "never repeat". The "vacation" command was OK
5498 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5499 (I found it when inspecting the code).
5501 20. There is now an overall timeout for performing a callout verification. It
5502 defaults to 4 times the callout timeout, which applies to individual SMTP
5503 commands during the callout. The overall timeout applies when there is more
5504 than one host that can be tried. The timeout is checked before trying the
5505 next host. This prevents very long delays if there are a large number of
5506 hosts and all are timing out (e.g. when the network connections are timing
5507 out). The value of the overall timeout can be changed by specifying an
5508 additional sub-option for "callout", called "maxwait". For example:
5510 verify = sender/callout=5s,maxwait=20s
5512 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5513 to the end before writing, but this should make it even safer).
5515 22. Exim was forgetting that it had advertised PIPELINING for the second and
5516 subsequent messages on an SMTP connection. It was also not resetting its
5517 memory on STARTTLS and an internal HELO.
5519 23. When Exim logs an SMTP synchronization error within a session, it now
5520 records whether PIPELINING has been advertised or not.
5522 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5523 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5524 rather than long int.
5526 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5528 26. Added the never_mail option to autoreply.
5534 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5535 crash if the getsockname() call failed; this can happen if a connection is
5536 closed very soon after it is established. The problem was simply in the
5537 order in which certain operations were done, causing Exim to try to write
5538 to the SMTP stream before it had set up the file descriptor. The bug has
5539 been fixed by making things happen in the correct order.
5545 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5546 before the connection was closed, thus losing the rejection response.
5548 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5549 some early Solaris releases, but causes trouble in current releases where
5550 socklen_t is defined.
5552 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5555 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5558 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5559 that releasing the top of it at the end releases what was used for sub-
5560 expansions (unless the block got too big). However, discard this block if
5561 the first thing is a variable or header, so that we can use its block when
5562 it is dynamic (useful for very large $message_headers, for example).
5564 6. Lookups now cache *every* query, not just the most recent. A new, separate
5565 store pool is used for this. It can be recovered when all lookup caches are
5566 flushed. Lookups now release memory at the end of their result strings.
5567 This has involved some general refactoring of the lookup sources.
5569 7. Some code has been added to the store_xxx() functions to reduce the amount
5570 of flapping under certain conditions.
5572 8. log_incoming_interface used to affect only the <= reception log lines. Now
5573 it causes the local interface and port to be added to several more SMTP log
5574 lines, for example "SMTP connection from", and rejection lines.
5576 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5578 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5580 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5581 could be overwritten at the end of the current message (or the start of a
5582 new message if it was set in a HELO ACL). The value is now preserved for
5583 the duration of the SMTP connection.
5585 12. If a transport had a headers_rewrite setting, and a matching header line
5586 contained an unqualified address, that address was qualified, even if it
5587 did not match any rewriting rules. The underlying bug was that the values
5588 of the flags that permit the existence of unqualified sender and recipient
5589 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5590 non-local messages, and by -bnq for local messages) were not being
5591 preserved with the message after it was received.
5593 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5594 "next input=" as part of the text comprising the host identity instead of
5595 the correct text. The code was using the same buffer for two different
5596 strings. However, depending on which order the printing function evaluated
5597 its arguments, the bug did not always show up. Under Linux, for example, my
5598 test suite worked just fine.
5600 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5601 doesn't work with some older versions of Perl. It has been changed to "my",
5602 which in any case is probably the better facility to use.
5604 15. A really picky compiler found some instances of statements for creating
5605 error messages that either had too many or two few arguments for the format
5608 16. The size of the buffer for calls to the DNS resolver has been increased
5609 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5610 for addresses that have a lot of PTR records. This alleviates a problem; it
5611 does not fully solve it.
5613 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5614 buffer now truncates the list and logs the incident, which is the same
5615 action as happens when Exim is looking up a host name and its aliases.
5616 Previously in this situation something unpredictable would happen;
5617 sometimes it was "internal error: store_reset failed".
5619 18. If a server dropped the connection unexpectedly when an Exim client was
5620 using GnuTLS and trying to read a response, the client delivery process
5621 crashed while trying to generate an error log message.
5623 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5624 string, for example:
5626 warn message = H1: something\nH2: something
5628 the text was added as a single header line from Exim's point of view
5629 though it ended up OK in the delivered message. However, searching for the
5630 second and subsequent header lines using $h_h2: did not work. This has been
5631 fixed. Similarly, if a system filter added multiple headers in this way,
5632 the routers could not see them.
5634 20. Expanded the error message when iplsearch is called with an invalid key to
5635 suggest using net-iplsearch in a host list.
5637 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5638 ACLs are no longer actually imposed (and a message to that effect is
5641 22. If a "gecos" field in a passwd entry contained escaped characters, in
5642 particular, if it contained a \" sequence, Exim got it wrong when building
5643 a From: or a Sender: header from that name. A second bug also caused
5644 incorrect handling when an unquoted " was present following a character
5645 that needed quoting.
5647 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5648 was not being matched caselessly.
5650 24. Arranged for all hyphens in the exim.8 source to be escaped with
5653 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5654 properly. Recipient callout cache records were still being keyed to include
5655 the sender, even when use_sender was set false. This led to far more
5656 callouts that were necessary. The sender is no longer included in the key
5657 when use_sender is false.
5659 26. Added "control = submission" modifier to ACLs.
5661 27. Added the ${base62d: operator to decode base 62 numbers.
5663 28. dnsdb lookups can now access SRV records.
5665 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5666 the configuration file.
5668 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5669 (-v) mode. This makes the output for a verbose queue run more intelligible.
5671 31. Added a use_postmaster feature to recipient callouts.
5673 32. Added the $body_zerocount variable, containing the number of binary zero
5674 bytes in the message body.
5676 33. The time of last modification of the "new" subdirectory is now used as the
5677 "mailbox time last read" when there is a quota error for a maildir
5680 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5682 35. Added +ignore_unknown as a special item in host lists.
5684 36. Code for decoding IPv6 addresses in host lists is now included, even if
5685 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5686 address was recognized as an IP address, but was then not correctly decoded
5687 into binary, causing unexpected and incorrect effects when compared with
5694 1. Very minor rewording of debugging text in manualroute to say "list of
5695 hosts" instead of "hostlist".
5697 2. If verify=header_syntax was set, and a header line with an unqualified
5698 address (no domain) and a large number of spaces between the end of the
5699 name and the colon was received, the reception process suffered a buffer
5700 overflow, and (when I tested it) crashed. This was caused by some obsolete
5701 code that should have been removed. The fix is to remove it!
5703 3. When running in the test harness, delay a bit after writing a bounce
5704 message to get a bit more predictability in the log output.
5706 4. Added a call to search_tidyup() just before forking a reception process. In
5707 theory, someone could use a lookup in the expansion of smtp_accept_max_
5708 per_host which, without the tidyup, could leave open a database connection.
5710 5. Added the variables $recipient_data and $sender_data which get set from a
5711 lookup success in an ACL "recipients" or "senders" condition, or a router
5712 "senders" option, similar to $domain_data and $local_part_data.
5714 6. Moved the writing of debug_print from before to after the "senders" test
5717 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5718 problems for message scanning, either using a data ACL, or using
5719 local_scan() because the Received: header was not generated till after they
5720 were called (in order to set the time as the time of reception completion).
5721 I have revised the way this works. The header is now generated after the
5722 body is received, but before the ACL or local_scan() are called. After they
5723 are run, the timestamp in the header is updated.
5729 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5730 before starting a queue runner without re-exec. This happened only when
5731 deliver_drop_privilege was set or when the Exim user was set to root. The
5732 effect of the bug was that timeouts during subsequent deliveries caused
5733 crashes instead of being properly handled. The handler is now left at its
5734 default (and expected) setting.
5736 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5737 message, again when deliver_drop_privilege is set or Exim is run as root.
5738 The bug described in (1) was not present in this case, but the tidying up
5739 of the other signals was missing. I have made the two cases consistent.
5741 3. The ignore_target_hosts setting on a manualroute router was being ignored
5742 for hosts that were looked up using the /MX notation.
5744 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5747 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5748 operated on the sender address. After changing the $sender_address to <>
5749 for the sender address verify, Exim was re-instated it as the original
5750 (before rewriting) address, but remembering that it had rewritten it, so it
5751 wasn't rewriting it again. This bug also had the effect of breaking the
5752 sender address verification caching when the sender address was rewritten.
5754 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5755 This has been changed so that if the ip literal address matches
5756 ignore_target_hosts, the router declines.
5758 7. Added expansion conditions match_domain, match_address, and match_local_
5759 part (NOT match_host).
5761 8. The placeholder for the Received: header didn't have a length field set.
5763 9. Added code to Exim itself and to exim_lock to test for a specific race
5764 condition that could lead to file corruption when using MBX delivery. The
5765 issue is with the lockfile that is created in /tmp. If this file is removed
5766 after a process has opened it but before that process has acquired a lock,
5767 there is the potential for a second process to recreate the file and also
5768 acquire a lock. This could lead to two Exim processes writing to the file
5769 at the same time. The added code performs the same test as UW imapd; it
5770 checks after acquiring the lock that its file descriptor still refers to
5771 the same named file.
5773 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5774 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5777 11. Added the smtp_active_hostname option. If used, this will typically be made
5778 to depend on the incoming interface address. Because $interface_address is
5779 not set up until the daemon has forked a reception process, error responses
5780 that can happen earlier (such as "too many connections") no longer contain
5783 12. If an expansion in a condition on a "warn" statement fails because a lookup
5784 defers, the "warn" statement is abandoned, and the next ACL statement is
5785 processed. Previously this caused the whole ACL to be aborted.
5787 13. Added the iplsearch lookup type.
5789 14. Added ident_timeout as a log selector.
5791 15. Added tls_certificate_verified as a log selector.
5793 16. Added a global option tls_require_ciphers (compare the smtp transport
5794 option of the same name). This controls incoming TLS connections.
5796 17. I finally figured out how to make tls_require_ciphers do a similar thing
5797 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5798 before starting the TLS session.
5800 18. Tabs are now shown as \t in -bP output.
5802 19. If the log selector return_path_on_delivery was set, Exim crashed when
5803 bouncing a message because it had too many Received: header lines.
5805 20. If two routers both had headers_remove settings, and the first one included
5806 a superfluous trailing colon, the final name in the first list and the
5807 first name in the second list were incorrectly joined into one item (with a
5808 colon in the middle).
5814 1. Added -C and -D options to the exinext utility, mainly to make it easier
5815 to include in the automated testing, but these could be helpful when
5816 multiple configurations are in use.
5818 2. The exinext utility was not formatting the output nicely when there was
5819 an alternate port involved in the retry record key, nor when there was a
5820 message id as well (for retries that were specific to a specific message
5821 and a specific host). It was also confused by IPv6 addresses, because of
5822 the additional colons they contain. I have fixed the IPv4 problem, and
5823 patched it up to do a reasonable job for IPv6.
5825 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5826 delivery, the log line now contains "pipelined" if PIPELINING was used.
5828 4. An SMTP transport process used to panic and die if the bind() call to set
5829 an explicit outgoing interface failed. This has been changed; it is now
5830 treated in the same way as a connect() failure.
5832 5. A reference to $sender_host_name in the part of a conditional expansion
5833 that was being skipped was still causing a DNS lookup. This no longer
5836 6. The def: expansion condition was not recognizing references to header lines
5837 that used bh_ and bheader_.
5839 7. Added the _cache feature to named lists.
5841 8. The code for checking quota_filecount in the appendfile transport was
5842 allowing one more file than it should have been.
5844 9. For compatibility with Sendmail, the command line option
5852 and sets the incoming protocol and host name (for trusted callers). The
5853 host name and its colon can be omitted when only the protocol is to be set.
5854 Note the Exim already has two private options, -pd and -ps, that refer to
5855 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5856 "s", but I don't think that's a major issue.
5858 10. A number of refactoring changes to the code, none of which should affect
5861 (a) The number of logging options was getting close to filling up the
5862 32-bit word that was used as a bit map. I have split them into two classes:
5863 those that are passed in the argument to log_write(), and those that are
5864 only ever tested independently outside of that function. These are now in
5865 separate 32-bit words, so there is plenty of room for expansion again.
5866 There is no change in the user interface or the logging behaviour.
5868 (b) When building, for example, log lines, the code previously used a
5869 macro that called string_cat() twice, in order to add two strings. This is
5870 not really sufficiently general. Furthermore, there was one instance where
5871 it was actually wrong because one of the argument was used twice, and in
5872 one call a function was used. (As it happened, calling the function twice
5873 did not affect the overall behaviour.) The macro has been replaced by a
5874 function that can join an arbitrary number of extra strings onto a growing
5877 (c) The code for expansion conditions now uses a table and a binary chop
5878 instead of a serial search (which was left over from when there were very
5879 few conditions). Also, it now recognizes conditions like "pam" even when
5880 the relevant support is not compiled in: a suitably worded error message is
5881 given if an attempt is made to use such a condition.
5883 11. Added ${time_interval:xxxxx}.
5885 12. A bug was causing one of the ddress fields not to be passed back correctly
5886 from remote delivery subprocesses. The field in question was not being
5887 subsequently used, so this caused to problems in practice.
5889 13. Added new log selectors queue_time and deliver_time.
5891 14. Might have fixed a bug in maildirsizefile handling that threw up
5892 "unexpected character" debug warnings, and recalculated the data
5893 unnecessarily. In any case, I expanded the warning message to give more
5896 15. Added the message "Restricted characters in address" to the statements in
5897 the default ACL that block characters like @ and % in local parts.
5899 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5900 Three changes have been made:
5902 (a) There was a serious bug; a negative response to MAIL caused the whole
5903 recipient domain to be cached as invalid, thereby blocking all messages
5904 to all local parts at the same domain, from all senders. This bug has
5905 been fixed. The domain is no longer cached after a negative response to
5906 MAIL if the sender used is not empty.
5908 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5911 (c) A new callout option, "use_sender" has been added for people who want
5912 the modified behaviour.
5918 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5921 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5922 indeed breaks things for older releases.
5924 3. Added additional logging to the case where there is a problem reading data
5925 from a filter that is running in a subprocess using a pipe, in order to
5926 try to track down a specific problem.
5928 4. Testing facility fudge: when running in the test harness and attempting
5929 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5930 getting "No route to host". Convert this to a timeout.
5932 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5935 6. Some OS don't have socklen_t but use size_t instead. This affects the
5936 fifth argument of getsockopt() amongst other things. This is now
5937 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5938 can be set for individual OS. I have set it for SunOS5, OSF1, and
5939 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5940 some earlier ones do not.
5942 7. Change 4.30/15 was not doing the test caselessly.
5944 8. The standard form for an IPv6 address literal was being rejected by address
5945 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5946 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5947 this, as well as the form without the "IPv6" on the front (but only when
5948 address literals are enabled, of course).
5950 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5952 10. Exim crashed if a message with an empty sender address specified by -f
5953 encountered a router with an errors_to setting. This could be provoked only
5954 by a command such as
5958 where an empty string was supplied; "<>" did not hit this bug.
5960 11. Installed PCRE release 4.5.
5962 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5963 remained set. It is now erased.
5965 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5966 times from message ids (which are base 36 rather than the normal 62).
5968 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5969 were being counted as actual protocol errors, and logged if the log
5970 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5971 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5972 and DATA following a set of rejected RCPTs do not count as protocol errors.
5973 In other words, Exim assumes they were pipelined, though this may not
5974 actually be the case. Of course, in all cases the client gets an
5975 appropriate error code.
5977 15. If a lookup fails in an ACL condition, a message about the failure may
5978 be available; it is used if testing the ACL cannot continue, because most
5979 such messages specify what the cause of the deferral is. However, some
5980 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5981 that caused an old message to be retained and used if a later statement
5982 caused a defer, replacing the real cause of the deferral.
5984 16. If an IP address had so many PTR records that the DNS lookup buffer
5985 was not large enough to hold them, Exim could crash while trying to process
5986 the truncated data. It now detects and logs this case.
5988 17. Further to 4.21/58, another change has been made: if (and only if) the
5989 first line of a message (the first header line) ends with CRLF, a bare LF
5990 in a subsequent header line has a space inserted after it, so as not to
5991 terminate the header.
5993 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5994 unnecessarily, used atoi() instead of strtol(), and didn't check the
5995 termination when getting file sizes from file names by regex.
5997 19. Completely re-implemented the support for maildirsize files, in the light
5998 of a number of problems with the previous contributed implementation
5999 (4.30/29). In particular:
6001 . If the quota is zero, the maildirsize file is maintained, but no quota is
6004 . If the maildir directory does not exist, it is created before any attempt
6005 to write a maildirsize file.
6007 . The quota value in the file is just a cache; if the quota is changed in
6008 the transport, the new value overrides.
6010 . A regular expression is available for excluding directories from the
6013 20. The autoreply transport checks the characters in options that define the
6014 message's headers; it allows continued headers, but it was checking with
6015 isspace() after an embedded newline instead of explicitly looking for a
6018 21. If all the "regular" hosts to which an address was routed had passed their
6019 expiry times, and had not reached their retry times, the address was
6020 bounced, even if fallback hosts were defined. Now Exim should go on to try
6023 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6024 equivalent code in the SMTP transport. Some hosts send humungous responses
6025 to HELO/EHLO, more than 1024 it seems.
6027 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6028 gives compiler warnings in some environments. I've now done it "properly",
6031 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6032 (because of problems with the built-in one) was declared to return uschar *
6033 instead of char *, causing compiler failure.
6035 25. Fixed a file descriptor leak when processing alias/forward files.
6037 26. Fixed a minor format string issue in dbfn.c.
6039 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6041 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6042 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6045 29. When a sender address is verified, it is cached, to save repeating the test
6046 when there is more than one recipient in a message. However, when the
6047 verification involves a callout, it is possible for different callout
6048 options to be set for different recipients. It is too complicated to keep
6049 track of this in the cache, so now Exim always runs a verification when a
6050 callout is required, relying on the callout cache for the optimization.
6051 The overhead is duplication of the address routing, but this should not be
6054 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6055 to be verified with callout=postmaster, and the main callout worked but the
6056 postmaster check failed, the verification correctly failed. However, if a
6057 subsequent RCPT command asked for sender verification *without* the
6058 postmaster check, incorrect caching caused this verification also to fail,
6061 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6062 it was not caching the DNS options (qualify_single, search_parents) that
6063 were used when the lookup failed. A subsequent lookup with different
6064 options therefore always gave the same answer, though there were cases
6065 where it should not have. (Example: a "domains = !$mx_any" option on a
6066 dnslookup router: the "domains" option is always processed without any
6067 widening, but the router might have qualify_single set.) Now Exim uses the
6068 cached value only when the same options are set.
6070 32. Added John Jetmore's "exipick" utility to the distribution.
6072 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6073 than a timeout (e.g. a certificate is required, and is not provided), an
6074 Exim server now closes the connection immediately. Previously it waited for
6075 the client to close - but if the client is SSL, it seems that they each
6076 wait for each other, leading to a delay before one of them times out.
6078 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6079 maintained 0.8.x compatibility because I don't think many are using it, and
6080 it is clearly obsolete.
6082 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6085 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6086 client certificate was expired. A simple patch fixes this, though I don't
6087 understand the full logic of why the verify callback is called multiple
6090 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6091 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6092 which causes problems with some clients (such as the Certicom SSL Plus
6093 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6094 disables the coutermeasure allowing Eudora to connect."
6096 38. Exim was not checking that a write() to a log file succeeded. This could
6097 lead to Bad Things if a log got too big, in particular if it hit a file
6098 size limit. Exim now panics and dies if it cannot write to a log file, just
6099 as it does if it cannot open a log file.
6101 39. Modified OS/Makefile-Linux so that it now contains
6103 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6105 The two -D definitions ensure that Exim is compiled with large file
6106 support, which makes it possible to handle log files that are bigger than
6109 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6110 instance) a domain was checked against a named list that involved a lookup,
6111 causing $domain_data to be set, then another domain was checked against the
6112 same list, then the first domain was re-checked, the value of $domain_data
6113 after the final check could be wrong. In particular, if the second check
6114 failed, it could be set empty. This bug probably also applied to
6117 41. The strip_trailing_dot option was not being applied to the address given
6118 with the -f command-line option.
6120 42. The code for reading a message's header from the spool was incrementing
6121 $received_count, but never initializing it. This meant that the value was
6122 incorrect (doubled) while delivering a message in the same process in which
6123 it was received. In the most common configuration of Exim, this never
6124 happens - a fresh exec is done - but it can happen when
6125 deliver_drop_privilege is set.
6127 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6128 it now includes up to 150 characters of the unexpected data in the log
6131 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6132 and building data strings. The size of both of these buffers was 10 000
6133 bytes - far larger than anybody would *ever* want, thought I. Needless to
6134 say, somebody hit the limit. I have increased the maximum line length to
6135 20 000 and the maximum data length of concatenated lines to 100 000. I have
6136 also fixed two bugs, because there was no checking on these buffers. Tsk,
6137 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6138 buffer is too small.
6140 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6141 lsearch lookups. Now it does.
6143 46. When parsing a route_list item in a manualroute router, a fixed-length
6144 buffer was used for the list of hosts. I made this 1024 bytes long,
6145 thinking that nobody would ever have a list of hosts that long. Wrong.
6146 Somebody had a whole pile of complicated expansion conditions, and the
6147 string was silently truncated, leading to an expansion error. It turns out
6148 that it is easier to change to an unlimited length (owing to other changes
6149 that have happened since this code was originally written) than to build
6150 structure for giving a limitation error. The length of the item that
6151 expands into the list of hosts is now unlimited.
6153 47. The lsearch lookup could not handle data where the length of text line was
6154 more than 4095 characters. Such lines were truncated, leading to shortened
6155 data being returned. It should now handle lines of any length.
6157 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6158 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6161 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6162 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6163 no longer happens in comment lines. A list of replacements is now placed
6164 at the head of all of the source files, except those whose only change is
6165 to replace PERL_COMMAND in the very first #! line.
6167 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6168 messages on the queue, with a bottom-up merge sort, using code contributed
6169 by Michael Haardt. This should make operations like -bp somewhat faster on
6170 large queues. It won't affect queue runners, except when queue_run_in_order
6173 51. Installed eximstats 1.31 in the distribution.
6175 52. Added support for SRV lookups to the dnslookup router.
6177 53. If an ACL referred to $message_body or $message_body_end, the value was not
6178 reset for any messages that followed in the same SMTP session.
6180 54. The store-handling optimization for building very long strings was not
6181 differentiating between the different store pools. I don't think this
6182 actually made any difference in practice, but I've tidied it.
6184 55. While running the routers to verify a sender address, $sender_address
6185 was still set to the sender address. This is wrong, because when routing to
6186 send a bounce to the sender, it would be empty. Therefore, I have changed
6187 it so that, while verifying a sender address, $sender_address is set to <>.
6188 (There is no change to what happens when verifying a recipient address.)
6190 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6191 target A or AAAA records (if not already returned) without resetting the
6192 qualify_single or search_parents options of the DNS resolver. These are
6193 inappropriate in this case because the targets of MX and SRV records must
6194 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6195 target that, when qualified, matched something in the local domain. These
6196 two options are now turned off when doing these lookups.
6198 57. It seems that at least some releases of Reiserfs (which does not have the
6199 concept of a fixed number of inodes) returns zero and not -1 for the
6200 number of available inodes. This interacted badly with check_spool_inodes,
6201 which assumed that -1 was the "no such thing" setting. What I have done is
6202 to check that the total number of inodes is greater than zero before doing
6203 the test of how many are available.
6205 58. When a "warn" ACL statement has a log_message modifier, the message is
6206 remembered, and not repeated. This is to avoid a lot of repetition when a
6207 message has many recipients that cause the same warning to be written.
6208 However, Exim was preserving the list of already written lines for an
6209 entire SMTP session, which doesn't seem right. The memory is now reset if a
6210 new message is started.
6212 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6213 debugging output unless log_rewrite was also set.
6215 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6216 of (int)(handle) when we know that handle contains (void *)(-1).
6218 61. The Exim daemon panic-logs an error return when it closes the incoming
6219 connection. However "connection reset by peer" seems to be common, and
6220 isn't really an error worthy of noting specially, so that particular error
6223 62. When Exim is trying to find all the local interfaces, it used to panic and
6224 die if the ioctl to get the interface flags failed. However, it seems that
6225 on at least one OS (Solaris 9) it is possible to have an interface that is
6226 included in the list of interfaces, but for which you get a failure error
6227 for this call. This happens when the interface is not "plumbed" into a
6228 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6229 failure of the "get flags" call assumes that the interface is down.
6231 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6232 makes life easier for people who are doing arithmetic on fields extracted
6233 from dates, where you often get leading zeros that should not be
6234 interpreted as octal.
6236 64. Added qualify_domain to the redirect router, to override the global
6239 65. If a pathologically long header line contained very many addresses (the
6240 report of this problem mentioned 10 000) and each of them was rewritten,
6241 Exim could use up a very large amount of memory. (It kept on making new
6242 copies of the header line as it rewrote, and never released the old ones.)
6243 At the expense of a bit more processing, the header rewriting function has
6244 been changed so that it no longer eats memory in this way.
6246 66. The generation of the Received: header has been moved from the time that a
6247 message starts to be received, to the time that it finishes. The timestamp
6248 in the Received: header should now be very close to that of the <= log
6249 line. There are two side-effects of this change:
6251 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6252 logged header lines no longer include the local Received: line, because
6253 it has not yet been created. The same applies to a copy of the message
6254 that is returned to a non-SMTP sender when a message is rejected.
6256 (b) When a filter file is tested using -bf, no additional Received: header
6257 is added to the test message. After some thought, I decided that this
6260 This change does not affect the value of $received_for. It is still set
6261 after address rewriting, but before local_scan() is called.
6263 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6265 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6266 gave an unhelpful panic error message, and a defer error. I have managed to
6267 change this behaviour so that it now rejects any supplied certificate,
6268 which seems right, as the list of acceptable certificates is empty.
6270 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6271 gave an unhelpful defer error. I have not managed to make this reject any
6272 supplied certificates, but the error message it gives is "no certificate
6273 supplied", which is not helpful.
6275 70. exigrep's output now also includes lines that are not associated with any
6276 message, but which match the given pattern. Implemented by a patch from
6277 Martin Sluka, which also tidied up the Perl a bit.
6279 71. Recipient callout verification, like sender verification, was using <> in
6280 the MAIL FROM command. This isn't really the right thing, since the actual
6281 sender may affect whether the remote host accepts the recipient or not. I
6282 have changed it to use the actual sender in the callout; this means that
6283 the cache record is now keyed on a recipient/sender pair, not just the
6284 recipient address. There doesn't seem to be a real danger of callout loops,
6285 since a callout by the remote host to check the sender would use <>.
6286 [SEE ABOVE: changed after hitting problems.]
6288 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6289 temporary errors. However, in the case of such a code being given after
6290 the end of a data transmission (i.e. after ".") Exim was failing to write
6291 a retry record for the message. (Yes, there was some broken host that was
6292 actually sending 8xx at this point.)
6294 73. An unknown lookup type in a host list could cause Exim to panic-die when
6295 the list was checked. (An example that provoked this was putting <; in the
6296 middle of a list instead of at the start.) If this happened during a DATA
6297 ACL check, a -D file could be left lying around. This kind of configuration
6298 error no longer causes Exim to die; instead it causes a defer error. The
6299 incident is still logged to the main and panic logs.
6301 74. Buglet left over from Exim 3 conversion. The message "too many messages
6302 in one connection" was written to the rejectlog but not the mainlog, except
6303 when address rewriting (yes!) was being logged.
6305 75. Added write_rejectlog option.
6307 76. When a system filter was run not as root (that is, when system_filter_user
6308 was set), the values of the $n variables were not being returned to the
6309 main process; thus, they were not subsequently available in the $sn
6312 77. Added +return_path_on_delivery log selector.
6314 78. A connection timeout was being treated differently from recipients deferred
6315 when testing hosts_max_try with a message that was older than the host's
6316 retry timeout. (The host should not be counted, thus allowing all hosts to
6317 be tried at least once before bouncing.) This may have been the cause of an
6318 occasionally reported bug whereby a message would remain on the queue
6319 longer than the retry timeout, but would be bounced if a delivery was
6320 forced. I say "may" because I never totally pinned down the problem;
6321 setting up timeout/retry tests is difficult. See also the next item.
6323 79. The ultimate address timeout was not being applied to errors that involved
6324 a combination of host plus message (for example, a timeout on a MAIL
6325 command). When an address resolved to a number of possible hosts, and they
6326 were not all tried for each delivery (e.g. because of hosts_max_try), a
6327 message could remain on the queue longer than the retry timeout.
6329 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6332 81. Fixed an obscure SMTP outgoing bug which required at least the following
6333 conditions: (a) there was another message waiting for the same server;
6334 (b) the server returned 5xx to all RCPT commands in the first message so
6335 that the message was not completed; (c) the server dropped the connection
6336 or gave a negative response to the RSET that Exim sends to abort the
6337 transaction. The observed case was a dropped connection after DATA that had
6338 been sent in pipelining mode. That is, the server had advertised PIPELINING
6339 but was not implementing it correctly. The effect of the bug was incorrect
6340 behaviour, such as trying another host, and this could lead to a crash.
6346 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6347 and daemon.c were passed as pointers to ints; they should have been
6348 pointers to socklen_t variables (which are typically unsigned ints).
6350 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6353 3. Fixed a really odd bug that affected only the testing scheme; patching a
6354 certain fixed string in the binary changed the value of another string that
6355 happened to be identical to the end of the original first string.
6357 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6358 name", it returns that address as the IP address. On some operating
6359 systems (e.g. Solaris), it also passes back the IP address string as the
6360 "host name". However, on others (e.g. Linux), it passes back an empty
6361 string. Exim wasn't checking for this, and was changing the host name to an
6362 empty string, assuming it had been canonicalized.
6364 5. Although rare, it is permitted to have more than one PTR record for a given
6365 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6366 all the names associated with an address, because they do in Solaris.
6367 However, it seems that they do not in Linux for data that comes from the
6368 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6369 I found this out when I moved to a new Linux workstation and tried to run
6370 the Exim test suite.
6372 To get round this problem I have changed the code so that it now does its
6373 own call to the DNS to look up PTR records when searching for a host name.
6374 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6375 addresses that are only in /etc/hosts are still found.
6377 This behaviour is, however, controlled by an option called host_lookup_
6378 order, which defaults to "bydns:byaddr". If people want to use the other
6379 order, or indeed, just use one or the other means of lookup, they can
6380 specify it in this variable.
6382 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6383 some operating systems, this comes back from gethostbyaddr() as an empty
6384 string, and this is what Exim used to test for. However, it seems that in
6385 other systems, "." is yielded. Exim now tests for this case too.
6387 7. The values of check_spool_space and check_log_space are now held internally
6388 as a number of kilobytes instead of an absolute number of bytes. If a
6389 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6390 kilobyte. This means that much larger values can be stored.
6392 8. Exim monitor: an attempt to get the action menu when not actually pointing
6393 at a message produces an empty menu entitled "No message selected". This
6394 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6395 no entries in it ("Shell widget menu has zero width and/or height"). So I
6396 have added a single, blank menu entry in this case.
6398 9. Added ${quote_local_part.
6400 10. MIME decoding is now applied to the contents of Subject: header lines when
6403 11. Now that a reference to $sender_host_address automatically causes a reverse
6404 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6405 host lookup before query-style lookups in lists that might use this
6406 variable. This has therefore been abolished, and the "net-" prefix is no
6407 longer necessary for query-style lookups.
6409 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6410 have been a typo for LFLAGS, so it has been changed.
6412 13. The install script calls Exim with "-C /dev/null" in order to find the
6413 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6414 to be output. However, since Exim outputs its version number before the
6415 error, it didn't break the script. It just looked ugly. I fixed this by
6416 always allowing "-C /dev/null" if the caller is root.
6418 14. Ignore overlarge ACL variable number when reading spool file - insurance
6419 against a later release with more variables having written the file.
6421 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6422 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6423 this, as well as the form without the "IPv6" on the front.
6425 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6426 OS/Makefile-Darwin file.
6428 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6429 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6431 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6432 variable. However, some OS use ints for inodes. I've added cast to long int
6433 to get rid of the compiler warning.
6435 19. I had forgotten to lock out "/../" in configuration file names when
6436 ALT_CONFIG_PREFIX was set.
6438 20. Routers used for verification do not need to specify transports. However,
6439 if such a router generated a host list, and callout was configured, Exim
6440 crashed, because it could not find a port number from the (non-existent)
6441 transport. It now assumes port 25 in this circumstance.
6443 21. Added the -t option to exigrep.
6445 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6446 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6447 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6448 them is set, LOOKUP_LSEARCH is forced.
6450 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6452 24. Added sender and host information to the "rejected by local_scan()" log
6453 line; previously there was no indication of these.
6455 25. Added .include_if_exists.
6457 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6458 Linux. It turns out that not all file systems support this. Apparently some
6459 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6460 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6461 not supported on the file descriptor, is now ignored when Exim is trying to
6462 sync a directory. This applies only to Linux.
6464 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6466 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6467 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6470 29. Added support for maildirsize files from supplied patch (modified a bit).
6472 30. The use of :fail: followed by an empty string could lead Exim to respond to
6473 sender verification failures with (e.g.):
6475 550 Verification failed for <xxx>
6476 550 Sender verify failed
6478 where the first response line was missing the '-' that indicates it is not
6479 the final line of the response.
6481 31. The loop for finding the name of the user that called Exim had a hardwired
6482 limit of 10; it now uses the value of finduser_retries, which is used for
6483 all other user lookups.
6485 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6488 33. Exim was neglecting to zero errno before one call of strtol() when
6489 expanding a string and expecting an integer value. On some systems this
6490 resulted in spurious "integer overflow" errors. Also, it was casting the
6491 result into an int without checking.
6493 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6494 did not work. The code looks as if it has *never* worked, though it appears
6495 to have been documented since at least release 1.62. I have made it work.
6497 35. The "timeout_DNS" error in retry rules, also documented since at least
6498 1.62, also never worked. As it isn't clear exactly what this means, and
6499 clearly it isn't a major issue, I have abolished the feature by treating it
6500 as "timeout", and writing a warning to the main and panic logs.
6502 36. The display of retry rules for -brt wasn't always showing the error code
6505 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6506 timeout_connect_A, timeout_connect_MX.
6508 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6509 to the empty sender.
6511 39. The daemon was not analysing the content of -oX till after it had closed
6512 stderr and disconnected from the controlling terminal. This meant that any
6513 syntax errors were only noted on the panic log, and the return code from
6514 the command was 0. By re-arranging the code a little, I've made the
6515 decoding happen first, so such errors now appear on stderr, and the return
6516 code is 1. However, the actual setting up of the sockets still happens in
6517 the disconnected process, so errors there are still only recorded on the
6520 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6521 connections (as happens on some IP stacks) was logged at start up time as
6522 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6523 it from "IPv6 and IPv4", which means that two separate sockets are being
6526 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6527 says whether AF_INET or AF_INET6 was passed as an argument.
6529 42. Exiwhat output was messed up when time zones were included in log
6532 43. Exiwhat now gives more information about the daemon's listening ports,
6533 and whether -tls-on-connect was used.
6535 44. The "port" option of the smtp transport is now expanded.
6537 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6538 silently ignored. Now an error message is written to the main and panic
6541 46. There's a new ACL modifier called "logwrite" which writes to a log file
6542 as soon as it is encountered.
6544 47. Added $local_user_uid and $local_user_gid at routing time.
6546 48. Exim crashed when trying to verify a sender address that was being
6549 49. Exim was recognizing only a space character after ".include". It now also
6550 recognizes a tab character.
6552 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6553 extracting the relevant information from the specification. The man page no
6554 longer contains scrambled data for the -d option, and I've added a section
6555 at the front about calling Exim under different names.
6557 51. Added "extra_headers" argument to the "mail" command in filter files.
6559 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6562 53. Installed eximstats 1.29.
6564 54. Added transport_filter_timeout as a generic transport option.
6566 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6567 Cc: header lines. This was required by RFC 822, but it not required by RFC
6570 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6571 incoming messages that did not have them. Now it does so only if the
6572 message originates locally, that is, if there is no associated remote host
6573 address. When Resent- header lines are present, this applies to the Resent-
6574 lines rather than the non-Resent- lines.
6576 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6577 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6579 58. Messages for configuration errors now include the name of the main
6580 configuration file - useful now that there may be more than one file in a
6581 list (.included file names were always shown).
6583 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6584 for those rare installations that do not start the daemon as root or run it
6585 setuid root. I've cut out the call to initgroups() if the daemon is not
6588 60. The Exim user and group can now be bound into the binary as text strings
6589 that are looked up at the start of Exim's processing.
6591 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6593 62. Added $mailstore_basename variable.
6595 63. Installed patch to sieve.c from Michael Haardt.
6597 64. When Exim failed to open the panic log after failing to open the main log,
6598 the original message it was trying to log was written to stderr and debug
6599 output, but if they were not available (the usual case in production), it
6600 was lost. Now it is written to syslog before the two lines that record the
6601 failures to open the logs.
6603 65. Users' Exim filters run in subprocesses under the user's uid. It is
6604 possible for a "deliver" command or an alias in a "personal" command to
6605 provoke an address rewrite. If logging of address rewriting is configured,
6606 this fails because the process is not running as root or exim. There may be
6607 a better way of dealing with this, but for the moment (because 4.30 needs
6608 to be released), I have disabled address rewrite logging when running a
6609 filter in a non-root, non-exim process.
6615 1. The buildconfig auxiliary program wasn't quoting the value set for
6616 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6617 not defined. This bug was masked in 4.22 by the effect that was fixed in
6620 2. Some messages that were rejected after a message id was allocated were
6621 shown as "incomplete" by exigrep. It no longer does this for messages that
6622 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6624 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6625 have allow_domain_literals set, the ID did not get logged in the <= line.
6626 Domain literals are now always recognized in Message-ID: header lines.
6628 4. The first argument for a ${extract expansion item is the key name or field
6629 number. Leading and trailing spaces in this item were not being ignored,
6630 causing some misleading effects.
6632 5. When deliver_drop_privilege was set, single queue runner processes started
6633 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6634 same command in the process it spins off) were not dropping privilege.
6636 6. When the daemon running as "exim" started a queue runner, it always
6637 re-executed Exim in the spun-off process. This is a waste of effort when
6638 deliver_drop_privilege is set. The new process now just calls the
6639 queue-runner function directly.
6645 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6648 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6649 ignored. Though the use of -oP was forcing the writing of a pid file, it
6650 was always written to the default place.
6652 3. If the message "no IP address found for host xxxx" is generated during
6653 incoming verification, it is now followed by identification of the incoming
6654 connection (so you can more easily find what provoked it).
6656 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6658 5. Added some features to "harden" Exim a bit more against certain attacks:
6660 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6661 be put in Local/Makefile. This is like the never_users runtime option,
6662 but it cannot be overridden. The default setting is "root".
6664 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6665 prefix string with which any file named in a -C command line option
6668 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6669 is retained for -C and -D only if the caller of Exim is root. Without
6670 it, the exim user may also use -C and -D and retain privilege.
6672 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6673 command line option is disabled.
6675 6. Macro names set by the -D option must start with an upper case letter, just
6676 like macro names defined in the configuration file.
6678 7. Added "dereference=" facility to LDAP.
6680 8. Two instances of the typo "uknown" in the source files are fixed.
6682 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6683 the Configure-Makefile script screwed up while processing it.
6685 10. Incorporated PCRE 4.4.
6687 11. The SMTP synchronization check was not operating right at the start of an
6688 SMTP session. For example, it could not catch a HELO sent before the client
6689 waited for the greeting. There is now a check for outstanding input at the
6690 point when the greeting is written. Because of the duplex, asynchronous
6691 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6692 way, but not yet received, when the check is performed.
6694 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6695 on TCP/IP sockets, because this apparently causes some broken clients to
6698 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6699 unchanged) from the Cygwin maintainer.
6701 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6702 was supported instead of testing for mbx. Effectively a typo.
6704 15. The spa authenticator server code was not checking that the input it
6705 received was valid base64.
6707 16. The debug output line for the "set" modifier in ACLs was not showing the
6708 name of the variable that was being set.
6710 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6712 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6713 lookup on its own. Something else was needed to trigger the lookup. For
6714 example, a match in host_lookup or the need for a host name in a host list.
6715 Now, if $sender_host_name is referenced and the host name has not yet been
6716 looked up, a lookup is performed. If the lookup fails, the variable remains
6717 empty, and $host_lookup_failed is set to "1".
6719 19. Added "eqi" as a case-independent comparison operator.
6721 20. The saslauthd authentication condition could segfault if neither service
6722 nor realm was specified.
6724 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6725 error message that was logged was misleading, and incoming SMTP
6726 connections were dropped. The message is now more accurate, and temporary
6727 errors are given to SMTP connections.
6729 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6730 (except RSET) with a 421 error, until QUIT is received. However, it was
6731 failing to send a response to QUIT.
6733 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6734 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6735 ACL is now run after the helo_try_verify_hosts code.
6737 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6738 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6739 case-independent, so other case variants are also recognized). Apparently
6740 some systems use these upper case variants.
6742 25. If more than two messages were waiting for the same host, and a transport
6743 filter was specified for the transport, Exim sent two messages over the
6744 same TCP/IP connection, and then failed with "socket operation on non-
6745 socket" when it tried to send the third.
6747 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6749 27. The extern definition of crypt16() in expand.c was not being excluded when
6750 the OS had its own crypt16() function.
6752 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6753 as a preferred synonym for return_size_limit, both as an option and as an
6756 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6758 30. Changed the default configuration ACL to relax the local part checking rule
6759 for addresses that are not in any local domains. For these addresses,
6760 slashes and pipe symbols are allowed within local parts, but the sequence
6761 /../ is explicitly forbidden.
6763 31. SPA server authentication was not clearing the challenge buffer before
6766 32. log_message in a "warn" ACL statement was writing to the reject log as
6767 well as to the main log, which contradicts the documentation and doesn't
6768 seem right (because no rejection is happening). So I have stopped it.
6770 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6771 However, I am unable to do any testing of this.
6773 34. Fixed an infelicity in the appendfile transport. When checking directories
6774 for a mailbox, to see if any needed to be created, it was accidentally
6775 using path names with one or more superfluous leading slashes; tracing
6776 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6778 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6779 contents are added to the log line that is written for every discarded
6780 recipient. (Previously a log_message setting was ignored.)
6782 36. The ${quote: operator now quotes the string if it is empty.
6784 37. The install script runs exim in order to find its version number. If for
6785 some reason other than non-existence or emptiness, which it checks, it
6786 could not run './exim', it was installing it with an empty version number,
6787 i.e. as "exim-". This error state is now caught, and the installation is
6790 38. An argument was missing from the function that creates an error message
6791 when Exim fails to connect to the socket for saslauthd authentication.
6792 This could cause Exim to crash, or give a corrupted message.
6794 39. Added isip, isip4, and isip6 to ${if conditions.
6796 40. The ACL variables $acl_xx are now saved with the message, and can be
6797 accessed later in routers, transports, and filters.
6799 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6800 strings in the file are not string-expanded.
6802 42. If a MAIL command specified a SIZE value that was too large to fit into an
6803 int variable, the check against message_size_limit failed. Such values are
6804 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6805 day this will have to be increased, but I don't think I want to be around
6806 when emails are that large.
6813 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6814 iconv() is not standard in FreeBSD.
6816 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6817 IPv6 enabled. The observed symptom was a segmentation fault on return from
6818 the function os_common_find_running_interfaces() in src/os.c.
6820 3. In the check_special_case() function in daemon.c I had used "errno" as an
6821 argument name, which causes warnings on some systems. This was basically a
6822 typo, since it was named "eno" in the comments!
6824 4. The code that waits for the clock to tick (at a resolution of some fraction
6825 of a second) so as to ensure message-id uniqueness was always waiting for
6826 at least one whole tick, when it could have waited for less. [This is
6827 almost certainly not relevant at current processor speeds, where it is
6828 unlikely to ever wait at all. But we try to future-proof.]
6830 5. The function that sleeps for a time interval that includes fractions of a
6831 second contained a race. It did not block SIGALRM between setting the
6832 timer, and suspending (a couple of lines later). If the interval was short
6833 and the sigsuspend() was delayed until after it had expired, the suspension
6834 never ended. On busy systems this could lead to processes getting stuck for
6837 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6838 process, before it forks any delivery processes. The open lookup caching
6839 mechanism meant that the open file or database connection was passed into
6840 the delivery process. The problem was that delivery processes always tidy
6841 up cached lookup data. This could cause a problem for the next delivery
6842 process started by the queue runner, because the external queue runner
6843 process does not know about the closure. So the next delivery process
6844 still has data in the lookup cache. In the case of a file lookup, there was
6845 no problem because closing a file descriptor in a subprocess doesn't affect
6846 the parent. However, if the lookup was caching a connection to a database,
6847 the connection was closed, and the second delivery process was likely to
6848 see errors such as "PGSQL: query failed: server closed the connection
6849 unexpectedly". The problem has been fixed by closing all cached lookups
6850 in a queue runner before running a delivery process.
6852 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6853 seem to have the "const" qualifier which it has on other OS. I've
6856 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6857 *of the same type* (client or server) with the same public name that an
6858 error should be diagnosed.
6860 9. When Exim looked up a host name for an IP address, but failed to find the
6861 original IP address when looking up the host name (a safety check), it
6862 output the message "<ip address> does not match any IP for NULL", which was
6863 confusing, to say the least. The bug was that the host name should have
6864 appeared instead of "NULL".
6866 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6867 user that is built into the binary, and the -C or -D options is used, root
6868 privilege is dropped before the configuration file is read. In addition,
6869 logging is switched to stderr instead of the normal log files. If the
6870 configuration then re-defines the Exim user, the unprivileged environment
6871 is probably not what is expected, so Exim logs a panic warning message (but
6874 However, if deliver_drop_privilege is set, the unprivileged state may well
6875 be exactly what is intended, so the warning has been cut out in that case,
6876 and Exim is allowed to try to write to its normal log files.
6882 1. smtp_return_error_details was not giving details for temporary sender
6883 or receiver verification errors.
6885 2. Diagnose a configuration error if two authenticators have the same public
6888 3. Exim used not to create the message log file for a message until the first
6889 delivery attempt. This could be confusing when incoming messages were held
6890 for policy or load reasons. The message log file is now created at the time
6891 the message is received, and an initial "Received" line is written to it.
6893 4. The automatically generated man page for command line options had a minor
6894 bug that caused no ill effects; however, a more serious problem was that
6895 the procedure for building the man page automatically didn't always
6896 operate. Consequently, release 4.20 contains an out-of-date version. This
6897 shouldn't happen again.
6899 5. When building Exim with embedded Perl support, the script that builds the
6900 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6901 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6903 6. The freeze_tell option was not being used for messages that were frozen on
6904 arrival, either by an ACL or by local_scan().
6906 7. Added the smtp_incomplete_transaction log selector.
6908 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6909 was accepting AUTH without a new EHLO.
6911 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6912 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6915 10. Logging of TCP/IP connections (when configured) now happens in the main
6916 daemon process instead of the child process, so that the TCP/IP connection
6917 count is more accurate (but it can never be perfect).
6919 11. The use of "drop" in a nested ACL was not being handled correctly in the
6920 outer ACL. Now, if condition failure induced by the nested "drop" causes
6921 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6922 or "require"), the connection is dropped.
6924 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6925 that yield "discard" can now be used with an "accept" or a "discard" verb,
6926 but an error is generated for any others (because I can't see a useful way
6927 to define what should happen).
6929 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6930 are now processed in the same way as lines in the Exim configuration file.
6931 In particular, continuation lines are supported.
6933 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6935 15. Added -ti meaning -t -i.
6937 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6938 domains, and warn by logging if others are found.
6940 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6941 returned from GIFCONF when Exim is trying to find the list of interfaces on
6942 a host. The code in os.c has been modified to copy each ifreq to an aligned
6943 structure in all cases.
6945 Also, in some cases, the returned ifreq's were being copied to a 'struct
6946 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6947 means the last couple of bytes of an IPv6 address could be chopped if the
6948 ifreq contained only a normal sockaddr (14 bytes storage).
6950 18. Named domain lists were not supported in the hosts_treat_as_local option.
6951 An entry such as +xxxx was not recognized, and was treated as a literal
6954 19. Ensure that header lines added by a DATA ACL are included in the reject log
6955 if the ACL subsequently rejects the message.
6957 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6958 MD5 (which is deprecated).
6960 21. When testing a filter file using -bf, Exim was writing a message when it
6961 took the sender from a "From " line in the message, but it was not doing so
6962 when it took $return_path from a Return-Path: header line. It now does.
6964 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6965 with a valid header line field name (a series of printing characters
6966 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6968 23. Changed "disc" in the source to "disk" to conform to the documentation and
6969 the book and for uniformity.
6971 24. Ignore Sendmail's -Ooption=value command line item.
6973 25. When execve() failed while trying to run a command in a pipe transport,
6974 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6975 could be confused with a return value of 69 from the command itself. This
6976 has been changed to 127, the value the shell returns if it is asked to run
6977 a non-existent command. The wording for the related log line suggests a
6978 non-existent command as the problem.
6980 26. If received_header_text expands to an empty string, do not add a Received:
6981 header line to the message. (Well, it adds a token one on the spool, but
6982 marks it "old" so that it doesn't get used or transmitted.)
6984 27. Installed eximstats 1.28 (addition of -nt option).
6986 28. There was no check for failure on the call to getsockname() in the daemon
6987 code. This can fail if there is a shortage of resources on the system, with
6988 ENOMEM, for example. A temporary error is now given on failure.
6990 29. Contrary to the C standard, it seems that in some environments, the
6991 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6992 program. Exim now does this explicitly; it affects the formatting of
6993 timestamps using strftime().
6995 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6996 complaints. I've now initialized all the variables, to avoid this.
6998 32. Header lines added by a system filter were not being "seen" during
6999 transport-time rewrites.
7001 33. The info_callback() function passed to OpenSSL is set up with type void
7002 (*)(SSL *, int, int), as described somewhere. However, when calling the
7003 function (actually a macro) that sets it up, the type void(*)() is
7004 expected. I've put in a cast to prevent warnings from picky compilers.
7006 34. If a DNS black list lookup found a CNAME record, but there were no A
7007 records associated with the domain it pointed at, Exim crashed.
7009 35. If a DNS black list lookup returned more than one A record, Exim ignored
7010 all but the first. It now scans all returned addresses if a particular IP
7011 value is being sought. In this situation, the contents of the
7012 $dnslist_value variable are a list of all the addresses, separated by a
7015 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7016 to accept a host name and all its aliases if the forward lookup for any of
7017 them yielded the IP address of the incoming connection. Now it accepts only
7018 those names whose forward lookup yields the correct IP address. Any other
7019 names are discarded. This closes a loophole whereby a rogue DNS
7020 administrator could create reverse DNS records to break through a
7021 wildcarded host restriction in an ACL.
7023 37. If a user filter or a system filter that ran in a subprocess used any of
7024 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7025 the wrong values were passed to the pipe command ($thisaddress had the
7026 value of $0, $0 had the value of $1, etc). This bug was introduced by
7027 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7029 38. Improved the line breaking for long SMTP error messages from ACLs.
7030 Previously, if there was no break point between 40 and 75 characters, Exim
7031 left the rest of the message alone. Two changes have been made: (a) I've
7032 reduced the minimum length to 35 characters; (b) if it can't find a break
7033 point between 35 and 75 characters, it looks ahead and uses the first one
7034 that it finds. This may give the occasional overlong line, but at least the
7035 remaining text gets split now.
7037 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7038 file descriptors might be low, and that setting 1000 would always raise it.
7039 It turns out that in some environments, the limit is already over 1000 and
7040 that lowering it causes trouble. So now Exim takes care not to decrease it.
7042 40. When delivering a message, the value of $return_path is set to $sender_
7043 address at the start of routing (routers may change the value). By an
7044 oversight, this default was not being set up when an address was tested by
7045 -bt or -bv, which affected the outcome if any router or filter referred to
7048 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7049 log only when "message" or "log_message" are set. However, if one of the
7050 conditions was an address verification, or a call to a nested ACL, the
7051 messages generated by the underlying test were being passed through. This
7052 no longer happens. The underlying message is available in $acl_verify_
7053 message for both "message" and "log_message" expansions, so it can be
7054 passed through if needed.
7056 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7057 new expansion $bh_ to give the encoded byte string without charset
7058 translation. Translation happens only if iconv() is available; HAVE_ICONV
7059 indicates this at build time. HEADERS_CHARSET gives the charset to
7060 translate to; headers_charset can change it in the configuration, and
7061 "headers charset" can change it in an individual filter file.
7063 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7064 that creates RFC 2047 encoded "words" labels them as that charset instead
7065 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7066 expansion operator; (ii) when Exim creates a From: line for a local
7067 message; (iii) when a header line is rewritten to include a "phrase" part.
7069 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7070 buggy, causing it to skip the first lines of messages whose message ID
7071 ended in 'D'. This would not have bitten before Exim release 4.14, because
7072 message IDs were unlikely to end in 'D' before then. The effect was to have
7073 incorrect size information for certain domains.
7075 45. #include "config.h" was missing at the start of the crypt16.c module. This
7076 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7079 46. If there was a timeout during a "random" callout check, Exim treated it as
7080 a failure of the random address, and carried on sending RSET and the real
7081 address. If the delay was just some slowness somewhere, the response to the
7082 original RCPT would be taken as a response to RSET and so on, causing
7083 mayhem of various kinds.
7085 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7086 when I implemented it. It didn't allow for the fact that some option values
7087 may legitimately be negative (e.g. size_addition), and it didn't even do
7088 the right test for positive values.
7090 48. Domain names in DNS records are case-independent. Exim always looks them up
7091 in lower case. Some resolvers return domain names in exactly the case they
7092 appear in the zone file, that is, they may contain uppercase letters. Not
7093 all resolvers do this - some return always lower case. Exim was treating a
7094 change of case by a resolver as a change of domain, similar to a widening
7095 of a domain abbreviation. This triggered its re-routing code and so it was
7096 trying to route what was effectively the same domain again. This normally
7097 caused routing to fail (because the router wouldn't handle the domain
7098 twice). Now Exim checks for this case specially, and just changes the
7099 casing of the domain that it ultimately uses when it transmits the message
7102 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7105 50. If a filter generated a file delivery with a non-absolute name (possible if
7106 no home directory exists for the router), the forbid_file option was not
7109 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7110 the existing equality matching.
7112 52. Exim was using ints instead of ino_t variables in some places where it was
7113 dealing with inode numbers.
7115 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7116 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7117 called TMPDIR, and if it finds it is different, it changes its value.
7119 54. The smtp_printf() function is now made available to local_scan() so
7120 additional output lines can be written before returning. There is also an
7121 smtp_fflush() function to enable the detection of a dropped connection.
7122 The variables smtp_input and smtp_batched_input are exported to
7125 55. Changed the default runtime configuration: the message "Unknown user"
7126 has been removed from the ACL, and instead placed on the localuser router,
7127 using the cannot_route_message feature. This means that any verification
7128 failures that generate their own messages won't get overridden. Similarly,
7129 the "Unrouteable address" message that was in the ACL for unverifiable
7130 relay addresses has also been removed.
7132 56. Added hosts_avoid_esmtp to the smtp transport.
7134 57. The exicyclog script was not checking for the esoteric option
7135 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7136 will work only if exicyclog is run under the appropriate euid.
7138 58. Following a discussion on the list, the rules by which Exim recognises line
7139 endings on incoming messages have been changed. The -dropcr and drop_cr
7140 options are now no-ops, retained only for backwards compatibility. The
7141 following line terminators are recognized: LF CRLF CR. However, special
7142 processing applies to CR:
7144 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7145 nor a local message in the state where . is a terminator.
7147 (ii) If a bare CR is encountered in a header line, an extra space is added
7148 after the line terminator so as not to end the header. The reasoning
7149 behind this is that bare CRs in header lines are most likely either
7150 to be mistakes, or people trying to play silly games.
7152 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7153 was being incorrectly given as 18 bytes larger than it should have been.
7154 This is a VOB (very old bug).
7156 60. This may never have affected anything current, but just in case it has:
7157 When the local host is found other than at the start of a list of hosts,
7158 the local host, those with the same MX, and any that follow, are discarded.
7159 When the list in question was part of a longer list of hosts, the following
7160 hosts (not currently being processed) were also being discarded. This no
7161 longer happens. I'm not sure if this situation could ever has previously
7164 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7167 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7168 header. This is something that is recommended in a new Internet Draft, and
7169 is something that is documented as being done by Sendmail. There are two
7170 possible values. For messages generated by the autoreply transport, Exim
7173 Auto-Submitted: auto-replied
7175 whereas for all other generated messages (e.g. bounces) it adds
7177 Auto-Submitted: auto-generated
7179 63. The "personal" condition in filters now includes a test for the
7180 Auto-Submitted: header. If it contains the string "auto-" the message it
7181 not considered personal.
7183 64. Added rcpt_include_affixes as a generic transport option.
7185 65. Added queue_only_override (default true).
7187 66. Added the syslog_duplication option.
7189 67. If what should have been the first header line of a message consisted of
7190 a space followed by a colon, Exim was mis-interpreting it as a header line.
7191 It isn't of course - it is syntactically invalid and should therefore be
7192 treated as the start of the message body. The misbehaviour could have
7193 caused a number of strange effects, including loss of data in subsequent
7194 header lines, and spool format errors.
7196 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7197 client host had authenticated. This control can now be exercised by an ACL
7198 for more flexibility.
7200 69. By default, callouts do not happen when testing with -bh. There is now a
7201 variant, -bhc, which does actually run the callout code, including
7202 consulting and updating the callout cache.
7204 70. Added support for saslauthd authentication, courtesy of Alexander
7207 71. If statvfs() failed on the spool or log directories while checking their
7208 size for availability, Exim confusingly gave the error "space shortage".
7209 Furthermore, in debugging mode it crashed with a floating point exception.
7210 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7211 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7212 serious problem, Exim now writes to the main and panic logs when this
7213 happens, with details of the failure. It then refuses to accept the
7214 incoming message, giving the message "spool directory problem" or "log
7215 directory problem" with a 421 code for SMTP messages.
7217 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7218 0, 1, and 2 exist, because some OS complain for execs without them (see
7219 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7220 descriptors. However, the code omitted to check that the open succeeded,
7221 causing mysterious errors if for some reason the permissions on /dev/null
7222 got screwed. Now Exim writes a message to the main and panic logs, and
7223 bombs out if it can't open /dev/null.
7225 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7226 interact so that it is all more flexible. It is supposed to remain
7227 backwards compatible. Also added extra_local_interfaces.
7229 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7230 to bomb out with an assertion failure - to the client this appears as a
7231 connection drop. This problem occurs in the part of the code that was taken
7232 from the Samba project. Fortunately, the assertion is in a very simple
7233 function, so I have fixed this by reproducing the function inline in the
7234 one place where it is called, and arranging for authentication to fail
7235 instead of killing the process with assert().
7237 75. The SPA client code was not working when the server requested OEM rather
7238 than Unicode encoding.
7240 76. Added code to make require_files with a specific uid setting more usable in
7241 the case where statting the file as root fails - usually a non-root-mounted
7242 NFS file system. When this happens and the failure is EACCES, Exim now
7243 forks a subprocess and does the per-uid checking as the relevant uid.
7245 77. Added process_log_path.
7247 78. If log_file_path was not explicitly set, a setting of check_log_space or
7248 check_log_inodes was ignored.
7250 79. If a space check for the spool or log partitions fails, the incident is now
7251 logged. Of course, in the latter case the data may get lost...
7253 80. Added the %p formatting code to string_format() so that it can be used to
7254 print addresses in debug_print(). Adjusted all the address printing in the
7255 debugging in store.c to use %p rather than %d.
7257 81. There was a concern that a line of code in smtp_in.c could overflow a
7258 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7259 initially expressed, the concern was not well-founded, because trailing
7260 spaces are removed early. However, if the trailing spaces were followed by
7261 a NULL, they did not get removed, so the overflow was possible. Two fixes
7264 (a) I re-wrote the offending code in a cleaner fashion.
7265 (b) If an incoming SMTP command contains a NULL character, it is rejected
7268 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7269 runs initgroups(), so that if the Exim user is in any additional groups,
7270 they will be used during message reception.
7276 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob