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.
94 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
95 When redis_servers is set to a list of > 1 element, and the Redis servers
96 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
97 case of MOVED into a DEFER case instead, thus moving the query onto the
98 next server in the list. For a cluster of N elements, all N servers must
99 be defined in redis_servers.
101 GF/02 Catch and remove uninitialized value warning in exiqsumm
102 Check for existence of @ARGV before looking at $ARGV[0]
104 JH/01 Replace the store_release() internal interface with store_newblock(),
105 which internalises the check required to safely use the old one, plus
106 the allocate and data copy operations duplicated in both (!) of the
107 extant use locations.
109 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
110 modifier. This matches the restriction on the commandline.
112 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
113 Previously only the last row was returned.
115 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
116 we assumed that tags in the header were well-formed, and parsed the
117 element content after inspecting only the first char of the tag.
118 Assumptions at that stage could crash the receive process on malformed
121 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
122 While running the DKIM ACL we operate on the Permanent memory pool so that
123 variables created with "set" persist to the DATA ACL. Also (at any time)
124 DNS lookups that fail create cache records using the Permanent pool. But
125 expansions release any allocations made on the current pool - so a dnsdb
126 lookup expansion done in the DKIM ACL releases the memory used for the
127 DNS negative-cache, and bad things result. Solution is to switch to the
128 Main pool for expansions.
129 While we're in that code, add checks on the DNS cache during store_reset,
130 active in the testsuite.
131 Problem spotted, and debugging aided, by Wolfgang Breyha.
133 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
134 When none of the hosts presented to a transport match an already-open
135 connection, close it and proceed with the list. Previously we would
136 queue the message. Spotted by Lena with Yahoo, probably involving
139 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
140 Previously a spurious "250 OK id=" response was appended to the proper
143 JH/08 The "support for" informational output now, which built with Content
144 Scanning support, has a line for the malware scanner interfaces compiled
145 in. Interface can be individually included or not at build time.
147 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
148 by the template makefile "src/EDITME". The "STREAM" support for an older
149 ClamAV interface method is removed.
151 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
152 rows affected is given instead).
154 JH/11 The runtime Berkeley DB library version is now additionally output by
155 "exim -d -bV". Previously only the compile-time version was shown.
157 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
158 SMTP connection. Previously, when one had more receipients than the
159 first, an abortive onward connection was made. Move to full support for
160 multiple onward connections in sequence, handling cutthrough connection
161 for all multi-message initiating connections.
163 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
164 routers. Previously, a multi-recipient message would fail to match the
165 onward-connection opened for the first recipient, and cause its closure.
167 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
168 a timeout on read on a GnuTLS initiating connection, resulting in the
169 initiating connection being dropped. This mattered most when the callout
170 was marked defer_ok. Fix to keep the two timeout-detection methods
173 JH/15 Relax results from ACL control request to enable cutthrough, in
174 unsupported situations, from error to silently (except under debug)
175 ignoring. This covers use with PRDR, frozen messages, queue-only and
178 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
180 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
181 metadata, resulting in a crash in free().
183 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
184 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
185 Broken also in d185889f4, with init system revamp.
187 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
188 Previously we abruptly closed the connection after reading a malware-
189 found indication; now we go on to read the "scan ok" response line,
192 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
193 ACL. Previously, a crash would result.
195 JH/19 Speed up macro lookups during configuration file read, by skipping non-
196 macro text after a replacement (previously it was only once per line) and
197 by skipping builtin macros when searching for an uppercase lead character.
199 JH/20 DANE support moved from Experimental to mainline. The Makefile control
200 for the build is renamed.
202 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
203 was allocated for every new TLS startup, meaning one per message. Fix
204 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
206 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
207 reported the original. Fix to report (as far as possible) the ACL
208 result replacing the original.
210 JH/23 Fix memory leak during multi-message connections using STARTTLS under
211 OpenSSL. Certificate information is loaded for every new TLS startup,
212 and the resources needed to be freed.
214 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
216 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
219 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
220 DATA response info to the (existing) per-recipient response info for
221 the "C=" log element. It can have useful tracking info from the
222 destination system. Patch from Simon Arlott.
224 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
225 length value. Previously this would segfault.
227 HS/02 Support Avast multiline protoocol, this allows passing flags to
228 newer versions of the scanner.
230 JH/28 Ensure that variables possibly set during message acceptance are marked
231 dead before release of memory in the daemon loop. This stops complaints
232 about them when the debug_store option is enabled. Discovered specifically
233 for sender_rate_period, but applies to a whole set of variables.
234 Do the same for the queue-runner and queue-list loops, for variables set
235 from spool message files. Do the same for the SMTP per-message loop, for
236 certain variables indirectly set in ACL operations.
238 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
239 as a multi-recipient message from a mailinglist manager). The coding had
240 an arbitrary cutoff number of characters while checking for more input;
241 enforced by writing a NUL into the buffer. This corrupted long / fast
242 input. The problem was exposed more widely when more pipelineing of SMTP
243 responses was introduced, and one Exim system was feeding another.
244 The symptom is log complaints of SMTP syntax error (NUL chars) on the
245 receiving system, and refused recipients seen by the sending system
246 (propating to people being dropped from mailing lists).
247 Discovered and pinpointed by David Carter.
249 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
250 replaced by the ${authresults } expansion.
252 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
254 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
255 allows proper process termination in container environments.
257 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
258 Previously the "final dot" had a newline after it; ensure it is CR,LF.
260 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
261 and "err_perm", deprecated since 4.83 when the RFC-defined words
262 "temperror" and "permerror" were introduced.
264 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
265 transport-filters or DKIM-signing. The restriction was lost in the
266 consolidation of verify-callout and delivery SMTP handling.
267 Extend the restriction to also cover ARC-signing.
269 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
270 in defer=pass mode supply a 450 to the initiator. Previously the message
273 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
274 tls_require_ciphers is used as before.
276 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
277 "pass_unscanned". Only tmpfails from the scanner are written to
278 the paniclog, as they may require admin intervention (permission
279 denied, license issues). Other scanner errors (like decompression
280 bombs) do not cause a paniclog entry.
282 JH/36 Fix reinitialisation of DKIM logging variable between messages.
283 Previously it was possible to log spurious information in receive log
286 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
287 triggered odd behaviour from Outlook Express clients.
289 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
292 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
293 since the IETF WG has not yet settled on that versus the original
294 "bare" representation.
296 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
297 Previously the millisecond value corrupted the output.
298 Fix also for syslog_pid=no and log_selector +pid, for which the pid
299 corrupted the output.
305 JH/01 Rework error string handling in TLS interface so that the caller in
306 more cases is responsible for logging. This permits library-sourced
307 string to be attached to addresses during delivery, and collapses
308 pairs of long lines into single ones.
310 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
311 during configuration. Wildcards are allowed and expanded.
313 JH/02 Rework error string handling in DKIM to pass more info back to callers.
314 This permits better logging.
316 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
317 do not close it down and have the child transport start it up again on
318 the passed-on TCP connection. Instead, proxy the child (and any
319 subsequent ones) for TLS via a unix-domain socket channel. Logging is
320 affected: the continued delivery log lines do not have any DNSSEC, TLS
321 Certificate or OCSP information. TLS cipher information is still logged.
323 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
324 identical IP addresses on different listening ports. Will also affect
327 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
328 add noisy ifdef guards to special-case this sillyness.
329 Patch from Bernd Kuhls.
331 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
332 than 255 are no longer allowed.
334 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
335 Disable the session-cache too, which might reduce our load. Since we
336 currrectly use a new context for every connection, both as server and
337 client, there is no benefit for these.
338 GnuTLS appears to not support tickets server-side by default (we don't
339 call gnutls_session_ticket_enable_server()) but client side is enabled
340 by default on recent versions (3.1.3 +) unless the PFS priority string
343 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
344 <https://reproducible-builds.org/specs/source-date-epoch/>.
346 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
347 the check for any unsuccessful recipients did not notice the limit, and
348 erroneously found still-pending ones.
350 JH/08 Pipeline CHUNKING command and data together, on kernels that support
351 MSG_MORE. Only in-clear (not on TLS connections).
353 JH/09 Avoid using a temporary file during transport using dkim. Unless a
354 transport-filter is involved we can buffer the headers in memory for
355 creating the signature, and read the spool data file once for the
356 signature and again for transmission.
358 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
359 4.77 as the kernel support then wasn't solid, having issues in 64bit
360 mode. Now, it's been long enough. Add support for FreeBSD also.
362 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
363 case where the routing stage had gathered several addresses to send to
364 a host before calling the transport for the first, we previously failed
365 to close down TLS in the old transport process before passing the TCP
366 connection to the new process. The new one sent a STARTTLS command
367 which naturally failed, giving a failed delivery and bloating the retry
368 database. Investigation and fix prototype from Wolfgang Breyha.
370 JH/12 Fix check on SMTP command input synchronisation. Previously there were
371 false-negatives in the check that the sender had not preempted a response
372 or prompt from Exim (running as a server), due to that code's lack of
373 awareness of the SMTP input buffering.
375 PP/04 Add commandline_checks_require_admin option.
376 Exim drops privileges sanely, various checks such as -be aren't a
377 security problem, as long as you trust local users with access to their
378 own account. When invoked by services which pass untrusted data to
379 Exim, this might be an issue. Set this option in main configuration
380 AND make fixes to the calling application, such as using `--` to stop
383 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
384 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
385 responses to those, into a single TLS record each way (this usually means
386 a single packet). As a side issue, smtp_enforce_sync now works on TLS
389 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
390 affects you only if you're dancing at the edge of the param size limits.
391 If you are, and this message makes sense to you, then: raise the
392 configured limit or use OpenSSL 1.1. Nothing we can do for older
395 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
396 cmdline element to get the documented default one. Previously it was
399 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
400 are unsuable for non-message-specific reasons - eg. connection timeout,
401 banner-time rejection.
403 JH/16 Fix logging of delivery remote port, when specified by router, under
406 PP/06 Repair manualroute's ability to take options in any order, even if one
407 is the name of a transport.
410 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
412 JH/17 Change the list-building routines interface to use the expanding-string
413 triplet model, for better allocation and copying behaviour.
415 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
416 Previously it was constructed the first time a possibly-matching string
417 was met in the configuration file input during startup; now it is done
420 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
421 compatible one, to avoid the (poorly documented) possibility of a config
422 file in the working directory redirecting the DB files, possibly correpting
423 some existing file. CVE-2017-10140 assigned for BDB.
425 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
426 cache-hot. Previously, although the result was properly cached, the
427 initial verify call returned a defer.
429 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
430 the main verify for receipient in uncached-mode.
432 JH/22 Retire historical build files to an "unsupported" subdir. These are
433 defined as "ones for which we have no current evidence of testing".
435 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
436 if present. Previously it was ignored.
438 JH/24 Start using specified-initialisers in C structure init coding. This is
439 a C99 feature (it's 2017, so now considered safe).
441 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
442 if was a fixed-sized field and bitmask ops via macros; it is now more
445 PP/07 GitHub PR 56: Apply MariaDB build fix.
446 Patch provided by Jaroslav Škarvada.
448 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
449 during Coverity cleanups [4.87 JH/47]
450 Diagnosis and fix provided by Michael Fischer v. Mollard.
452 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
453 the right size to place the terminating semicolon on its own folded
454 line, the header hash was calculated to an incorrect value thanks to
455 the (relaxed) space the fold became.
457 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
458 and confused the parent.
460 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
461 which could crash as a result. This could lead to undeliverable messages.
463 JH/28 Logging: "next input sent too soon" now shows where input was truncated
466 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
467 matters on fast-turnover and PID-randomising systems, which were getting
468 out-of-order delivery.
470 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
471 a possibly-overlapping copy. The symptom was that "Remote host closed
472 connection in response to HELO" was logged instead of the actual 4xx
475 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
476 Previously only that bufferd was discarded, resulting in SYMTP command
479 JH/32 DKIM: when a message has multiple signatures matching an identity given
480 in dkim_verify_signers, run the dkim acl once for each. Previously only
481 one run was done. Bug 2189.
483 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
484 "panic the current process" to "deliberately defer". The panic log is
485 still written with the problem list name; the mail and reject logs now
486 get a temp-reject line for the message that was being handled, saying
487 something like "domains check lookup or other defer". The SMTP 451
488 message is still "Temporary local problem".
490 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
491 A crafted sequence of BDAT commands could result in in-use memory beeing
492 freed. CVE-2017-16943.
494 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
495 from SMTP input. Previously it was always done; now only done for DATA
496 and not BDAT commands. CVE-2017-16944.
498 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
499 to the message (such as an overlong header line). Previously this was
500 not done and we did not exit BDAT mode. Followon from the previous item
501 though a different problem.
507 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
508 than -2003 did; needs libidn2 in addition to libidn.
510 JH/02 The path option on a pipe transport is now expanded before use.
512 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
513 Patch provided by "Björn", documentation fix added too.
515 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
516 missing a wire-to-host endian conversion.
518 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
519 close after a BDAT command line could be taken as a following command,
520 giving a synch failure. Fix by only checking for synch immediately
521 before acknowledging the chunk.
523 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
524 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
525 macro. Patches provided by Josh Soref.
527 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
528 Previously we did not; the RFC seems ambiguous and VRFY is not listed
529 by IANA as a service extension. However, John Klensin suggests that we
532 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
533 the dkim code may be unix-mode line endings rather than smtp wire-format
534 CRLF, so prepend a CR to any bare LF.
536 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
537 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
539 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
540 the first were themselves being wrongly included in the feed into dkim
541 processing; with most chunk sizes in use this resulted in an incorrect
542 body hash calculated value.
544 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
545 DKIM signature block, for verification. Although advised against by
546 standards it is specifically not ruled illegal.
548 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
550 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
551 missing a body hash (the bh= tag).
553 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
554 It seems that HAProxy sends the Proxy Protocol information in clear and
555 only then does a TLS startup, so do the same.
557 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
558 TCP connections (such as for Spamd) unless the daemon successfully set
559 Fast Open mode on its listening sockets. This fixes breakage seen on
560 too-old kernels or those not configured for Fast Open, at the cost of
561 requiring both directions being enabled for TFO, and TFO never being used
562 by non-daemon-related Exim processes.
564 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
565 endings, at least on the first header line. Try to canonify any that get
566 past that check, despite the cost.
568 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
569 now limited to an arbitrary five deep, while parsing addresses with the
570 strip_excess_angle_brackets option enabled.
572 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
573 instead leave the unprompted TLS handshake in socket buffer for the
574 TLS library to consume.
576 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
578 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
580 JH/16 Drop variables when they go out of scope. Memory management drops a whole
581 region in one operation, for speed, and this leaves assigned pointers
582 dangling. Add checks run only under the testsuite which checks all
583 variables at a store-reset and panics on a dangling pointer; add code
584 explicitly nulling out all the variables discovered. Fixes one known
585 bug: a transport crash, where a dangling pointer for $sending_ip_address
586 originally assigned in a verify callout, is re-used.
588 PP/06 Drop '.' from @INC in various Perl scripts.
590 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
592 PP/08 Reduce a number of compilation warnings under clang; building with
593 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
594 should be warning-free.
596 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
598 HS/01 Fix portability problems introduced by PP/08 for platforms where
599 realloc(NULL) is not equivalent to malloc() [SunOS et al].
601 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
602 chunk. This allows us to accept broken chunked messages. We need a more
603 general solution here.
605 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
606 already-broken messages in the queue.
608 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
610 JH/19 Fix reference counting bug in routing-generated-address tracking.
616 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
617 supports it and a size is available (ie. the sending peer gave us one).
619 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
620 years of being deprecated). The replacements are the ACLs
621 acl_smtp_mime and acl_not_smtp_mime.
623 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
624 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
625 or even an in-clear connection were permitted. Now, if the host lookup
626 was dnssec and dane was requested then the host is only used if the
627 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
628 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
629 if one fails this test.
630 This means that a poorly-configured remote DNS will make it incommunicado;
631 but it protects against a DNS-interception attack on it.
633 JH/04 Bug 1810: make continued-use of an open smtp transport connection
634 non-noisy when a race steals the message being considered.
636 JH/05 If main configuration option tls_certificate is unset, generate a
637 self-signed certificate for inbound TLS connections.
639 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
640 in rewrites and routers.
642 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
643 and logged a warning sing 4.83; now they are a configuration file error.
645 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
646 (lacking @domain). Apply the same qualification processing as RCPT.
648 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
650 JH/10 Support ${sha256:} applied to a string (as well as the previous
653 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
654 a cutthrough deliver is pending, as we always want to make a connection.
655 This also avoids re-routing the message when later placing the cutthrough
656 connection after a verify cache hit.
657 Do not update it with the verify result either.
659 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
660 when routing results in more than one destination address.
662 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
663 signing (which inhibits the cutthrough capability). Previously only
664 the presence of an option was tested; now an expansion evaluating as
665 empty is permissible (obviously it should depend only on data available
666 when the cutthrough connection is made).
668 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
669 the relevant preceding SMTP command did not note the pipelining mode.
671 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
672 Previously they were not counted.
674 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
675 as one having no matching records. Previously we deferred the message
676 that needed the lookup.
678 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
679 distinguished as "(=".
681 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
682 for missing MX records. Previously it only worked for missing A records.
684 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
686 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
687 after the data-go-ahead and data-ack. Patch from Jason Betts.
689 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
690 even for a "none" policy. Patch from Tony Meyer.
692 JH/22 Fix continued use of a connection for further deliveries. If a port was
693 specified by a router, it must also match for the delivery to be
696 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
697 When one of the recipients of a message was unsuitable for the connection
698 (has no matching addresses), we lost track of needing to mark it
699 deferred. As a result mail would be lost.
701 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
703 JH/25 Decoding ACL controls is now done using a binary search; the source code
704 takes up less space and should be simpler to maintain. Merge the ACL
705 condition decode tables also, with similar effect.
707 JH/26 Fix problem with one_time used on a redirect router which returned the
708 parent address unchanged. A retry would see the parent address marked as
709 delivered, so not attempt the (identical) child. As a result mail would
712 JH/27 Fix a possible security hole, wherein a process operating with the Exim
713 UID can gain a root shell. Credit to http://www.halfdog.net/ for
714 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
717 JH/28 Enable {spool,log} filesystem space and inode checks as default.
718 Main config options check_{log,spool}_{inodes,space} are now
719 100 inodes, 10MB unless set otherwise in the configuration.
721 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
722 Previously it only applied to the main-section connection policy
725 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
727 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
728 by me. Added RFC7919 DH primes as an alternative.
730 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
731 are not in the system include path.
733 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
734 GnuTLS, when a session startup failed (eg because the client disconnected)
735 Exim did stdio operations after fclose. This was exposed by a recent
736 change which nulled out the file handle after the fclose.
738 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
739 signed directly by the cert-signing cert, rather than an intermediate
740 OCSP-signing cert. This is the model used by LetsEncrypt.
742 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
744 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
745 an incoming connection.
747 HS/02 Bug 1802: Do not half-close the connection after sending a request
750 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
751 fallback to "prime256v1".
753 JH/34 SECURITY: Use proper copy of DATA command in error message.
754 Could leak key material. Remotely exploitable. CVE-2016-9963.
760 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
761 and 3.4.4 - once the server is enabled to respond to an OCSP request
762 it does even when not requested, resulting in a stapling non-aware
763 client dropping the TLS connection.
765 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
766 support variable-length bit vectors. No functional change.
768 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
769 The I= interface field on outgoing lines is now after the H= remote
770 host field, same as incoming lines. There is a separate
771 outgoing_interface log selector which allows you to disable the
774 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
775 If not running log_selector +smtp_connection the mainlog would be held
776 open indefinitely after a "too many connections" event, including to a
777 deleted file after a log rotate. Leave the per net connection logging
778 leaving it open for efficiency as that will be quickly detected by the
779 check on the next write.
781 HS/01 Bug 1671: Fix post transport crash.
782 Processing the wait-<transport> messages could crash the delivery
783 process if the message IDs didn't exist for some reason. When
784 using 'split_spool_directory=yes' the construction of the spool
785 file name failed already, exposing the same netto behaviour.
787 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
788 mime_regex ACL conditions.
790 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
791 to DSN fail messages (bounces): remote IP, remote greeting, remote response
792 to HELO, local diagnostic string.
794 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
795 log line to debug. Even when configured with a tls authenticator many
796 client connections are expected to not authenticate in this way, so
797 an authenticate fail is not an error.
799 HS/02 Add the Exim version string to the process info. This way exiwhat
800 gives some more detail about the running daemon.
802 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
803 matter for fast-change records such as DNSBLs.
805 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
806 retry record, even if constant. There may be multiple transports with
807 different interface settings and the retry behaviour needs to be kept
810 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
812 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
814 JH/10 Bug 840: fix log_defer_output option of pipe transport
816 JH/11 Bug 830: use same host for all RCPTS of a message, even under
817 hosts_randomize. This matters a lot when combined with mua_wrapper.
819 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
820 ${quote_pgsql:<string>} operator.
822 JH/13 Bug 1708: avoid misaligned access in cached lookup.
824 JH/14 Change header file name for freeradius-client. Relevant if compiling
825 with Radius support; from the Gentoo tree and checked under Fedora.
827 JH/15 Bug 1712: Introduce $prdr_requested flag variable
829 JH/16 Bug 1714: Permit an empty string as expansion result for transport
830 option transport_filter, meaning no filtering.
832 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
834 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
835 defaults to "*" (all hosts). The variable is now available when not built
836 with TLS, default unset, mainly to enable keeping the testsuite sane.
837 If a server certificate is not supplied (via tls_certificate) an error is
838 logged, and clients will find TLS connections fail on startup. Presumably
839 they will retry in-clear.
840 Packagers of Exim are strongly encouraged to create a server certificate
841 at installation time.
843 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
844 with the $config_file variable.
846 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
847 in transport context, after the attempt, and per-recipient. The latter type
848 is per host attempted. The event data is the error message, and the errno
849 information encodes the lookup type (A vs. MX) used for the (first) host,
850 and the trailing two digits of the smtp 4xx response.
852 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
853 to write to mainlog (or rejectlog, paniclog) in the window between file
854 creation and permissions/ownership being changed. Particularly affects
855 installations where exicyclog is run as root, rather than exim user;
856 result is that the running daemon panics and dies.
858 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
860 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
861 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
862 "pri" and "weight". Note that the previous implicit priority given by the
863 list order is no longer honoured.
865 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
868 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
869 by defining SUPPORT_SOCKS.
871 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
872 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
873 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
874 variables are renamed to proxy_{local,external}_{address,port}.
876 JH/27 Move Internationalisation support from Experimental to mainline, enabled
877 for a build by defining SUPPORT_I18N
879 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
880 of the query string, and make ${quote_redis:} do that quoting.
882 JH/29 Move Events support from Experimental to mainline, enabled by default
883 and removable for a build by defining DISABLE_EVENT.
885 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
887 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
888 cached by the daemon.
890 JH/32 Move Redis support from Experimental to mainline, enabled for a build
891 by defining LOOKUP_REDIS. The libhiredis library is required.
893 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
894 keys are given for lookup.
896 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
897 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
898 only supported when built with TLS support. The PolarSSL SHA routines
899 are still used when the TLS library is too old for convenient support.
901 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
902 openssl_options), for security. OpenSSL forces this from version 1.1.0
903 server-side so match that on older versions.
905 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
906 allocation for $value could be released as the expansion processing
907 concluded, but leaving the global pointer active for it.
909 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
910 and to use the domains and local_parts ACL conditions.
912 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
913 incorrectly not doubled on cutthrough transmission, hence seen as a
914 body-termination at the receiving system - resulting in truncated mails.
915 Commonly the sender saw a TCP-level error, and retransmitted the message
916 via the normal store-and-forward channel. This could result in duplicates
917 received - but deduplicating mailstores were liable to retain only the
918 initial truncated version.
920 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
922 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
924 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
925 we're in there, support oversigning also; bug 1309.
927 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
929 HS/04 Add support for keep_environment and add_environment options.
931 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
932 either intentional arithmetic overflow during PRNG, or testing config-
935 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
936 delivery resulted in actual delivery. Cancel cutthrough before DATA
939 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
940 rejects a recipient: pass the reject to the originator.
942 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
943 Many were false-positives and ignorable, but it's worth fixing the
946 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
947 for the new environment-manipulation done at startup. Move the routines
948 from being local to tls.c to being global via the os.c file.
950 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
951 an extract embedded as result-arg for a map, the first arg for extract
952 is unavailable so we cannot tell if this is a numbered or keyed
953 extraction. Accept either.
959 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
962 JH/02 The smtp transport option "multi_domain" is now expanded.
964 JH/03 The smtp transport now requests PRDR by default, if the server offers
967 JH/04 Certificate name checking on server certificates, when exim is a client,
968 is now done by default. The transport option tls_verify_cert_hostnames
969 can be used to disable this per-host. The build option
970 EXPERIMENTAL_CERTNAMES is withdrawn.
972 JH/05 The value of the tls_verify_certificates smtp transport and main options
973 default to the word "system" to access the system default CA bundle.
974 For GnuTLS, only version 3.0.20 or later.
976 JH/06 Verification of the server certificate for a TLS connection is now tried
977 (but not required) by default. The verification status is now logged by
978 default, for both outbound TLS and client-certificate supplying inbound
981 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
984 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
985 Status Notification (bounce) messages are now MIME format per RFC 3464.
986 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
987 under the control of the dsn_advertise_hosts option, and routers may
988 have a dsn_lasthop option.
990 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
991 default, modifiable by a malware= option. The list separator for
992 the options can now be changed in the usual way. Bug 68.
994 JH/10 The smtp_receive_timeout main option is now expanded before use.
996 JH/11 The incoming_interface log option now also enables logging of the
997 local interface on delivery outgoing connections.
999 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1000 if the interface and destination host and port all match.
1002 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1005 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1006 Patch from Andrew Lewis.
1008 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1009 now supports optional time-restrictions, weighting, and priority
1010 modifiers per server. Patch originally by <rommer@active.by>.
1012 JH/16 The spamd_address main option now supports a mixed list of local
1013 and remote servers. Remote servers can be IPv6 addresses, and
1014 specify a port-range.
1016 JH/17 Bug 68: The spamd_address main option now supports an optional
1017 timeout value per server.
1019 JH/18 Bug 1581: Router and transport options headers_add/remove can
1020 now have the list separator specified.
1022 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1025 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1028 JH/21 Support for the A6 type of dns record is withdrawn.
1030 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1031 rather than the verbs used.
1033 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1034 from 255 to 1024 chars.
1036 JH/24 Verification callouts now attempt to use TLS by default.
1038 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1039 are generic router options now. The defaults didn't change.
1041 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1042 Original patch from Alexander Shikoff, worked over by JH.
1044 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1045 files if ZCAT_COMMAND is not executable.
1047 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1049 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1051 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1052 Normally benign, it bites when the pair was led to by a CNAME;
1053 modern usage is to not canonicalize the domain to a CNAME target
1054 (and we were inconsistent anyway for A-only vs AAAA+A).
1056 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1058 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1059 when evaluating $sender_host_dnssec.
1061 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1062 $sender_helo_dnssec variable.
1064 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1066 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1068 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1070 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1071 documented as working, but never had. Support all but $spam_report.
1073 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1074 added for tls authenticator.
1076 HS/03 Add perl_taintmode main config option
1082 TL/01 When running the test suite, the README says that variables such as
1083 no_msglog_check are global and can be placed anywhere in a specific
1084 test's script, however it was observed that placement needed to be near
1085 the beginning for it to behave that way. Changed the runtest perl
1086 script to read through the entire script once to detect and set these
1087 variables, reset to the beginning of the script, and then run through
1088 the script parsing/test process like normal.
1090 TL/02 The BSD's have an arc4random API. One of the functions to induce
1091 adding randomness was arc4random_stir(), but it has been removed in
1092 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1093 function when detected.
1095 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1096 cause callback expansion.
1098 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1099 syntax errors in an expansion can be treated as a string instead of
1100 logging or causing an error, due to the internal use of bool_lax
1101 instead of bool when processing it.
1103 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1104 server certificates when making smtp deliveries.
1106 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1108 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1110 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1112 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1113 Merged patch from Sebastian Wiedenroth.
1115 JH/05 Fix results-pipe from transport process. Several recipients, combined
1116 with certificate use, exposed issues where response data items split
1117 over buffer boundaries were not parsed properly. This eventually
1118 resulted in duplicates being sent. This issue only became common enough
1119 to notice due to the introduction of connection certificate information,
1120 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1122 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1123 size buffer was used, resulting in syntax errors when an expansion
1126 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1127 version 3.3.6 or later.
1129 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1130 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1131 both become "event_action", the variables become $event_name, $event_data
1132 and $event_defer_errno. There is a new variable $verify_mode, usable in
1133 routers, transports and related events. The tls:cert event is now also
1134 raised for inbound connections, if the main configuration event_action
1137 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1138 early OCSP support, but no stapling (appears to be less than 1.0.0).
1140 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1141 server certificate names available under the smtp transport option
1142 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1145 JH/10 Time-related extraction expansions from certificates now use the main
1146 option "timezone" setting for output formatting, and are consistent
1147 between OpenSSL and GnuTLS compilations. Bug 1541.
1149 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1150 encoded parameter in the incoming message. Bug 1558.
1152 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1153 include certificate info, eximon was claiming there were spoolfile
1156 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1158 JH/14 Log delivery-related information more consistently, using the sequence
1159 "H=<name> [<ip>]" wherever possible.
1161 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1162 are problematic for Debian distribution, omit them from the release
1165 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1167 JH/16 Fix string representation of time values on 64bit time_t architectures.
1170 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1171 output list separator was used.
1176 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1177 checkers that were complaining about end of non-void function with no
1180 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1181 This was a regression introduced in 4.83 by another bugfix.
1183 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1185 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1186 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1192 TF/01 Correctly close the server side of TLS when forking for delivery.
1194 When a message was received over SMTP with TLS, Exim failed to clear up
1195 the incoming connection properly after forking off the child process to
1196 deliver the message. In some situations the subsequent outgoing
1197 delivery connection happened to have the same fd number as the incoming
1198 connection previously had. Exim would try to use TLS and fail, logging
1199 a "Bad file descriptor" error.
1201 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1202 utilities have not been installed.
1204 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1205 temporary space as the ACL may create new global variables.
1207 TL/01 LDAP support uses per connection or global context settings, depending
1208 upon the detected version of the libraries at build time.
1210 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1211 to extract and use the src ip:port in logging and expansions as if it
1212 were a direct connection from the outside internet. PPv2 support was
1213 updated based on HAProxy spec change in May 2014.
1215 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1217 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1218 Properly escape header and check for NULL return.
1220 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1223 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1225 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1226 characters in header names, implemented as a verify condition.
1227 Contributed by Michael Fischer v. Mollard.
1229 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1230 results permerror and temperror. Previous values are deprecated but
1231 still accepted. In a future release, err_perm and err_temp will be
1232 completely removed, which will be a backward incompatibility if the
1233 ACL tests for either of these two old results. Patch contributed by
1234 user bes-internal on the mailing list.
1236 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1238 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1239 selectors, in both main and reject logs.
1241 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1244 JH/07 Add malware type "sock" for talking to simple daemon.
1246 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1248 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1249 routers/transports under cutthrough routing.
1251 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1252 numbers. Touch up "bool" conditional to keep the same definition.
1254 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1256 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1258 JH/12 Expand items in router/transport headers_add or headers_remove lists
1259 individually rather than the list as a whole. Bug 1452.
1261 Required for reasonable handling of multiple headers_ options when
1262 they may be empty; requires that headers_remove items with embedded
1263 colons must have them doubled (or the list-separator changed).
1265 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1266 view the policy declared in the DMARC record. Currently, $dmarc_status
1267 is a combined value of both the record presence and the result of the
1270 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1272 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1273 dnslookup router and the smtp transport (applying to the forward
1276 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1277 of ldap servers used for a specific lookup. Patch provided by Heiko
1280 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1281 New variable $lookup_dnssec_authenticated for observability.
1283 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1284 Patch submitted by Lars Timman.
1286 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1288 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1289 Requires trusted mode and valid format message id, aborts otherwise.
1290 Patch contributed by Heiko Schlichting.
1292 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1293 certextract with support for various fields. Bug 1358.
1295 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1296 is requested by default, modifiable by smtp transport option
1299 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1300 operate on certificate variables to give certificate fingerprints
1301 Also new ${sha256:cert_variable}.
1303 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1305 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1306 Christian Aistleitner.
1308 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1310 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1311 file. Patch from Wolfgang Breyha.
1313 JH/25 Expand the coverage of the delivery $host and $host_address to
1314 client authenticators run in verify callout. Bug 1476.
1316 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1317 align with daemon_smtp_ports. Bug 72.
1319 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1320 support and error reporting did not work properly.
1322 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1323 and is readable. Patch from Andrew Colin Kissa.
1325 TL/14 Enhance documentation of ${run expansion and how it parses the
1326 commandline after expansion, particularly in the case when an
1327 unquoted variable expansion results in an empty value.
1329 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1331 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1332 by renaming away from it. Addresses GH issue 15, reported by
1335 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1337 TL/15 SECURITY: prevent double expansion in math comparison functions
1338 (can expand unsanitized data). Not remotely exploitable.
1345 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1347 PP/02 Make -n do something, by making it not do something.
1348 When combined with -bP, the name of an option is not output.
1350 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1353 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1354 $sender_host_name and config options to manage this, and basic check
1357 PP/05 DSCP support for outbound connections and control modifier for inbound.
1359 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1360 (Only plugin which currently uses this is kerberos4, which nobody should
1361 be using, but we should make it available and other future plugins might
1362 conceivably use it, even though it would break NAT; stuff *should* be
1363 using channel bindings instead).
1365 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1366 name; added for Sendmail compatibility; requires admin caller.
1367 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1368 just ignore it); requires trusted caller.
1369 Also parse but ignore: -Ac -Am -X<logfile>
1372 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1374 TL/02 Add +smtp_confirmation as a default logging option.
1376 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1377 Patch by Magnus Holmgren from 2007-02-20.
1379 TL/04 Bugzilla 1281 - Spec typo.
1380 Bugzilla 1283 - Spec typo.
1381 Bugzilla 1290 - Spec grammar fixes.
1383 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1385 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1387 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1388 mailing list by by Dmitry Isaikin.
1390 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1392 JH/02 Support "G" suffix to numbers in ${if comparisons.
1394 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1396 NM/01 Bugzilla 1197 - Spec typo
1397 Bugzilla 1196 - Spec examples corrections
1399 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1401 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1402 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1405 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1406 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1408 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1409 "acl {{name}{arg}...}", and optional args on acl condition
1412 JH/05 Permit multiple router/transport headers_add/remove lines.
1414 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1416 JH/07 Avoid using a waiting database for a single-message-only transport.
1417 Performance patch from Paul Fisher. Bugzilla 1262.
1419 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1422 JH/09 Add $headers_added variable, with content from use of ACL modifier
1423 add_header (but not yet added to the message). Bugzilla 199.
1425 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1426 Pulled from Bugzilla 817 by Wolfgang Breyha.
1428 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1430 (nb: this is the same fix as in Exim 4.80.1)
1432 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1435 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1438 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1440 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1441 advertise SMTP AUTH mechanism to us, instead of a generic
1442 protocol violation error. Also, make Exim more robust to bad
1443 data from the Dovecot auth socket.
1445 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1447 When a queue runner is handling a message, Exim first routes the
1448 recipient addresses, during which it prunes them based on the retry
1449 hints database. After that it attempts to deliver the message to
1450 any remaining recipients. It then updates the hints database using
1453 So if a recipient address works intermittently, it can get repeatedly
1454 deferred at routing time. The retry hints record remains fresh so the
1455 address never reaches the final cutoff time.
1457 This is a fairly common occurrence when a user is bumping up against
1458 their storage quota. Exim had some logic in its local delivery code
1459 to deal with this. However it did not apply to per-recipient defers
1460 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1462 This change adds a proper retry rule check during routing so that the
1463 final cutoff time is checked against the message's age. We only do
1464 this check if there is an address retry record and there is not a
1465 domain retry record; this implies that previous attempts to handle
1466 the address had the retry_use_local_parts option turned on. We use
1467 this as an approximation for the destination being like a local
1468 delivery, as in LMTP.
1470 I suspect this new check makes the old local delivery cutoff check
1471 redundant, but I have not verified this so I left the code in place.
1473 TF/02 Correct gecos expansion when From: is a prefix of the username.
1475 Test 0254 submits a message to Exim with the header
1479 When I ran the test suite under the user fanf2, Exim expanded
1480 the header to contain my full name, whereas it should have added
1481 a Resent-Sender: header. It erroneously treats any prefix of the
1482 username as equal to the username.
1484 This change corrects that bug.
1486 GF/01 DCC debug and logging tidyup
1487 Error conditions log to paniclog rather than rejectlog.
1488 Debug lines prefixed by "DCC: " to remove any ambiguity.
1490 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1492 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1493 Bug spotted by Jeremy Harris; was flawed since initial commit.
1494 Would have resulted in OCSP responses post-SNI triggering an Exim
1495 NULL dereference and crash.
1497 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1499 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1500 Bug detection, analysis and fix by Samuel Thibault.
1501 Bugzilla 1331, Debian bug #698092.
1503 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1505 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1506 Server implementation by Todd Lyons, client by JH.
1507 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1508 config variable "prdr_enable" controls whether the server
1509 advertises the facility. If the client requests PRDR a new
1510 acl_data_smtp_prdr ACL is called once for each recipient, after
1511 the body content is received and before the acl_smtp_data ACL.
1512 The client is controlled by both of: a hosts_try_prdr option
1513 on the smtp transport, and the server advertisement.
1514 Default client logging of deliveries and rejections involving
1515 PRDR are flagged with the string "PRDR".
1517 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1518 fclose(). Diagnosis by Todd Lyons.
1520 PP/17 Update configure.default to handle IPv6 localhost better.
1521 Patch by Alain Williams (plus minor tweaks).
1524 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1525 This is now consistent with GnuTLS, and is now documented: the
1526 previous undocumented portable approach to treating the option as
1527 unset was to force an expansion failure. That still works, and
1528 an empty string is now equivalent.
1530 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1531 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1532 not performing validation itself.
1534 PP/20 Added force_command boolean option to pipe transport.
1535 Patch from Nick Koston, of cPanel Inc.
1537 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1540 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1542 PP/21 Fix eximon continuous updating with timestamped log-files.
1543 Broken in a format-string cleanup in 4.80, missed when I repaired the
1544 other false fix of the same issue.
1545 Report and fix from Heiko Schlichting.
1548 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1549 Report from Prashanth Katuri.
1551 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1552 It's SecureTransport, so affects any MacOS clients which use the
1553 system-integrated TLS libraries, including email clients.
1555 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1556 using a MIME ACL for non-SMTP local injection.
1557 Report and assistance in diagnosis by Warren Baker.
1559 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1561 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1563 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1564 last id that failed so it may be referenced in subsequent ACL's.
1566 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1569 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1570 ldap library initialization, allowing self-signed CA's to be
1571 used. Also properly sets require_cert option later in code by
1572 using NULL (global ldap config) instead of ldap handle (per
1573 session). Bug diagnosis and testing by alxgomz.
1575 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1576 the src/util/ subdirectory.
1578 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1579 renamed to Transport Post Delivery Action by Jeremy Harris, as
1582 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1583 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1584 redis_servers = needs to be configured which will be used by the redis
1585 lookup. Patch from Warren Baker, of The Packet Hub.
1587 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1589 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1590 hostname or reverse DNS when processing a host list. Used suggestions
1591 from multiple comments on this bug.
1593 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1595 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1596 Missed a few lines, added it to make the runtest require no keyboard
1599 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1600 contains upper case chars. Make router use caseful_local_part.
1602 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1603 support when GnuTLS has been built with p11-kit.
1609 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1611 This, or similar/improved, will also be change PP/11 of 4.82.
1617 PP/01 Handle short writes when writing local log-files.
1618 In practice, only affects FreeBSD (8 onwards).
1619 Bugzilla 1053, with thanks to Dmitry Isaikin.
1621 NM/01 Bugzilla 949 - Documentation tweak
1623 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1626 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1628 PP/02 Implemented gsasl authenticator.
1630 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1632 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1633 `pkg-config foo` for cflags/libs.
1635 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1636 with rest of GSASL and with heimdal_gssapi.
1638 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1639 `pkg-config foo` for cflags/libs for the TLS implementation.
1641 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1642 properties get this fed in as external SSF. A number of robustness
1643 and debugging improvements to the cyrus_sasl authenticator.
1645 PP/08 cyrus_sasl server now expands the server_realm option.
1647 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1648 Patch by Jeremy Harris.
1650 PP/10 Added dbmjz lookup type.
1652 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1654 PP/12 MAIL args handles TAB as well as SP, for better interop with
1655 non-compliant senders.
1656 Analysis and variant patch by Todd Lyons.
1658 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1659 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1660 Patch from Dirk Mueller <dmueller@suse.com>
1662 PP/13 tls_peerdn now print-escaped for spool files.
1663 Observed some $tls_peerdn in wild which contained \n, which resulted
1664 in spool file corruption.
1666 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1667 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1668 or write after TLS renegotiation, which otherwise led to messages
1671 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1672 as a tracking header (ie: a signed header comes before the signature).
1673 Patch from Wolfgang Breyha.
1675 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1676 comma-sep list; embedded commas doubled.
1678 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1680 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1682 Report and patch from Dmitry Banschikov.
1684 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1685 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1686 failures. We appear to now support TLS1.1+ with Exim.
1688 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1689 lets Exim select keys and certificates based upon TLS SNI from client.
1690 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1691 before an outbound SMTP session. New log_selector, +tls_sni.
1693 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1694 NULL dereference. Report and patch from Alun Jones.
1696 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1697 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1698 Not seeing resolver debug output on NetBSD, but suspect this is a
1699 resolver implementation change.
1701 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1702 Left warnings. Added "eximon gdb" invocation mode.
1704 PP/21 Defaulting "accept_8bitmime" to true, not false.
1706 PP/22 Added -bw for inetd wait mode support.
1708 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1709 locate the relevant includes and libraries. Made this the default.
1711 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1712 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1714 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1715 This may cause build issues on older platforms.
1717 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1718 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1719 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1720 Added SNI support via GnuTLS too.
1721 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1723 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1725 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1726 Applied second patch from Janne, implementing suggestion to default
1727 multiple-strings-in-record handling to match SPF spec.
1729 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1731 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1732 read-only, out of scope).
1733 Patch from Wolfgang Breyha, report from Stuart Northfield.
1735 PP/29 Fix three issues highlighted by clang analyser static analysis.
1736 Only crash-plausible issue would require the Cambridge-specific
1737 iplookup router and a misconfiguration.
1738 Report from Marcin Mirosław.
1740 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1742 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1743 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1744 As part of this, removing so much warning spew let me fix some minor
1745 real issues in debug logging.
1747 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1748 assignment on my part. Fixed.
1750 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1751 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1752 Janne Snabb (who went above and beyond: thank you).
1754 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1755 string otherwise requires a connection and a bunch more work and it's
1756 relatively easy to get wrong. Should also expose TLS library linkage
1759 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1760 64-bit ${eval} (JH/03).
1762 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1763 GNU libc to support some of the 64-bit stuff, should not lead to
1764 conflicts. Defined before os.h is pulled in, so if a given platform
1765 needs to override this, it can.
1767 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1768 protection layer was required, which is not implemented.
1769 Bugzilla 1254, patch from Wolfgang Breyha.
1771 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1772 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1773 tls_dhparam take prime identifiers. Also unbreak combination of
1774 OpenSSL+DH_params+TLSSNI.
1776 PP/39 Disable SSLv2 by default in OpenSSL support.
1782 PP/01 Solaris build fix for Oracle's LDAP libraries.
1783 Bugzilla 1109, patch from Stephen Usher.
1785 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1787 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1790 TF/02 Fix a couple more cases where we did not log the error message
1791 when unlink() failed. See also change 4.74-TF/03.
1793 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1794 lock up or crash if it happened to be inside a call to libc when it
1795 got a SIGUSR1 from exiwhat.
1797 The SIGUSR1 handler appends the current process status to the process
1798 log which is later printed by exiwhat. It used to use the general
1799 purpose logging code to do this, but several functions it calls are
1800 not safe for signals.
1802 The new output code in the SIGUSR1 handler is specific to the process
1803 log, and simple enough that it's easy to inspect for signal safety.
1804 Removing some special cases also simplifies the general logging code.
1805 Removing the spurious timestamps from the process log simplifies
1808 TF/04 Improved ratelimit ACL condition.
1810 The /noupdate option has been deprecated in favour of /readonly which
1811 has clearer semantics. The /leaky, /strict, and /readonly update modes
1812 are mutually exclusive. The update mode is no longer included in the
1813 database key; it just determines when the database is updated. (This
1814 means that when you upgrade Exim will forget old rate measurements.)
1816 Exim now checks that the per_* options are used with an update mode that
1817 makes sense for the current ACL. For example, when Exim is processing a
1818 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1819 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1820 must specify per_mail/readonly. If you omit the update mode it defaults to
1821 /leaky where that makes sense (as before) or /readonly where required.
1823 The /noupdate option is now undocumented but still supported for
1824 backwards compatibility. It is equivalent to /readonly except that in
1825 ACLs where /readonly is required you may specify /leaky/noupdate or
1826 /strict/noupdate which are treated the same as /readonly.
1828 A useful new feature is the /count= option. This is a generalization
1829 of the per_byte option, so that you can measure the throughput of other
1830 aggregate values. For example, the per_byte option is now equivalent
1831 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1833 The per_rcpt option has been generalized using the /count= mechanism
1834 (though it's more complicated than the per_byte equivalence). When it is
1835 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1836 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1837 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1838 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1839 Note that using per_rcpt with a non-readonly update mode in more than
1840 one ACL will cause the recipients to be double-counted. (The per_mail
1841 and per_byte options don't have this problem.)
1843 The handling of very low rates has changed slightly. If the computed rate
1844 is less than the event's count (usually one) then this event is the first
1845 after a long gap. In this case the rate is set to the same as this event's
1846 count, so that the first message of a spam run is counted properly.
1848 The major new feature is a mechanism for counting the rate of unique
1849 events. The new per_addr option counts the number of different
1850 recipients that someone has sent messages to in the last time period. It
1851 behaves like per_rcpt if all the recipient addresses are different, but
1852 duplicate recipient addresses do not increase the measured rate. Like
1853 the /count= option this is a general mechanism, so the per_addr option
1854 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1855 example, measure the rate that a client uses different sender addresses
1856 with the options per_mail/unique=$sender_address. There are further
1857 details in the main documentation.
1859 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1861 TF/06 Removed a few PCRE remnants.
1863 TF/07 Automatically extract Exim's version number from tags in the git
1864 repository when doing development or release builds.
1866 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1867 Bugzilla 879. Patch from Paul Fisher.
1869 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1870 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1873 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1875 PP/05 Variable $av_failed, true if the AV scanner deferred.
1876 Bugzilla 1078. Patch from John Horne.
1878 PP/06 Stop make process more reliably on build failure.
1879 Bugzilla 1087. Patch from Heiko Schlittermann.
1881 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1882 Bugzilla 1089. Patch from Heiko Schlittermann.
1884 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1885 Bugzilla 1131. Patch from Holger Weiß.
1887 PP/09 Handle IPv6 addresses with SPF.
1888 Bugzilla 860. Patch from Wolfgang Breyha.
1890 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1892 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1895 PP/11 match_* no longer expand right-hand-side by default.
1896 New compile-time build option, EXPAND_LISTMATCH_RHS.
1897 New expansion conditions, "inlist", "inlisti".
1899 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1901 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1903 PP/14 fix log_write() format string regression from TF/03.
1904 Bugzilla 1152. Patch from Dmitry Isaikin.
1910 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1912 PP/02 Harmonised TLS library version reporting; only show if debugging.
1913 Layout now matches that introduced for other libraries in 4.74 PP/03.
1915 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1917 PP/04 New "dns_use_edns0" global option.
1919 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1922 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1923 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1925 TK/01 Updated PolarSSL code to 0.14.2.
1926 Bugzilla 1097. Patch from Andreas Metzler.
1928 PP/07 Catch divide-by-zero in ${eval:...}.
1929 Fixes bugzilla 1102.
1931 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1934 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1935 format-string attack -- SECURITY: remote arbitrary code execution.
1937 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1938 time unintentionally subject to list matching rules, letting the header
1939 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1940 arbitrary string expansion). This allowed for information disclosure.
1942 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1943 INT_MIN/-1 -- value coerced to INT_MAX.
1949 NM/01 Workaround for PCRE version dependency in version reporting
1952 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1953 This fixes portability to compilers other than gcc, notably
1954 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1956 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1957 makefiles for portability to HP-UX and POSIX correctness.
1959 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1960 Also via indented variable definition in the Makefile.
1961 (Debugging by Oliver Heesakkers).
1963 PP/02 Restore caching of spamd results with expanded spamd_address.
1964 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1966 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1967 Improves build reliability. Fix from: Frank Elsner
1969 NM/02 Fix wide character breakage in the rfc2047 coding
1970 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1972 NM/03 Allow underscore in dnslist lookups
1973 Fixes bug 1026. Patch from Graeme Fowler
1975 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1976 Code patches from Adam Ciarcinski of NetBSD.
1978 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1981 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1982 is logged, to avoid truncation. Patch from John Horne.
1984 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1985 Patch from Jakob Hirsch.
1987 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1988 SQL string expansion failure details.
1989 Patch from Andrey Oktyabrski.
1991 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1992 Patch from Simon Arlott.
1994 PP/09 New lookups functionality failed to compile on old gcc which rejects
1995 extern declarations in function scope.
1996 Patch from Oliver Fleischmann
1998 PP/10 Use sig_atomic_t for flags set from signal handlers.
1999 Check getgroups() return and improve debugging.
2000 Fixed developed for diagnosis in bug 927 (which turned out to be
2003 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2004 Patch from Mark Zealey.
2006 PP/12 Bugzilla 1056: Improved spamd server selection.
2007 Patch from Mark Zealey.
2009 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2010 Based on patch from Heiko Schlittermann.
2012 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2013 Patch from Uwe Doering, sign-off by Michael Haardt.
2015 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2016 variable declaration deep within a block. Bug and patch from
2019 PP/15 lookups-Makefile IRIX compatibility coercion.
2021 PP/16 Make DISABLE_DKIM build knob functional.
2023 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2024 Patch by Simon Arlott
2026 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2027 variable argument macros. Our copy now differs from upstream.
2033 TF/01 Failure to get a lock on a hints database can have serious
2034 consequences so log it to the panic log.
2036 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2037 controlled using the smtp_confirmation log selector.
2039 TF/03 Include the error message when we fail to unlink a spool file.
2041 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2042 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2043 for maintaining out-of-tree patches for some time.
2045 PP/01 Bugzilla 139: Documentation and portability issues.
2046 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2047 Handle per-OS dynamic-module compilation flags.
2049 PP/02 Let /dev/null have normal permissions.
2050 The 4.73 fixes were a little too stringent and complained about the
2051 permissions on /dev/null. Exempt it from some checks.
2052 Reported by Andreas M. Kirchwitz.
2054 PP/03 Report version information for many libraries, including
2055 Exim version information for dynamically loaded libraries. Created
2056 version.h, now support a version extension string for distributors
2057 who patch heavily. Dynamic module ABI change.
2059 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2060 privilege escalation vulnerability whereby the Exim run-time user
2061 can cause root to append content of the attacker's choosing to
2064 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2067 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2068 If dropping privileges for untrusted macros, we disabled normal logging
2069 on the basis that it would fail; for the Exim run-time user, this is not
2070 the case, and it resulted in successful deliveries going unlogged.
2071 Fixed. Reported by Andreas Metzler.
2077 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2078 only prepend for the Resent-* case. Fixes regression introduced in
2079 Exim 4.70 by NM/22 for Bugzilla 607.
2081 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2082 increasing numbers of administrators be bitten by this.
2084 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2086 PP/03 Bugzilla 994: added openssl_options main configuration option.
2088 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2090 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2092 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2094 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2095 without a peer certificate, leading to a segfault because of an
2096 assumption that peers always have certificates. Be a little more
2097 paranoid. Problem reported by Martin Tscholak.
2099 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2100 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2101 NB: ClamAV planning to remove STREAM in "middle of 2010".
2102 CL also introduces -bmalware, various -d+acl logging additions and
2103 more caution in buffer sizes.
2105 PP/09 Implemented reverse_ip expansion operator.
2107 PP/10 Bugzilla 937: provide a "debug" ACL control.
2109 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2111 PP/12 Bugzilla 973: Implement --version.
2113 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2115 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2117 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2119 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2120 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2121 ignore trailing whitespace.
2123 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2125 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2128 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2129 Notification from Dr Andrew Aitchison.
2131 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2132 ExtendedDetectionInfo response format.
2133 Notification from John Horne.
2135 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2138 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2139 XSL and documented dependency on system catalogs, with examples of how
2142 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2145 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2146 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2147 configuration file which is writeable by the Exim user or group.
2149 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2150 of configuration files to cover files specified with the -C option if
2151 they are going to be used with root privileges, not just the default
2154 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2155 option (effectively making it always true).
2157 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2158 files to be used while preserving root privileges.
2160 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2161 that rogue child processes cannot use them.
2163 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2164 run-time user, instead of root.
2166 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2167 Exim run-time user without dropping privileges.
2169 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2170 result string, instead of calling string_vformat() twice with the same
2173 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2174 for other users. Others should always drop root privileges if they use
2175 -C on the command line, even for a whitelisted configure file.
2177 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2179 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2185 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2186 $data_path, and $header_path variables; fixed documentation bugs and
2189 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2190 exipick to access non-standard spools, including the "frozen" queue
2193 NM/01 Bugzilla 965: Support mysql stored procedures.
2194 Patch from Alain Williams
2196 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2198 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2199 Patch from Andreas Metzler
2201 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2202 Patch from Kirill Miazine
2204 NM/05 Bugzilla 671: Added umask to procmail example.
2206 JJ/03 installed exipick 20100323.0, fixing doc bug
2208 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2209 directory. Notification and patch from Dan Rosenberg.
2211 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2213 TK/02 Improve log output when DKIM signing operation fails.
2215 MH/01 Treat the transport option dkim_domain as a colon separated
2216 list, not as a single string, and sign the message with each element,
2217 omitting multiple occurences of the same signer.
2219 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2220 Bugzilla 985, 986. Patch by Simon Arlott
2222 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2223 Patch by Simon Arlott
2225 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2226 MBX locking. Notification from Dan Rosenberg.
2232 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2234 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2236 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2238 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2240 NM/04 Bugzilla 915: Flag broken perl installation during build.
2246 TK/01 Added patch by Johannes Berg that expands the main option
2247 "spamd_address" if it starts with a dollar sign.
2249 TK/02 Write list of recipients to X-Envelope-Sender header when building
2250 the mbox-format spool file for content scanning (suggested by Jakob
2253 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2254 (http://www.dcc-servers.net/) support via dccifd. Activated by
2255 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2257 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2258 by Mark Daniel Reidel <mr@df.eu>.
2260 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2261 When building exim an external PCRE library is now needed -
2262 PCRE is a system library on the majority of modern systems.
2263 See entry on PCRE_LIBS in EDITME file.
2265 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2266 conversation. Added nologin parameter to request.
2267 Patch contributed by Kirill Miazine.
2269 TF/01 Do not log submission mode rewrites if they do not change the address.
2271 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2273 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2274 log files in place. Contributed by Roberto Lima.
2276 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2278 TF/03 Bugzilla 615: When checking the local_parts router precondition
2279 after a local_part_suffix or local_part_prefix option, Exim now
2280 does not use the address's named list lookup cache, since this
2281 contains cached lookups for the whole local part.
2283 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2284 Robert Millan. Documentation is in experimental-spec.txt.
2286 TF/04 Bugzilla 668: Fix parallel build (make -j).
2288 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2290 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2291 Patch provided by Jan Srzednicki.
2293 TF/05 Leading white space used to be stripped from $spam_report which
2294 wrecked the formatting. Now it is preserved.
2296 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2297 that they are available at delivery time.
2299 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2301 TF/08 TLS error reporting now respects the incoming_interface and
2302 incoming_port log selectors.
2304 TF/09 Produce a more useful error message if an SMTP transport's hosts
2305 setting expands to an empty string.
2307 NM/06 Bugzilla 744: EXPN did not work under TLS.
2308 Patch provided by Phil Pennock.
2310 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2311 Patch provided by Richard Godbee.
2313 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2314 acl_smtp_notquit, added index entry.
2316 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2317 Patch provided by Eugene Bujak.
2319 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2320 accept(). Patch provided by Maxim Dounin.
2322 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2323 Patch provided by Phil Pennock.
2325 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2327 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2328 Patch provided by Brad "anomie" Jorsch.
2330 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2331 Patch provided by Dean Brooks.
2333 TK/05 Add native DKIM support (does not depend on external libraries).
2335 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2336 Patch provided by Graeme Fowler.
2338 NM/16 Bugzilla 851: Documentation example syntax fix.
2340 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2342 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2345 NM/19 Bugzilla 745: TLS version reporting.
2346 Patch provided by Phil Pennock.
2348 NM/20 Bugzilla 167: bool: condition support.
2349 Patch provided by Phil Pennock.
2351 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2352 clients. Patch provided by Phil Pennock.
2354 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2355 Patch provided by Brad "anomie" Jorsch.
2357 NM/23 Bugzilla 687: Fix misparses in eximstats.
2358 Patch provided by Heiko Schlittermann.
2360 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2361 Patch provided by Heiko Schlittermann.
2363 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2364 plus update to original patch.
2366 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2368 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2369 Patch provided by David Brownlee.
2371 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2373 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2375 NM/30 Bugzilla 888: TLS documentation bugfixes.
2377 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2379 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2380 Unlike the original bugzilla I have changed all shell scripts in src tree.
2382 NM/33 Bugzilla 898: Transport filter timeout fix.
2383 Patch by Todd Rinaldo.
2385 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2386 Patch by Serge Demonchaux.
2388 NM/35 Bugzilla 39: Base64 decode bug fixes.
2389 Patch by Jakob Hirsch.
2391 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2393 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2395 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2397 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2403 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2404 ALT-N's libdkim that I have put here:
2405 http://duncanthrax.net/exim-experimental/
2407 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2408 were called 'true' and it seems that C99 defines that as a reserved
2409 keyword to be used with 'bool' variable types. That means you could
2410 not include C99-style headers which use bools without triggering
2411 build errors in sieve.c.
2413 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2414 as mailq or other aliases. Changed the --help handling significantly
2415 to do whats expected. exim_usage() emits usage/help information.
2417 SC/01 Added the -bylocaldomain option to eximstats.
2419 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2421 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2423 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2429 PH/01 Another patch from the Sieve maintainer.
2431 PH/02 When an IPv6 address is converted to a string for single-key lookup
2432 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2433 dots are used instead of colons so that keys in lsearch files need not
2434 contain colons. This was done some time before quoting was made available
2435 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2436 (notated using the quote facility) so as to distinguish them from IPv4
2437 keys. This meant that lookups for IP addresses in host lists did not work
2438 for iplsearch lookups.
2440 This has been fixed by arranging for IPv6 addresses to be expressed with
2441 colons if the lookup type is iplsearch. This is not incompatible, because
2442 previously such lookups could never work.
2444 The situation is now rather anomalous, since one *can* have colons in
2445 ordinary lsearch keys. However, making the change in all cases is
2446 incompatible and would probably break a number of configurations.
2448 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2451 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2452 conversion specification without a maximum field width, thereby enabling
2453 a rogue spamd server to cause a buffer overflow. While nobody in their
2454 right mind would setup Exim to query an untrusted spamd server, an
2455 attacker that gains access to a server running spamd could potentially
2456 exploit this vulnerability to run arbitrary code as the Exim user.
2458 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2459 $primary_hostname instead of what libspf2 thinks the hosts name is.
2461 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2462 a directory entry by the name of the lookup key. Previously, if a
2463 symlink pointed to a non-existing file or a file in a directory that
2464 Exim lacked permissions to read, a lookup for a key matching that
2465 symlink would fail. Now it is enough that a matching directory entry
2466 exists, symlink or not. (Bugzilla 503.)
2468 PH/03 The body_linecount and body_zerocount variables are now exported in the
2471 PH/04 Added the $dnslist_matched variable.
2473 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2474 This means they are set thereafter only if the connection becomes
2477 PH/06 Added the client_condition to authenticators so that some can be skipped
2478 by clients under certain conditions.
2480 PH/07 The error message for a badly-placed control=no_multiline_responses left
2481 "_responses" off the end of the name.
2483 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2485 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2486 (without spaces) instead of just copying the configuration text.
2488 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2490 PH/11 Added $max_received_linelength.
2492 PH/12 Added +ignore_defer and +include_defer to host lists.
2494 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2495 way in which PCRE > 7.0 is built.
2497 PH/14 Implemented queue_only_load_latch.
2499 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2500 MAIL command. The effect was to mangle the value on 64-bit systems.
2502 PH/16 Another patch from the Sieve maintainer.
2504 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2506 PH/18 If a system quota error occurred while trying to create the file for
2507 a maildir delivery, the message "Mailbox is full" was not appended to the
2508 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2509 only to a quota excession during the actual writing of the file.
2511 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2512 characters?) which causes problems in log lines. The DN values are now
2513 passed through string_printing() before being added to log lines.
2515 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2516 and InterBase are left for another time.)
2518 PH/21 Added message_body_newlines option.
2520 PH/22 Guard against possible overflow in moan_check_errorcopy().
2522 PH/23 POSIX allows open() to be a macro; guard against that.
2524 PH/24 If the recipient of an error message contained an @ in the local part
2525 (suitably quoted, of course), incorrect values were put in $domain and
2526 $local_part during the evaluation of errors_copy.
2532 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2533 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2536 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2537 issue a MAIL command.
2539 PH/02 In an ACL statement such as
2541 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2543 if a client was not listed at all, or was listed with a value other than
2544 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2545 the condition was not true (as it should be), so access was not denied.
2546 The bug was that the ! inversion was incorrectly passed on to the second
2547 item. This has been fixed.
2549 PH/03 Added additional dnslists conditions == and =& which are different from
2550 = and & when the dns lookup returns more than one IP address.
2552 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2553 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2555 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2556 FSYNC, which compiles an option called disable_fsync that allows for
2557 bypassing fsync(). The documentation is heavily laced with warnings.
2559 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2561 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2562 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2563 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2564 including adding "make clean"; (3) Added -fPIC when compiling the test
2565 dynamically loaded module, to get rid of a warning.
2567 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2568 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2569 = 0s. The bug is otherwise harmless.
2571 PH/07 There was a bug in the dovecot authenticator such that the value of
2572 $auth1 could be overwritten, and so not correctly preserved, after a
2573 successful authentication. This usually meant that the value preserved by
2574 the server_setid option was incorrect.
2576 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2578 PH/09 Installed PCRE release 7.0.
2580 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2581 run for batched SMTP input. It is now run at the start of every message
2582 in the batch. While fixing this I discovered that the process information
2583 (output by running exiwhat) was not always getting set for -bs and -bS
2584 input. This is fixed, and it now also says "batched" for BSMTP.
2586 PH/11 Added control=no_pipelining.
2588 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2589 patch, slightly modified), and move the expansion of helo_data till after
2590 the connection is made in the smtp transport (so it can use these
2593 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2595 PH/14 Added log_selector = +pid.
2597 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2599 PH/16 Add ${if forany and ${if forall.
2601 PH/17 Added dsn_from option to vary the From: line in DSNs.
2603 PH/18 Flush SMTP output before performing a callout, unless control =
2604 no_callout_flush is set.
2606 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2607 was true (the default) a successful delivery failed to delete the retry
2608 item, thus causing premature timeout of the address. The bug is now
2611 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2613 PH/21 Long custom messages for fakedefer and fakereject are now split up
2614 into multiline responses in the same way that messages for "deny" and
2615 other ACL rejections are.
2617 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2618 with slight modification.
2620 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2621 draft, changing the syntax and factoring some duplicate code".
2623 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2624 for deliveries of the second and subsequent messages over the same SMTP
2627 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2628 ${reduce, with only minor "tidies".
2630 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2632 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2633 expansion side effects.
2635 PH/27 When a message times out after an over-quota error from an Exim-imposed
2636 quota, the bounce message says "mailbox is full". This message was not
2637 being given when it was a system quota that was exceeded. It now should
2640 MH/03 Made $recipients available in local_scan(). local_scan() already has
2641 better access to the recipient list through recipients_list[], but
2642 $recipients can be useful in postmaster-provided expansion strings.
2644 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2645 in the case of a MAIL command with additional options following the
2646 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2647 were accidentally chopped off.
2649 PH/29 SMTP synchronization checks are implemented when a command is read -
2650 there is a check that no more input is waiting when there shouldn't be
2651 any. However, for some commands, a delay in an ACL can mean that it is
2652 some time before the response is written. In this time, more input might
2653 arrive, invalidly. So now there are extra checks after an ACL has run for
2654 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2655 pipelining has not been advertised.
2657 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2659 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2660 returned in the SMTP response when smtp_return_error_details was false.
2661 This has been fixed.
2663 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2664 instead of the C I/O that was originally supplied, because problems were
2665 reported on Solaris.
2667 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2668 Exim which did not show up earlier: it was assuming that a call to
2669 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2670 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2671 was a macro that became an assignment, so it seemed to work. This has
2672 changed to a proper function call with a void return, hence the compile
2673 error. Exim's code has been fixed.
2675 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2678 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2680 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2682 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2683 because it turns out to be too big or there is a timeout) there is no
2684 "Completed" line in the log. When some messages of this type were
2685 selected by exigrep, they were listed as "not completed". Others were
2686 picked up by some special patterns. I have improved the selection
2687 criteria to be more general.
2689 PH/38 The host_find_failed option in the manualroute router can now be set
2690 to "ignore", to completely ignore a host whose IP address cannot be
2691 found. If all hosts are ignored, the behaviour is controlled by the new
2692 host_all_ignored option.
2694 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2695 homing or because of multiple MX records with /mx) generated more than
2696 one IP address, and the following item turned out to be the local host,
2697 all the secondary addresses of the first item were incorrectly removed
2698 from the list, along with the local host and any following hosts (which
2699 is what is supposed to happen).
2701 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2702 whoever called Exim into the -H file. In the case of the daemon it was
2703 behaving confusingly. When first started, it used values for whoever
2704 started the daemon, but after a SIGHUP it used the Exim user (because it
2705 calls itself on a restart). I have changed the code so that it now always
2708 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2709 message are rejected with the same error (e.g. no authentication or bad
2710 sender address), and a DATA command is nevertheless sent (as can happen
2711 with PIPELINING or a stupid MUA), the error message that was given to the
2712 RCPT commands is included in the rejection of the DATA command. This is
2713 intended to be helpful for MUAs that show only the final error to their
2716 PH/42 Another patch from the Sieve maintainer.
2718 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2719 Eximstats - Fixed some broken HTML links and added missing column headers
2721 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2722 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2724 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2726 PH/43 Yet another patch from the Sieve maintainer.
2728 PH/44 I found a way to check for a TCP/IP connection going away before sending
2729 the response to the final '.' that terminates a message, but only in the
2730 case where the client has not sent further data following the '.'
2731 (unfortunately, this is allowed). However, in many cases there won't be
2732 any further data because there won't be any more messages to send. A call
2733 to select() can be used: if it shows that the input is "ready", there is
2734 either input waiting, or the socket has been closed. An attempt to read
2735 the next input character can distinguish the two cases. Previously, Exim
2736 would have sent an OK response which the client would never have see.
2737 This could lead to message repetition. This fix should cure that, at
2738 least in a lot of common cases.
2740 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2741 advertised in response to EHLO.
2747 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2748 fixed by 4.65/MH/01 (is this a record?) are fixed:
2750 (i) An empty string was always treated as zero by the numeric comparison
2751 operators. This behaviour has been restored.
2753 (ii) It is documented that the numeric comparison operators always treat
2754 their arguments as decimal numbers. This was broken in that numbers
2755 starting with 0 were being interpreted as octal.
2757 While fixing these problems I realized that there was another issue that
2758 hadn't been noticed. Values of message_size_limit (both the global option
2759 and the transport option) were treated as octal if they started with 0.
2760 The documentation was vague. These values are now always treated as
2761 decimal, and I will make that clear in the documentation.
2767 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2768 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2771 MH/01 Don't check that the operands of numeric comparison operators are
2772 integers when their expansion is in "skipping" mode (fixes bug
2773 introduced by 4.64-PH/07).
2775 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2776 child addresses, Exim now panics and dies. Previously, because the count
2777 is held in a short int, deliveries were likely to be lost. As such a
2778 large number of recipients for a single message is ridiculous
2779 (performance will be very, very poor), I have chosen to impose a limit
2780 rather than extend the field.
2786 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2787 leftover -K file (the existence of which was triggered by #402).
2788 While we were at it, introduced process PID as part of the -K
2789 filename. This should rule out race conditions when creating
2792 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2793 processing considerably. Previous code took too long for large mails,
2794 triggering a timeout which in turn triggers #401.
2796 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2797 in the DK code in transports.c. sendfile() is not really portable,
2798 hence the _LINUX specificness.
2800 TF/01 In the add_headers option to the mail command in an Exim filter,
2801 there was a bug that Exim would claim a syntax error in any
2802 header after the first one which had an odd number of characters
2805 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2806 callout verification, Exim cached a "reject" for the entire domain. This
2807 is correct for most verifications, but it is not correct for a recipient
2808 verification with use_sender or use_postmaster set, because in that case
2809 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2810 case of MAIL FROM:<> rejection from other early rejections (e.g.
2811 rejection of HELO). When verifying a recipient using a non-null MAIL
2812 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2813 Whatever the result of the callout, the value of the domain cache is
2814 left unchanged (for any other kind of callout, getting as far as trying
2815 RCPT means that the domain itself is ok).
2817 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2820 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2821 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2822 session, the daemon ignores ECONNECT errors and logs others; it now
2823 ignores EPIPE as well.
2825 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2826 (quoted-printable decoding).
2828 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2829 later the small subsequent patch to fix an introduced bug.
2831 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2833 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2835 PH/08 An error is now given if message_size_limit is specified negative.
2837 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2838 to be given (somewhat) arbitrary names.
2840 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2843 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2844 miscellaneous code fixes
2846 PH/10 Added the log_reject_target ACL modifier to specify where to log
2849 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2850 hostname. This is wrong, because it relates to the incoming message (and
2851 probably the interface on which it is arriving) and not to the outgoing
2852 callout (which could be using a different interface). This has been
2853 changed to use the value of the helo_data option from the smtp transport
2854 instead - this is what is used when a message is actually being sent. If
2855 there is no remote transport (possible with a router that sets up host
2856 addresses), $smtp_active_hostname is used.
2858 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2859 tweaks were necessary in order to get it to work (see also 21 below):
2860 (a) The code assumed that strncpy() returns a negative number on buffer
2861 overflow, which isn't the case. Replaced with Exim's string_format()
2863 (b) There were several signed/unsigned issues. I just did the minimum
2864 hacking in of casts. There is scope for a larger refactoring.
2865 (c) The code used strcasecmp() which is not a standard C function.
2866 Replaced with Exim's strcmpic() function.
2867 (d) The code set only $1; it now sets $auth1 as well.
2868 (e) A simple test gave the error "authentication client didn't specify
2869 service in request". It would seem that Dovecot has changed its
2870 interface. Fortunately there's a specification; I followed it and
2871 changed what the client sends and it appears to be working now.
2873 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2876 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2877 address A is aliased to B and C, where B exists and C does not. Without
2878 -v the output is "A verified" because verification stops after a
2879 successful redirection if more than one address is generated. However,
2880 with -v the child addresses are also verified. Exim was outputting "A
2881 failed to verify" and then showing the successful verification for C,
2882 with its parentage. It now outputs "B failed to verify", showing B's
2883 parentage before showing the successful verification of C.
2885 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2886 look up a TXT record in a specific list after matching in a combined
2889 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2890 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2891 they consult the DNS. I had assumed they would set it the way they
2892 wanted; and indeed my experiments on Linux seem to show that in some
2893 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2894 To be on the safe side, however, I have now made the interface to
2895 host_find_byname() similar to host_find_bydns(), with an argument
2896 containing the DNS resolver options. The host_find_byname() function now
2897 sets these options at its start, just as host_find_bydns() does. The smtp
2898 transport options dns_qualify_single and dns_search_parents are passed to
2899 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2900 of host_find_byname() use the default settings of RES_DEFNAMES
2901 (qualify_single) but not RES_DNSRCH (search_parents).
2903 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2904 spool_read_header() do less string testing, by means of a preliminary
2905 switch on the second character of optional "-foo" lines. (This is
2906 overdue, caused by the large number of possibilities that now exist.
2907 Originally there were few.) While I was there, I also converted the
2908 str(n)cmp tests so they don't re-test the leading "-" and the first
2909 character, in the hope this might squeeze out yet more improvement.
2911 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2912 flag allowing group syntax was set by the header_syntax check but not
2913 turned off, possible causing trouble later; (2) The flag was not being
2914 set at all for the header_verify test, causing "group"-style headers to
2915 be rejected. I have now set it in this case, and also caused header_
2916 verify to ignore an empty address taken from a group. While doing this, I
2917 came across some other cases where the code for allowing group syntax
2918 while scanning a header line wasn't quite right (mostly, not resetting
2919 the flag correctly in the right place). These bugs could have caused
2920 trouble for malformed header lines. I hope it is now all correct.
2922 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2923 with the "reply" argument non-NULL. The code, however (which originally
2924 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2925 but it didn't always do it. This confused somebody who was copying the
2926 code for some other use. I have removed all the tests.
2928 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2929 feature that was used to support insecure browsers during the U.S. crypto
2930 embargo. It requires special client support, and Exim is probably the
2931 only MTA that supported it -- and would never use it because real RSA is
2932 always available. This code has been removed, because it had the bad
2933 effect of slowing Exim down by computing (never used) parameters for the
2934 RSA_EXPORT functionality.
2936 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2937 authenticator to fail if there's a tab character in the incoming data
2938 (there should never be unless someone is messing about, as it's supposed
2939 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2940 if the connection is using TLS or if the remote IP is the same as the
2941 local IP, and the "valid-client-cert option" if a client certificate has
2944 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2945 authenticators. This can be used for authorization after authentication
2946 succeeds. (In the case of plaintext, it servers for both authentication
2949 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2950 if any retry times were supplied.
2952 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2953 connection, where there is no client IP address to check. In this
2954 situation, the verify now always succeeds.
2956 PH/25 Applied John Jetmore's -Mset patch.
2958 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2960 PH/27 In a string expansion for a processed (not raw) header when multiple
2961 headers of the same name were present, leading whitespace was being
2962 removed from all of them, but trailing whitespace was being removed only
2963 from the last one. Now trailing whitespace is removed from each header
2964 before concatenation. Completely empty headers in a concatenation (as
2965 before) are ignored.
2967 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2968 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2970 PH/29 [Removed. This was a change that I later backed out, and forgot to
2971 correct the ChangeLog entry (that I had efficiently created) before
2972 committing the later change.]
2974 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2975 address errors (4xx response to RCPT) over the same connection as other
2976 messages routed to the same hosts. Such deliveries are always "forced",
2977 so retry times are not inspected. This resulted in far too many retries
2978 for the affected addresses. The effect occurred only when there were more
2979 hosts than the hosts_max_try setting in the smtp transport when it had
2980 the 4xx errors. Those hosts that it had tried were not added to the list
2981 of hosts for which the message was waiting, so if all were tried, there
2982 was no problem. Two fixes have been applied:
2984 (i) If there are any address or message errors in an SMTP delivery, none
2985 of the hosts (tried or untried) are now added to the list of hosts
2986 for which the message is waiting, so the message should not be a
2987 candidate for sending over the same connection that was used for a
2988 successful delivery of some other message. This seems entirely
2989 reasonable: after all the message is NOT "waiting for some host".
2990 This is so "obvious" that I'm not sure why it wasn't done
2991 previously. Hope I haven't missed anything, but it can't do any
2992 harm, as the worst effect is to miss an optimization.
2994 (ii) If, despite (i), such a delivery is accidentally attempted, the
2995 routing retry time is respected, so at least it doesn't keep
2996 hammering the server.
2998 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2999 in ${readsocket because some servers need this prod.
3001 PH/32 Added some extra debug output when updating a wait-xxx database.
3003 PH/33 The hint "could be header name not terminated by colon", which has been
3004 given for certain expansion errors for a long time, was not being given
3005 for the ${if def:h_colon_omitted{... case.
3007 PH/34 The spec says: "With one important exception, whenever a domain list is
3008 being scanned, $domain contains the subject domain." There was at least
3009 one case where this was not true.
3011 PH/35 The error "getsockname() failed: connection reset by peer" was being
3012 written to the panic log as well as the main log, but it isn't really
3013 panic-worthy as it just means the connection died rather early on. I have
3014 removed the panic log writing for the ECONNRESET error when getsockname()
3017 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3018 runs only) independently of the message's sender address. This meant
3019 that, if the 4xx error was in fact related to the sender, a different
3020 message to the same recipient with a different sender could confuse
3021 things. In particular, this can happen when sending to a greylisting
3022 server, but other circumstances could also provoke similar problems.
3023 I have changed the default so that the retry time for these errors is now
3024 based a combination of the sender and recipient addresses. This change
3025 can be overridden by setting address_retry_include_sender=false in the
3028 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3029 remote server are returned as part of bounce messages. This was not
3030 happening for LMTP over a pipe (the lmtp transport), but now it is the
3031 same for both kinds of LMTP.
3033 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3034 in header lines that were in fact CNAMEs. This is no longer the case.
3036 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3037 and queue runs started by the daemon processed all messages. This has
3038 been fixed so that -R and -S can now usefully be given with -q<time>.
3040 PH/40 Import PCRE release 6.7 (fixes some bugs).
3042 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3044 PH/42 Give an error if -q is specified more than once.
3046 PH/43 Renamed the variables $interface_address and $interface_port as
3047 $received_ip_address and $received_port, to make it clear that these
3048 values apply to message reception, and not to the outgoing interface when
3049 a message is delivered. (The old names remain recognized, of course.)
3051 PH/44 There was no timeout on the connect() call when using a Unix domain
3052 socket in the ${readsocket expansion. There now is.
3054 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3055 be meaningful with "accept".
3057 SC/01 Eximstats V1.43
3058 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3060 SC/02 Eximstats V1.44
3061 Use a glob alias rather than an array ref in the generated
3062 parser. This improves both readability and performance.
3064 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3065 Collect SpamAssassin and rejection statistics.
3066 Don't display local sender or destination tables unless
3067 there is data to show.
3068 Added average volumes into the top table text output.
3070 SC/04 Eximstats V1.46
3071 Collect data on the number of addresses (recipients)
3072 as well as the number of messages.
3074 SC/05 Eximstats V1.47
3075 Added 'Message too big' to the list of mail rejection
3076 reasons (thanks to Marco Gaiarin).
3078 SC/06 Eximstats V1.48
3079 Mainlog lines which have GMT offsets and are too short to
3080 have a flag are now skipped.
3082 SC/07 Eximstats V1.49 (Alain Williams)
3083 Added the -emptyok flag.
3085 SC/08 Eximstats V1.50
3086 Fixes for obtaining the IP address from reject messages.
3088 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3089 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3090 whitespace changes from 4.64-PH/27
3092 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3095 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3096 are found, allow negative numbers in numeric criteria)
3098 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3100 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3101 to match changes made in 4.64-PH/43
3103 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3105 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3106 "domain" is actually an IP address, and force a failure. This locks out
3107 those revolvers/nameservers that support "A-for-A" lookups, in
3108 contravention of the specifications.
3110 PH/48 When a host name was looked up from an IP address, and the subsequent
3111 forward lookup of the name timed out, the host name was left in
3112 $sender_host_name, contrary to the specification.
3114 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3115 restricted to single-key lookups, Exim was not diagnosing an error if
3116 * or *@ was used with a query-style lookup.
3118 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3120 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3121 long ago, but noone interested enough thought of it. Let's just say that
3122 the "1.1" means that there are some new functions that weren't there at
3123 some point in the past.
3125 PH/51 Error processing for expansion failure of helo_data from an smtp
3126 transport during callout processing was broken.
3128 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3129 tested/used via the -bh/-bhc/-bs options.
3131 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3132 bug, fixed in subsequent PCRE releases).
3134 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3135 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3137 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3143 SC/01 Use a glob alias rather than an array ref in eximstats generated
3144 parser. This improves both readability and performance.
3146 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3147 Don't display local sender or destination tables in eximstats unless
3148 there is data to show.
3149 Added average volumes into the eximstats top table text output.
3151 SC/03 Collect data on the number of addresses (recipients) as well
3152 as the number of messages in eximstats.
3154 TF/01 Correct an error in the documentation for the redirect router. Exim
3155 does (usually) call initgroups() when daemonizing.
3157 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3158 with consistent privilege compared to when running as a daemon.
3160 TF/03 Note in the spec that $authenticated_id is not set for local
3161 submissions from trusted users.
3163 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3164 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3166 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3167 by adding some example configuration directives to the default
3168 configuration file. A little bit of work is required to uncomment the
3169 directives and define how usernames and passwords are checked, but
3170 there is now a framework to start from.
3172 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3173 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3174 without this. I don't know how relevant this is to other LDAP libraries.
3176 PH/02 Add the verb name to the "unknown ACL verb" error.
3178 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3180 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3182 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3183 directory not expanded when it should be if an expanded home directory
3184 was set for the address (which is overridden by the transport).
3186 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3189 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3190 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3191 because it is too late at that time, and has no effect.
3193 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3194 security issue with \' (bugzilla #107). I could not use the
3195 PQescapeStringConn() function, because it needs a PGconn value as one of
3198 PH/08 When testing addresses using -bt, indicate those final addresses that
3199 are duplicates that would not cause an additional delivery. At least one
3200 person was confused, thinking that -bt output corresponded to deliveries.
3201 (Suppressing duplicates isn't a good idea as you lose the information
3202 about possibly different redirections that led to the duplicates.)
3204 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3205 systems where poll() doesn't work, in particular OS X.
3207 PH/10 Added more information to debugging output for retry time not reached.
3209 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3210 operations in malware.c.
3212 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3215 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3216 syslog_duplication set false, log lines that would normally be written
3217 both the the main log and to the reject log were not written to syslog at
3220 PH/14 In the default configuration, change the use of "message" in ACL warn
3221 statements to "add_header".
3223 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3224 not followed by a command (e.g. "seen endif").
3226 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3227 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3230 PH/17 Added extra conditions to the default value of delay_warning_condition
3234 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3235 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3236 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3239 The Auto-Submitted: and various List- headers are standardised, whereas I
3240 don't think Precedence: ever was.
3242 PH/18 Refactored debugging code in route_finduser() to show more information,
3243 in particular, the error code if getpwnam() issues one.
3245 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3246 This is apparently needed in addition to the PH/07 change above to avoid
3247 any possible encoding problems.
3249 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3250 but not after initializing Perl.
3252 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3253 output them only if debugging. By default they are written stderr,
3254 apparently, which is not desirable.
3256 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3259 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3262 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3264 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3265 authenticated or an ident call has been made. Suppress the default
3266 values for $authenticated_id and $authenticated_sender (but permit -oMai
3267 and -oMas) when testing with -bh.
3269 PH/24 Re-jigged the order of the tests in the default configuration so that the
3270 tests for valid domains and recipients precede the DNS black list and CSA
3271 tests, on the grounds that those ones are more expensive.
3273 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3274 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3275 This bug exists in every version of Exim that I still have, right back to
3278 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3279 However, an attempt to turn on case-sensitivity in a regex key by
3280 including (?-i) didn't work because the subject string was already
3281 lowercased, and the effects were non-intuitive. It turns out that a
3282 one-line patch can be used to allow (?-i) to work as expected.
3288 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3289 other effects) broke the use of negated acl sub-conditions.
3291 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3294 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3295 "Deny" causes Exim to reject the incoming connection with a 554 error.
3296 Unfortunately, if there is a major crisis, such as a disk failure,
3297 tcp-wrappers gives "deny", whereas what one would like would be some
3298 kind of temporary error. A kludge has been added to help with this.
3299 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3300 554 error is used if errno is still zero or contains ENOENT (which occurs
3301 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3304 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3306 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3307 errors. Otherwise a message that provokes a temporary error (when other
3308 messages do not) can cause a whole host to time out.
3310 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3311 addresses were routed directly to files or pipes from a redirect router.
3312 File deliveries just didn't batch; pipe deliveries might have suffered
3315 PH/06 A failure to get a lock for a hints database would erroneously always say
3316 "Failed to get write lock", even when it was really a read lock.
3318 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3319 of 0600. This has been changed to use the value of the lockfile_mode
3320 option (which defaults to 0600).
3322 PH/08 Applied small patch from the Sieve maintainer.
3324 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3325 folder from quota calculations, a direct delivery into this folder messed
3326 up the contents of the maildirsize file. This was because the regex was
3327 used only to exclude .Trash (or whatever) when the size of the mailbox
3328 was calculated. There was no check that a delivery was happening into an
3329 excluded directory. This bug has been fixed by ignoring all quota
3330 processing for deliveries into excluded directories.
3332 PH/10 Added the maildirfolder_create_regex option to appendfile.
3338 PH/01 The code for finding all the local interface addresses on a FreeBSD
3339 system running IPv6 was broken. This may well have applied to all BSD
3340 systems, as well as to others that have similar system calls. The broken
3341 code found IPv4 interfaces correctly, but gave incorrect values for the
3342 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3343 that it would not match correctly against @[] and not recognize the IPv6
3346 PH/02 The ipliteral router was not recognizing addresses of the form user@
3347 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3349 PH/03 Added disable_ipv6.
3351 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3352 decoded form, because it is most often used to construct To: headers
3353 lines in autoreplies, and the decoded form may well be syntactically
3354 invalid. However, $reply_address has leading white space removed, and all
3355 newlines turned into spaces so that the autoreply transport does not
3358 PH/05 If group was specified without a user on a router, and no group or user
3359 was specified on a transport, the group from the router was ignored.
3361 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3362 for visible compile-time settings that can be used to change these
3363 numbers, for those that want even more. Backwards compatibility with old
3364 spool files has been maintained. However, going back to a previous Exim
3365 release will lost any variables that are in spool files.
3367 PH/07 Two small changes when running in the test harness: increase delay when
3368 passing a TCP/IP connection to a new process, in case the original
3369 process has to generate a bounce, and remove special handling of
3370 127.0.0.2 (sic), which is no longer necessary.
3372 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3373 be the same on different OS.
3375 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3378 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3379 whether --show-vars was specified or not
3381 JJ/02 exipick: Added support for new ACL variable spool format introduced
3384 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3385 syntactically invalid From: or Reply-to: line, and a filter used this to
3386 generate an autoreply, and therefore failed to obtain an address for the
3387 autoreply, Exim could try to deliver to a non-existent relative file
3388 name, causing unrelated and misleading errors. What now happens is that
3389 it logs this as a hard delivery error, but does not attempt to create a
3392 PH/11 The exinext utility has a -C option for testing purposes, but although
3393 the given file was scanned by exinext itself; it wasn't being passed on
3394 when Exim was called.
3396 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3397 an end-of-file indication when reading a command response.
3399 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3400 compiled. In many other places in Exim, IPv6 addresses are always
3401 recognized, so I have changed this. It also means that IPv4 domain
3402 literals of the form [IPV4:n.n.n.n] are now always recognized.
3404 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3405 used if the router is not running as root, for example, when verifying at
3406 ACL time, or when using -bh. The debugging output from this situation was
3407 non-existent - all you got was a failure to exec. I have made two
3410 (a) Failures to set uid/gid, the current directory, or a process leader
3411 in a subprocess such as that created by queryprogram now generate
3412 suitable debugging output when -d is set.
3414 (b) The queryprogram router detects when it is not running as root,
3415 outputs suitable debugging information if -d is set, and then runs
3416 the subprocess without attempting to change uid/gid.
3418 PH/15 Minor change to Makefile for building test_host (undocumented testing
3421 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3422 additional section of a DNS packet that returns MX or SRV records.
3423 Instead, it always explicitly searches for A/AAAA records. This avoids
3424 major problems that occur when a DNS server includes only records of one
3425 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3426 fixed another bug: if SRV records were looked up and the corresponding
3427 address records were *not* found in the additional section, the port
3428 values from the SRV records were lost.
3430 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3431 using the correct key (the original address) when searching the retry
3432 rules in order to find which one to use for generating the retry hint.
3434 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3435 adding the default one. Similarly, if it contains a Reply-To: header, the
3436 errors_reply_to option, if set, is not used.
3438 PH/19 When calculating a retry time, Exim used to measure the "time since
3439 failure" by looking at the "first failed" field in the retry record. Now
3440 it does not use this if it is later than than the arrival time of the
3441 message. Instead it uses the arrival time. This makes for better
3442 behaviour in cases where some deliveries succeed, thus re-setting the
3443 "first failed" field. An example is a quota failure for a huge message
3444 when small messages continue to be delivered. Without this change, the
3445 "time since failure" will always be short, possible causing more frequent
3446 delivery attempts for the huge message than are intended.
3447 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3449 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3450 $1, $2, $3) because the numerical variables can be reset during some
3451 expansion items (e.g. "match"), thereby losing the authentication data.
3453 PH/21 Make -bV show the size of off_t variables so that the test suite can
3454 decide whether to run tests for quotas > 2G.
3456 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3457 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3458 filecount value is greater than 2G or if a quota value is greater than 2G
3459 on a system where the size of off_t is not greater than 4, a panic error
3462 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3463 never match. The debug and -bh output now contains an explicit error
3464 message indicating a malformed IPv4 address or mask.
3466 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3467 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3468 PH/23 above applies.
3470 PH/25 Do not write to syslog when running in the test harness. The only
3471 occasion when this arises is a failure to open the main or panic logs
3472 (for which there is an explicit test).
3474 PH/26 Added the /no_tell option to "control=freeze".
3476 PH/27 If a host name lookup failed very early in a connection, for example, if
3477 the IP address matched host_lookup and the reverse lookup yielded a name
3478 that did not have a forward lookup, an error message of the form "no IP
3479 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3480 could be logged. Now it outputs the IP address instead of "NULL".
3482 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3483 allows the sender and the authenticated sender to be set when
3484 submitting a message from within Exim. Since child_open_exim() is
3485 documented for local_scan(), the new function should be too.
3487 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3488 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3489 results in an empty string is now treated as unset.
3491 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3493 PH/31 Added sender_verify_fail logging option.
3495 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3496 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3497 tidied the source and removed it altogether.
3499 PH/33 When a queue run was abandoned because the load average was too high, a
3500 log line was always written; now it is written only if the queue_run log
3501 selector is set. In addition, the log line for abandonment now contains
3502 information about the queue run such as the pid. This is always present
3503 in "start" and "stop" lines but was omitted from the "abandon" line.
3505 PH/34 Omit spaces between a header name and the colon in the error message that
3506 is given when verify = headers_syntax fails (if there are lots of them,
3507 the message gets confusing).
3509 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3510 names, as there are now some PTR records that contain slashes. This check
3511 is only to protect against broken name servers that fall over on strange
3512 characters, so the fact that it applies to all lookups doesn't matter.
3514 PH/36 Now that the new test suite is complete, we can remove some of the
3515 special code in Exim that was needed for the old test suite. For example,
3516 sorting DNS records because real resolvers return them in an arbitrary
3517 order. The new test suite's fake resolver always returns records in the
3520 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3521 bounces) except when queue_only is set, to avoid logging races between
3522 the different processes.
3524 PH/38 Panic-die if .include specifies a non-absolute path.
3526 PH/39 A tweak to the "H" retry rule from its user.
3528 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3529 a label. They prevented compilation on older perls.
3531 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3532 a warning to be raised on newish perls.
3534 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3535 on queue. Changes to match documented behaviour of showing count of
3536 messages matching specified criteria.
3538 PH/40 Changed the default ident timeout from 30s to 5s.
3540 PH/41 Added support for the use of login_cap features, on those BSD systems
3541 that have them, for controlling the resources used by pipe deliveries.
3543 PH/42 The content-scanning code uses fopen() to create files in which to put
3544 message data. Previously it was not paying any attention to the mode of
3545 the files. Exim runs with umask(0) because the rest of the code creates
3546 files with open(), and sets the required mode explicitly. Thus, these
3547 files were ending up world-writeable. This was not a big issue, because,
3548 being within the spool directory, they were not world-accessible. I have
3549 created a function called modefopen, which takes an additional mode
3550 argument. It sets umask(777), creates the file, chmods it to the required
3551 mode, then resets the umask. All the relevant calls to fopen() in the
3552 content scanning code have been changed to use this function.
3554 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3555 to 24 hours. This avoids potential overflow problems when processing G
3556 and H retry rules. I suspect nobody ever tinkers with this value.
3558 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3560 PH/45 When the plaintext authenticator is running as a client, the server's
3561 challenges are checked to ensure they are valid base64 strings. By
3562 default, the authentication attempt is cancelled if an invalid string is
3563 received. Setting client_ignore_invalid_base64 true ignores these errors.
3564 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3565 they are received. Thus, the responses can be made to depend on the
3566 challenges. If an invalid string is ignored, an empty string is placed in
3569 PH/46 Messages that are created by the autoreply transport now contains a
3570 References: header, in accordance with RFCs 2822 and 3834.
3572 PH/47 Added authenticated_sender_force to the smtp transport.
3574 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3576 PH/49 Installed latest patch from the Sieve maintainer.
3578 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3579 was also set, the appendfile transport was unnecessarily scanning a
3580 directory of message files (e.g. for maildir delivery) to find the count
3581 of files (along with the size), even though it did not need this
3582 information. It now does the scan only if it needs to find either the
3583 size of the count of files.
3585 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3587 PH/52 Two bugs concerned with error handling when the smtp transport is
3590 (i) Exim was not creating retry information for temporary errors given
3591 for individual recipients after the DATA command when the smtp transport
3592 was used in LMTP mode. This meant that they could be retried too
3593 frequently, and not timed out correctly.
3595 (ii) Exim was setting the flag that allows error details to be returned
3596 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3597 recipients that were returned after the DATA command.
3599 PH/53 This is related to PH/52, but is more general: for any failing address,
3600 when detailed error information was permitted to be returned to the
3601 sender, but the error was temporary, then after the final timeout, only
3602 "retry timeout exceeded" was returned. Now it returns the full error as
3603 well as "retry timeout exceeded".
3605 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3606 do this, and (what is worse) MTAs that accept it.
3608 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3609 will now be deprecated.
3611 PH/56 New os.c-cygwin from the Cygwin maintainer.
3613 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3614 formats (previously only available in exim formats via -bpr, -bpru,
3615 and -bpra. Now also available in native and exiqgrep formats)
3617 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3618 with very large, slow to parse queues
3620 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3622 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3624 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3625 responses to authentication challenges, though it was showing the
3626 challenges; (ii) I've removed the CR characters from the debug output for
3629 PH/58 Allow for the insertion of a newline as well as a space when a string
3630 is turned into more than one encoded-word during RFC 2047 encoding. The
3631 Sieve code now uses this.
3633 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3634 data_4xx, lost_connection, tls_required.
3636 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3637 message was being sent as an SMTP response.
3639 PH/61 Add -l and -k options to exicyclog.
3641 PH/62 When verifying, if an address was redirected to one new address, so that
3642 verification continued, and the new address failed or deferred after
3643 having set something in $address_data, the value of $address_data was not
3644 passed back to the ACL. This was different to the case when no
3645 redirection occurred. The value is now passed back in both cases.
3647 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3648 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3649 use login_cap.h, so on its own it isn't the distinguishing feature. The
3650 new name refers directly to the setclassresources() function.
3652 PH/65 Added configuration files for NetBSD3.
3654 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3656 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3657 is preferred over IPv4.
3659 PH/68 The bounce_return_message and bounce_return_body options were not being
3660 honoured for bounces generated during the reception of non-SMTP messages.
3661 In particular, this applied to messages rejected by the ACL. This bug has
3662 been fixed. However, if bounce_return_message is true and bounce_return_
3663 body is false, the headers that are returned for a non-SMTP message
3664 include only those that have been read before the error was detected.
3665 (In the case of an ACL rejection, they have all been read.)
3667 PH/69 The HTML version of the specification is now built in a directory called
3668 spec_html instead of spec.html, because the latter looks like a path with
3669 a MIME-type, and this confuses some software.
3671 PH/70 Catch two compiler warnings in sieve.c.
3673 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3674 function verify_get_ident() calls ip_connect() to connect a socket, but
3675 if the "connect()" function timed out, ip_connect() used to close the
3676 socket. However, verify_get_ident() also closes the socket later, and in
3677 between Exim writes to the log, which may get opened at this point. When
3678 the socket was closed in ip_connect(), the log could get the same file
3679 descriptor number as the socket. This naturally causes chaos. The fix is
3680 not to close the socket in ip_connect(); the socket should be closed by
3681 the function that creates it. There was only one place in the code where
3682 this was missing, in the iplookup router, which I don't think anybody now
3683 uses, but I've fixed it anyway.
3685 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3686 well as to direct DNS lookups. Otherwise the handling of names in host
3687 lists is inconsistent and therefore confusing.
3693 PH/01 Two changes to the default runtime configuration:
3695 (1) Move the checks for relay_from_hosts and authenticated clients from
3696 after to before the (commented out) DNS black list checks.
3698 (2) Add control=submission to the relay_from_hosts and authenticated
3699 clients checks, on the grounds that messages accepted by these
3700 statements are most likely to be submissions.
3702 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3704 (1) Generate an error if the third argument for the ${prvs expansion is
3707 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3710 (3) Reset the variables that are obtained from the first argument of
3711 ${prvscheck and used in the second argument before leaving the code,
3712 because their memory is reclaimed, so using them afterwards may do
3715 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3716 one (it's much easier than Tom thought :-).
3718 (5) Because of (4), we can now allow for the use of $prvscheck_result
3719 inside the third argument.
3721 PH/03 For some reason, the default setting of PATH when running a command from
3722 a pipe transport was just "/usr/bin". I have changed it to
3725 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3726 anything to be listed in the output from -bV.
3728 PH/05 When a filter generated an autoreply, the entire To: header line was
3729 quoted in the delivery log line, like this:
3731 => >A.N.Other <ano@some.domain> <original@ddress> ...
3733 This has been changed so that it extracts the operative address. There
3734 may be more than one such address. If so, they are comma-separated, like
3737 => >ano@some.domain,ona@other.domain <original@ddress> ...
3739 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3740 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3741 not being looked up in the rDNS to get a host name, Exim was showing the
3742 IP address twice in Received: lines, even though the IP addresses were
3743 identical. For example:
3745 Received: from [1.2.3.4] (helo=[1.2.3.4])
3747 However, if the real host name was known, it was omitting the HELO data
3748 if it matched the actual IP address. This has been tidied up so that it
3749 doesn't show the same IP address twice.
3751 PH/07 When both +timestamp and +memory debugging was on, the value given by
3752 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3753 called by the debug printing, thereby overwriting the timestamp buffer.
3754 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3756 PH/08 When the original message was included in an autoreply transport, it
3757 always said "this is a copy of the message, including all the headers",
3758 even if body_only or headers_only was set. It now gives an appropriate
3761 PH/09 Applied a patch from the Sieve maintainer which:
3763 o fixes some comments
3764 o adds the (disabled) notify extension core
3765 o adds some debug output for the result of if/elsif tests
3766 o points to the current vacation draft in the documentation
3767 and documents the missing references header update
3771 o fixes a bug in processing the envelope test (when testing
3772 multiple envelope elements, the last element determined the
3775 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3776 Electronic Mail") by including:
3778 Auto-submitted: auto-generated
3780 in the messages that it generates (bounce messages and others, such as
3781 warnings). In the case of bounce messages for non-SMTP messages, there was
3782 also a typo: it was using "Auto_submitted" (underscore instead of
3783 hyphen). Since every message generated by Exim is necessarily in response
3784 to another message, thes have all been changed to:
3786 Auto-Submitted: auto-replied
3788 in accordance with these statements in the RFC:
3790 The auto-replied keyword:
3792 - SHOULD be used on messages sent in direct response to another
3793 message by an automatic process,
3795 - MUST NOT be used on manually-generated messages,
3797 - MAY be used on Delivery Status Notifications (DSNs) and Message
3798 Disposition Notifications (MDNs),
3800 - MUST NOT be used on messages generated by automatic or periodic
3801 processes, except for messages which are automatic responses to
3804 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3805 to the default Received: header definition.
3807 PH/12 Added log selector acl_warn_skipped (default on).
3809 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3810 variables because (a) they are in the wrong storage pool and (b) even if
3811 they were copied, it wouldn't work properly because of the caching.
3813 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3814 checking when decoding. Apparently there are clients that generate
3815 overlong encoded strings. Why am I not surprised?
3817 PH/15 If the first argument of "${if match_address" was not empty, but did not
3818 contain an "@" character, Exim crashed. Now it writes a panic log message
3819 and treats the condition as false.
3821 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3823 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3824 extension "envelope-auth". The code is finished and in agreement with
3825 other implementations, but there is no documentation so far and in fact,
3826 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3827 not changing the active code.
3829 Print executed "if" and "elsif" statements when debugging is used. This
3830 helps a great deal to understand what a filter does.
3832 Document more things not specified clearly in RFC3028. I had all this
3833 sorted out, when out of a sudden new issues came to my mind. Oops."
3835 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3838 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3839 canonical form (as documented). However, after a host name lookup from
3840 the IP address, check_host() was doing a simple string comparison with
3841 addresses acquired from the DNS when checking that the found name did
3842 have the original IP as one of its addresses. Since any found IPv6
3843 addresses are likely to be in abbreviated form, the comparison could
3844 fail. Luckily, there already exists a function for doing the comparison
3845 by converting both addresses to binary, so now that is used instead of
3846 the text comparison.
3848 PH/20 There was another similar case to PH/19, when a complete host name was
3849 given in a host list; looking up its IP address could give an abbreviated
3850 form, whereas the current host's name might or might not be abbreviated.
3851 The same fix has been applied.
3857 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3858 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3861 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3862 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3864 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3866 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3867 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3868 Also, for "Auto-Submitted", treat anything other than "no" as
3869 non-personal, in accordance with RFC 3834. (Previously it treated
3870 anything starting "auto-" as non-personal.)
3872 TF/01 The control=submission/name=... option had a problem with syntax
3873 errors if the name included a slash character. The /name= option
3874 now slurps the rest of the string, so it can include any characters
3875 but it must come last in the list of options (after /sender_retain
3878 PH/05 Some modifications to the interface to the fake nameserver for the new
3886 TK/01 Added the "success_on_redirect" address verification option. See
3887 NewStuff for rationale and an example.
3889 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3891 PH/02 Patch to exigrep to allow it to work on syslog lines.
3893 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3894 fread() to skip over the body file's header line, because in Cygwin the
3895 header line is locked and is inaccessible.
3897 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3898 co-exist for some time) to make it clear that it is the Exim ID that is
3899 referenced, not the Message-ID: header line.
3901 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3902 string_format() function, because snprintf() does not exist on all
3905 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3906 ${stat: expansion item.
3908 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3909 protocol synchronization error", to keep the pedants happy.
3911 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3912 well as for IRIX systems, when gcc is being used. See the host.c source
3915 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3917 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3920 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3921 transport and to the smtp transport in LMTP mode.
3923 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3925 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3926 run a filter in a subprocess. This could lead to confusion in subsequent
3927 lookups in the parent process. There should also be a search_tidyup() at
3928 the end of the subprocess.
3930 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3931 only if the host matched helo_try_verify_hosts, which caused the
3932 verification to occur when the EHLO/HELO command was issued. The ACL just
3933 tested the remembered result. Now, if a previous verification attempt has
3934 not happened, "verify = helo" does it there and then.
3936 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3938 TK/03 Fix log output including CR from clamd.
3940 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3941 exist provoked a memory error which could cause a segfault.
3943 PH/15 Installed PCRE 6.2
3945 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3947 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3948 of the problem. Specifically, suggested +O2 rather than +O1 for the
3951 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3953 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3954 record, the port was not correctly logged when the outgoing_port log
3955 selector was set (it logged the transort's default port).
3957 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3958 fallback_hosts, and "hosts" in the smtp transport.
3960 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3961 host errors such as "Connection refused".
3963 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3964 authentication with radiusclient 0.4.9:
3966 - Error returned from rc_read_config was caught wrongly
3967 - Username/password not passed on to radius server due to wrong length.
3969 The presumption is that some radiusclient API changes for 4.51/PH/17
3970 were not taken care of correctly. The code is still untested by me (my
3971 Linux distribution still has 0.3.2 of radiusclient), but it was
3972 contributed by a Radius user.
3974 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3975 expanding the "port" option of the smtp transport.
3977 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3978 while reading a MIME header. Thanks to Tom Hughes for a patch.
3980 PH/24 Include config.h inside local_scan.h so that configuration settings are
3983 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3984 that in an ACL for RCPT (for example), you can examine exactly what was
3987 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3988 commands, but it was not correctly comparing the address with the actual
3989 client host address. Thus, it would show the EHLO address in Received:
3990 header lines when this was not necessary.
3992 PH/27 Added the % operator to ${eval:}.
3994 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3995 it should be ignoring failures (because with -C, for example, it has lost
3996 privilege). It wasn't ignoring creation failures other than "already
3999 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4002 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4003 because an input line was too long, either on its own, or by virtue of
4004 too many continuations, the temporary file was not being removed, and the
4005 return code was incorrect.
4007 PH/31 Missing "BOOL" in function definition in filtertest.c.
4009 PH/32 Applied Sieve patches from the maintainer.
4011 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4013 PH/33 Added "verify = not_blind".
4015 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4016 Local/Makefile (with some defaults set). These are used in built scripts
4017 such as exicyclog, but they have never been used in the exim_install
4018 script (though there are many overriding facilities there). I have
4019 arranged that the exim_install script now takes note of these two
4022 PH/35 Installed configuration files for Dragonfly.
4024 PH/36 When a locally submitted message by a trusted user did not contain a
4025 From: header, and the sender address was obtained from -f or from an SMTP
4026 MAIL command, and the trusted user did not use -F to supply a sender
4027 name, $originator_name was incorrectly used when constructing a From:
4028 header. Furthermore, $originator_name was used for submission mode
4029 messages from external hosts without From: headers in a similar way,
4030 which is clearly wrong.
4032 PH/37 Added control=suppress_local_fixups.
4034 PH/38 When log_selector = +received_sender was set, and the addition of the
4035 sender made the log line's construction buffer exactly full, or one byte
4036 less than full, an overflow happened when the terminating "\n" was
4039 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4040 when the result of a list match is failure because a DNS lookup failed.
4042 PH/40 RM_COMMAND is now used in the building process.
4044 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4045 the "build-* directories that it finds.
4047 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4048 address in a domain literal was a prefix of an interface address.
4050 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4051 when verifying a sender address, unless rewrite_headers is false.
4053 PH/44 Wrote a long comment about why errors_to addresses are verified as
4054 recipients, not senders.
4056 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4057 the ratelimit ACL was added.
4059 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4061 PH/46 Added extra information about PostgreSQL errors to the error string.
4063 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4064 avoiding the need to install special zones in a real server. This is
4065 backwards compatible; if it can't find the fake resolver, it drops back.
4066 Thus, both old and new test suites can be run.
4068 TF/02 Added util/ratelimit.pl
4070 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4071 clock is set back in time.
4073 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4074 Candler <B.Candler@pobox.com>.
4076 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4077 OK for addresses that are the result of redirections.
4079 PH/48 A number of further additions for the benefit of the new test suite,
4080 including a fake gethostbyname() that interfaces to the fake DNS resolver
4083 TF/06 The fix for widen_domains has also been applied to qualify_single and
4084 search_parents which are the other dnslookup options that can cause
4087 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4090 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4092 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4093 service" if it encountered an MX record with an empty target hostname.
4094 The message is now "an MX or SRV record indicated no SMTP service".
4096 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4097 if the DNS of the sending site is misconfigured. This is quite a
4098 common situation. This change restores the behaviour of treating a
4099 helo verification defer as a failure.
4101 PH/51 If self=fail was set on a router, the bounce message did not include the
4102 actual error message.
4108 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4110 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4111 command itself ended in error, the underlying message about the transport
4112 filter timeout was being overwritten with the pipe command error. Now the
4113 underlying error message should be appended to the second error message.
4115 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4117 PH/02 Reduce the amount of output that "make" produces by default. Full output
4118 can still be requested.
4120 PH/03 The warning log line about a condition test deferring for a "warn" verb
4121 was being output only once per connection, rather than after each
4122 occurrence (because it was using the same function as for successful
4123 "warn" verbs). This seems wrong, so I have changed it.
4125 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4126 it should not have, which might have caused a crash in the right
4127 circumstances, but probably never did.
4129 PH/04 Installed a modified version of Tony Finch's patch to make submission
4130 mode fix the return path as well as the Sender: header line, and to
4131 add a /name= option so that you can make the user's friendly name appear
4134 TF/03 Added the control = fakedefer ACL modifier.
4136 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4137 Mark Lowes for thorough testing.
4139 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4141 TK/03 Merged latest SRS patch from Miles Wilton.
4143 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4144 with the definition in sysexits.h (which is #included earlier).
4145 Fortunately, Exim does not actually use EX_OK. The code used to try to
4146 preserve the sysexits.h value, by assuming that macro definitions were
4147 scanned for macro replacements. I have been disabused of this notion,
4148 so now the code just undefines EX_OK before #including unistd.h.
4150 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4151 in the smtp transport. When a block could not be written in a single
4152 write() function, the timeout was being re-applied to each part-write.
4153 This seems wrong - if the receiver was accepting one byte at a time it
4154 would take for ever. The timeout is now adjusted when this happens. It
4155 doesn't have to be particularly precise.
4157 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4158 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4160 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4161 without a domain if the check to <postmaster@domain> fails.
4163 SC/01 Eximstats: added -xls and the ability to specify output files
4164 (patch written by Frank Heydlauf).
4166 SC/02 Eximstats: use FileHandles for outputting results.
4168 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4170 SC/04 Eximstats: fixed display of large numbers with -nvr option
4172 SC/05 Eximstats: fixed merging of reports with empty tables.
4174 SC/06 Eximstats: added the -include_original_destination flag
4176 SC/07 Eximstats: removed tabs and trailing whitespace.
4178 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4180 TK/06 MBOX spool code: Add real "From " MBOX separator line
4181 so the .eml file is really in mbox format (even though
4182 most programs do not really care). Patch from Alex Miller.
4184 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4185 The latter is generated from $received_to and is only set if the
4186 message has one envelope recipient. SA can use these headers,
4187 obviously out-of-the-box. Patch from Alex Miller.
4189 PH/08 The ${def test on a variable was returning false if the variable's
4190 value was "0", contrary to what the specification has always said!
4191 The result should be true unless the variable is empty.
4193 PH/09 The syntax error of a character other than { following "${if
4194 def:variable_name" (after optional whitespace) was not being diagnosed.
4195 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4196 accidental colon was present, for example, could give incorrect results.
4198 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4199 result is used (not including appendfile, where other changes are about
4202 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4203 This involved changing a lot of size variables from int to off_t. It
4204 should work with maildirs and everything.
4206 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4207 spamd dying while we are connected to it.
4209 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4212 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4213 function for BDB 4.3.
4215 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4217 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4218 the sources for PCRE have been re-arranged and more files are now
4221 PH/15 The code I had for printing potentially long long variables in PH/11
4222 above was not the best (it lost precision). The length of off_t variables
4223 is now inspected at build time, and an appropriate printing format (%ld
4224 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4225 to be "long long int" or "long int". This is needed for the internal
4226 formatting function string_vformat().
4228 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4229 the configuration file to be ":syslog", then the script "guesses" where
4230 the logs files are, rather than using the compiled in default. In our
4231 case the guess is not the same as the compiled default, so the script
4232 suddenly stopped working when I started to use syslog. The patch checks
4233 to see if log_file_path is "". If so, it attempts to read it from exim
4234 with no configuration file to get the compiled in version, before it
4235 falls back to the previous guessing code."
4237 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4238 implementing BATV in an Exim configuration. See NewStuff for the gory
4241 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4242 Makefile that are specific to HP-UX.
4244 PH/18 If the "use_postmaster" option was set for a recipient callout together
4245 with the "random" option, the postmaster address was used as the MAIL
4246 FROM address for the random test, but not for the subsequent recipient
4247 test. It is now used for both.
4249 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4250 patch removes a few documentation additions to RFC 3028, because the
4251 latest draft now contains them. It adds the new en;ascii-case comparator
4252 and a new error check for 8bit text in MIME parts. Comparator and
4253 require names are now matched exactly. I enabled the subaddress
4254 extension, but it is not well tested yet (read: it works for me)."
4256 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4257 rework some of the code of TK/09 above to avoid the hardwired use of
4258 "%lld" and "long long". Replaced the call to snprintf() with a call to
4261 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4262 records point to non-existent hosts", "retry timeout exceeded", and
4263 "retry time not reached for any host after a long failure period".
4265 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4266 experimental DomainKeys support:
4268 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4269 (2) On an error such as an illegally used "control", the wrong name for
4270 the control was given.
4272 These problems did NOT occur unless DomainKeys support was compiled.
4274 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4276 PH/24 Added ${if match_ip condition.
4278 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4279 databases so that it will be absolutely obvious if a crash occurs in the
4280 DB library. This is a regular occurrence (often caused by mis-matched
4283 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4284 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4285 picked up on a user's system that detects such things. There doesn't seem
4286 to be a gcc warning option for this - only an attribute that has to be
4287 put on the function's prototype. It seems that in Fedora Core 4 they have
4288 set this on a number of new functions. No doubt there will be more in due
4291 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4292 specify a transport. However, if an address that was verified by such a
4293 router was the subject of a callout, Exim crashed because it tried to
4294 read the rcpt_include_affixes from the non-existent transport. Now it
4295 just assumes that the setting of that option is false. This bug was
4296 introduced by 4.51/PH/31.
4298 PH/28 Changed -d+all to exclude +memory, because that information is very
4299 rarely of interest, but it makes the output a lot bigger. People tend to
4300 do -d+all out of habit.
4302 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4303 code in os-type was giving problems when libc.so lives in lib64, like on
4306 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4307 aren't the modern standard, and it seems that some systems' include files
4308 don't always have them. Exim was already checking for some of the newer
4309 ones like T_AAAA, and defining it itself. I've added checks for all the
4310 record types that Exim uses.
4312 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4313 not automatically generating a new one, as it is supposed to. This
4314 prevented TLS from working. If the file did exist, but contained invalid
4315 data, a new version was generated, as expected. It was only the case of a
4316 non-existent file that was broken.
4318 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4319 with a change in libdomainkeys > 0.64.
4321 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4322 from DNS. If the selector record carries the flag, it now has
4323 precedence over the domain-wide flag.
4325 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4327 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4328 the use of a port name that isn't defined in /etc/services) occurred, the
4329 message was deferred as in a normal delivery, and thus remained on the
4330 spool, instead of being failed because of the mua_wrapper setting. This
4331 is now fixed, and I tidied up some of the mua_wrapper messages at the
4334 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4335 the messages in a hash of arrays rather than using individual hashes.
4336 This is a bit cleaner and results in dramatic memory savings, albeit
4337 at a slight CPU cost.
4339 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4340 as requested by Marc Sherman.
4342 SC/10 Eximstats: added histograms for user specified patterns as requested
4345 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4347 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4348 fopen() in the content-scanning modules that did not already have it.
4354 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4355 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4357 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4359 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4361 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4362 to test Sieve filters that use "vacation".
4364 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4365 that changes the way the GnuTLS parameters are stored in the cache file.
4366 The new format can be generated externally. For backward compatibility,
4367 if the data in the cache doesn't make sense, Exim assumes it has read an
4368 old-format file, and it generates new data and writes a new file. This
4369 means that you can't go back to an older release without removing the
4372 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4373 work if there are any delivery delays because "one_time" forces the
4374 parent to be marked "delivered", so its unseen clone is never tried
4375 again. For this reason, Exim now forbids the simultaneous setting of
4378 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4379 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4380 entry if you want to know the details. The fix, however, neglected to
4381 consider the case when local delivery batching is involved. The test for
4382 "previously delivered" was not happening when checking to see if an
4383 address could be batched with a previous (undelivered) one; under
4384 certain circumstances this could lead to multiple deliveries to the same
4387 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4388 in its include files, and this causes problems building Exim.
4390 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4391 header_syntax) but Exim was just ignoring anything given after a slash.
4392 In particular, this caused confusion with an attempt to use "verify =
4393 reverse_host_lookup/defer_ok". An error is now given when options are
4394 supplied for verify items that do not have them. (Maybe reverse_host_
4395 lookup should have a defer_ok option, but that's a different point.)
4397 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4398 defined by RFC 821) to 2048, because there were problems with some AUTH
4399 commands, and RFC 1869 says the size should be increased for extended
4400 SMTP commands that take arguments.
4402 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4405 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4406 "unknown" error; now it says that the functionality isn't in the binary.
4408 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4409 an address' error message when a string expansion fails (syntax or
4410 whatever). Otherwise the password may appear in the log. Following change
4411 PH/42 below, there is no longer a chance of it appearing in a bounce
4414 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4416 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4417 of its addresses was ever tried. (Bugzilla bug #2.)
4419 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4420 the result incorrectly in the debug output. (It correctly added a newline
4421 to what was transported.)
4423 TF/01 Added $received_time.
4425 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4426 commented out examples of how to interface to a virus scanner and to
4427 SpamAssassin. Also added commented examples of av_scanner and
4428 spamd_address settings.
4430 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4431 and controls are allowed in which ACLs. There were a couple of minor
4432 errors. Some of the entries in the conditions table (which is a table of
4433 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4434 negation of where the condition IS allowed.
4436 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4438 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4439 header file does not have a version number, so I've had to invent a new
4440 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4441 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4442 radiusclient), but it was contributed by a Radius user.
4444 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4445 files or directories, for OpenSSL.
4447 PH/19 When an Exim process that is running as root has to create an Exim log
4448 file, it does so in a subprocess that runs as exim:exim so as to get the
4449 ownership right at creation (otherwise, other Exim processes might see
4450 the file with the wrong ownership). There was no test for failure of this
4451 fork() call, which would lead to the process getting stuck as it waited
4452 for a non-existent subprocess. Forks do occasionally fail when resources
4453 run out. I reviewed all the other calls to fork(); they all seem to check
4456 PH/20 When checking for unexpected SMTP input at connect time (before writing
4457 the banner), Exim was not dealing correctly with a non-positive return
4458 from the read() function. If the client had disconnected by this time,
4459 the result was a log entry for a synchronization error with an empty
4460 string after "input=" when read() returned zero. If read() returned -1
4461 (an event I could not check), uninitialized data bytes were printed.
4462 There were reports of junk text (parts of files, etc) appearing after
4465 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4467 PH/22 Added support for macro redefinition, and (re)definition in between
4468 driver and ACL definitions.
4470 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4471 forgetting to use the resulting value; it was using the unexpanded value.
4473 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4474 hadn't been configured. The fix is from Juergen Kreileder, who
4475 understands it better than I do:
4477 "Here's what I see happening with three configured cyrus_sasl
4478 authenticators configured (plain, login, cram-md5):
4480 On startup auth_cyrus_sasl_init() gets called for each of these.
4481 This means three calls to sasl_listmech() without a specified mech_list.
4482 => SASL tests which mechs of all available mechs actually work
4483 => three warnings about OTP not working
4484 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4486 With the patch, sasl_listmech() also gets called three times. But now
4487 SASL's mech_list option is set to the server_mech specified in the the
4488 authenticator. Or in other words, the answer from sasl_listmech()
4489 gets limited to just the mech you're testing for (which is different
4491 => the return list contains just 'plain' or 'login', 'cram-md5' or
4492 nothing depending on the value of ob->server_mech.
4494 I've just tested the patch: Authentication still works fine,
4495 unavailable mechs specified in the exim configuration are still
4496 caught, and the auth.log warnings about OTP are gone."
4498 PH/25 When debugging is enabled, the contents of the command line are added
4499 to the debugging output, even when log_selector=+arguments is not
4502 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4503 answer is "GNU", and only if the return is "GNU/something" is the answer
4506 PH/27 $acl_verify_message is now set immediately after the failure of a
4507 verification in an ACL, and so is available in subsequent modifiers. In
4508 particular, the message can be preserved by coding like this:
4510 warn !verify = sender
4511 set acl_m0 = $acl_verify_message
4513 Previously, $acl_verify_message was set only while expanding "message"
4514 and "log_message" when a very denied access.
4516 PH/28 Modified OS/os.c-Linux with
4518 -#ifndef OS_LOAD_AVERAGE
4519 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4521 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4522 nomenclature these days.)
4524 PH/29 Installed patch from the Sieve maintainer that adds the options
4525 sieve_useraddress and sieve_subaddress to the redirect router.
4527 PH/30 In these circumstances:
4528 . Two addresses routed to the same list of hosts;
4529 . First host does not offer TLS;
4530 . First host accepts first address;
4531 . First host gives temporary error to second address;
4532 . Second host offers TLS and a TLS session is established;
4533 . Second host accepts second address.
4534 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4535 and peerdn, if requested) that were in fact used only for the second
4538 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4539 attention to any local part prefix or suffix that was matched by the
4540 router that accepted the address. It now behaves in the same way as it
4541 does for delivery: the affixes are removed from the local part unless
4542 rcpt_include_affixes is set on the transport.
4544 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4545 timeout during the DATA phase of an incoming message.
4547 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4548 applied a patch sanctioned by the Sieve maintainer.
4550 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4551 the uid or gid is negative. A case of a negative gid caused this to be
4552 noticed. The fix allows for either to be negative.
4554 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4555 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4558 PH/36 The change PH/12 above was broken. Fixed it.
4560 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4561 the grounds that routing the same address twice would always produce the
4562 same answer. This might have been true once, but it is certainly no
4563 longer true now. Routing a child address may depend on the previous
4564 routing that produced that child. Some complicated redirection strategies
4565 went wrong when messages had multiple recipients, and made Exim's
4566 behaviour dependent on the order in which the addresses were given.
4568 I have moved the duplicate checking until after the routing is complete.
4569 Exim scans the addresses that are assigned to local and remote
4570 transports, and removes any duplicates. This means that more work will be
4571 done, as duplicates will always all be routed, but duplicates are
4572 presumably rare, so I don't expect this is of any significance.
4574 For deliveries to pipes, files, and autoreplies, the duplicate checking
4575 still happens during the routing process, since they are not going to be
4578 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4579 It corrects a timeout issue with spamd. This is Ian's comment: "The
4580 background is that sometimes spamd either never reads data from a
4581 connection it has accepted, or it never writes response data. The exiscan
4582 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4583 blindly assumes that writes won't block so it may never time out."
4585 PH/39 Allow G after quota size as well as K and M.
4587 PH/40 The value set for $authenticated_id in an authenticator may not contain
4588 binary zeroes or newlines because the value is written to log lines and
4589 to spool files. There was no check on this. Now the value is run through
4590 the string_printing() function so that such characters are converted to
4591 printable escape sequences.
4593 PH/41 $message_linecount is a new variable that contains the total number of
4594 lines in the message. Compare $body_linecount, which is the count for the
4597 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4598 bounce and delay warning messages, except in certain special cases, which
4601 (a) An SMTP error message from a remote host;
4602 (b) A message specified in a :fail: redirection;
4603 (c) A message specified in a "fail" command in a system filter;
4604 (d) A message specified in a FAIL return from the queryprogram router;
4605 (e) A message specified by the cannot_route_message router option.
4607 In these cases only, Exim does include the error details in bounce and
4608 warning messages. There are also a few cases where bland messages such
4609 as "unrouteable address" or "local delivery error" are given.
4611 PH/43 $value is now also set for the "else" part of a ${run expansion.
4613 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4614 being worked on, but at least Exim now implements the latest version to
4617 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4618 process to complete was treated as a delivery failure, a timeout while
4619 writing the message to the pipe was logged, but erroneously treated as a
4620 successful delivery. Such timeouts include transport filter timeouts. For
4621 consistency with the overall process timeout, these timeouts are now
4622 treated as errors, giving rise to delivery failures by default. However,
4623 there is now a new Boolean option for the pipe transport called
4624 timeout_defer, which, if set TRUE, converts the failures into defers for
4625 both kinds of timeout. A transport filter timeout is now identified in
4628 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4629 systems where "make" and "gmake" are different, calling "gmake" at top
4630 level broke things. I've arranged for the value of $(MAKE) to be passed
4631 from the Makefile to this script so that it can call the same version of
4635 A note about Exim versions 4.44 and 4.50
4636 ----------------------------------------
4638 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4639 changes of various kinds. As a consequence, a big documentation update was
4640 needed. This delayed the release for rather longer than seemed good, especially
4641 in the light of a couple of (minor) security issues. Therefore, the changes
4642 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4643 release. So 4.44 and 4.50 are in effect two different branches that both start
4646 I have left the 4.50 change log unchanged; it contains all the changes since
4647 4.43. The change log for 4.44 is below; many of its items are identical to
4648 those for 4.50. This seems to be the most sensible way to preserve the
4649 historical information.
4655 1. Minor wording change to the doc/README.SIEVE file.
4657 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4658 computation of the current number of files was incorrect.
4660 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4661 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4664 4. Give more explanation in the error message when the command for a transport
4665 filter fails to execute.
4667 5. There are several places where Exim runs a non-Exim command in a
4668 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4669 was being done only for the command run by the queryprogram router. It is
4670 now done for all such subprocesses. The other cases are: ${run, transport
4671 filters, and the commands run by the lmtp and pipe transports.
4673 6. Added CONFIGURE_GROUP build-time option.
4675 7. Some older OS have a limit of 256 on the maximum number of file
4676 descriptors. Exim was using setrlimit() to set 1000 as a large value
4677 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4678 systems. I've change it so that if it can't get 1000, it tries for 256.
4680 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4681 was an oversight, and furthermore, ever since the addition of extra
4682 controls (e.g. 4.43/32), the checks on when to allow different forms of
4683 "control" were broken. There should now be diagnostics for all cases when a
4684 control that does not make sense is encountered.
4686 9. Added the /retain_sender option to "control=submission".
4688 10. $recipients is now available in the predata ACL (oversight).
4690 11. Tidy the search cache before the fork to do a delivery from a message
4691 received from the command line. Otherwise the child will trigger a lookup
4692 failure and thereby defer the delivery if it tries to use (for example) a
4693 cached ldap connection that the parent has called unbind on.
4695 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4696 of $address_data from the recipient verification was clobbered by the
4697 sender verification.
4699 13. The value of address_data from a sender verification is now available in
4700 $sender_address_data in subsequent conditions in the ACL statement.
4702 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4704 15. Added a new option "connect=<time>" to callout options, to set a different
4707 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4708 was its contents. (It was OK if the option was not defined at all.)
4710 17. A "Completed" log line is now written for messages that are removed from
4711 the spool by the -Mrm option.
4713 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4714 information about exactly what failed.
4716 19. Added -dd to debug only the daemon process.
4718 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4719 handles timeouts, both on the server side and network timeouts. Renamed the
4720 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4722 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4723 It is now set to "smtps".
4725 22. $host_address is now set to the target address during the checking of
4726 ignore_target_hosts.
4728 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4729 being passed; this would have caused $sender_host_name to have been used if
4730 matching the list had actually called for a host name (not very likely,
4731 since this list is usually IP addresses). A host name is now passed as
4734 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4735 code that specifies a non-restarting handler (typically sigaction() in
4736 modern systems) in an attempt to fix a rare and obscure crash bug.
4738 25. Narrowed the window for a race in the daemon that could cause it to ignore
4739 SIGCHLD signals. This is not a major problem, because they are used only to
4740 wake it up if nothing else does.
4742 26. A malformed maildirsize file could cause Exim to calculate negative values
4743 for the mailbox size or file count. Odd effects could occur as a result.
4744 The maildirsize information is now recalculated if the size or filecount
4747 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4748 support for a long time. Removed HAVE_SYS_VFS_H.
4750 28. Installed the latest version of exipick from John Jetmore.
4752 29. In an address list, if the pattern was not a regular expression, an empty
4753 subject address (from a bounce message) matched only if the pattern was an
4754 empty string. Non-empty patterns were not even tested. This was the wrong
4755 because it is perfectly reasonable to use an empty address as part of a
4756 database query. An empty address is now tested by patterns that are
4757 lookups. However, all the other forms of pattern expect the subject to
4758 contain a local part and a domain, and therefore, for them, an empty
4759 address still always fails if the pattern is not itself empty.
4761 30. Exim went into a mad DNS loop when attempting to do a callout where the
4762 host was specified on an smtp transport, and looking it up yielded more
4763 than one IP address.
4765 31. Re-factored the code for checking spool and log partition space into a
4766 function that finds that data and another that does the check. The former
4767 is then used to implement four new variables: $spool_space, $log_space,
4768 $spool_inodes, and $log_inodes.
4770 32. The RFC2047 encoding function was originally intended for short strings
4771 such as real names; it was not keeping to the 75-character limit for
4772 encoded words that the RFC imposes. It now respects the limit, and
4773 generates multiple encoded words if necessary. To be on the safe side, I
4774 have increased the buffer size for the ${rfc2047: expansion operator from
4777 33. It is now permitted to omit both strings after an "if" condition; if the
4778 condition is true, the result is "true". As before, when the second string
4779 is omitted, a false condition yields an empty string. This makes it less
4780 cumbersome to write custom ACL and router conditions.
4782 34. Failure to deliver a bounce message always caused it to be frozen, even if
4783 there was an errors_to setting on the router. The errors_to setting is now
4786 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4787 canonical form (fully expanded) before being placed in
4788 $sender_host_address.
4790 36. The table in the code that translates DNS record types into text (T_A to
4791 "A" for instance) was missing entries for NS and CNAME. It is just possible
4792 that this could have caused confusion if both these types were looked up
4793 for the same domain, because the text type is used as part of Exim's
4794 per-process caching. But the chance of anyone hitting this buglet seems
4797 37. The dnsdb lookup has been extended in a number of ways.
4799 (1) There is a new type, "zns", which walks up the domain tree until it
4800 finds some nameserver records. It should be used with care.
4802 (2) There is a new type, "mxh", which is like "mx" except that it returns
4803 just the host names, not the priorities.
4805 (3) It is now possible to give a list of domains (or IP addresses) to be
4806 looked up. The behaviour when one of the lookups defers can be
4807 controlled by a keyword.
4809 (4) It is now possible to specify the separator character for use when
4810 multiple records are returned.
4812 38. The dnslists ACL condition has been extended: it is now possible to supply
4813 a list of IP addresses and/or domains to be looked up in a particular DNS
4816 39. Added log_selector=+queue_time_overall.
4818 40. When running the queue in the test harness, wait just a tad after forking a
4819 delivery process, to get repeatability of debugging output.
4821 41. Include certificate and key file names in error message when GnuTLS fails
4822 to set them up, because the GnuTLS error message doesn't include the name
4823 of the failing file when there is a problem reading it.
4825 42. Allow both -bf and -bF in the same test run.
4827 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4829 44. The "Exiscan patch" is now merged into the mainline Exim source.
4831 45. Sometimes the final signoff response after QUIT could fail to get
4832 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4833 < 0 before doing a fflush(). This bug looks as though it goes back to the
4834 introduction of TLS in release 3.20, but "sometimes" must have been rare
4835 because the tests only now provoked it.
4837 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4838 (this can affect the format of dates).
4840 47. exim_tidydb, when checking for the continued existence of a message for
4841 which it has found a message-specific retry record, was not finding
4842 messages that were in split spool directories. Consequently, it was
4843 deleting retry records that should have stayed in existence.
4845 48. Steve fixed some bugs in eximstats.
4847 49. The SPA authentication driver was not abandoning authentication and moving
4848 on to the next authenticator when an expansion was forced to fail,
4849 contradicting the general specification for all authenticators. Instead it
4850 was generating a temporary error. It now behaves as specified.
4852 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4853 (the order specifies the preference for clients). The order is now AES256,
4854 AES128, 3DES, ARCFOUR128.
4856 51. Small patch to Sieve code - explicitly set From: when generating an
4859 52. Exim crashed if a remote delivery caused a very long error message to be
4860 recorded - for instance if somebody sent an entire SpamAssassin report back
4861 as a large number of 550 error lines. This bug was coincidentally fixed by
4862 increasing the size of one of Exim's internal buffers (big_buffer) that
4863 happened as part of the Exiscan merge. However, to be on the safe side, I
4864 have made the code more robust (and fixed the comments that describe what
4867 53. Now that there can be additional text after "Completed" in log lines (if
4868 the queue_time_overall log selector is set), a one-byte patch to exigrep
4869 was needed to allow it to recognize "Completed" as not the last thing in
4872 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4873 patch that reportedly fixes this has been added. I am not expert enough to
4874 create a test for it. This is what the patch creator wrote:
4876 "I found a little strange behaviour of ldap code when working with
4877 Windows 2003 AD Domain, where users was placed in more than one
4878 Organization Units. When I tried to give exim partial DN, the exit code
4879 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4880 But simultaneously result of request was absolutely normal ldap result,
4881 so I produce this patch..."
4883 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4884 REFERENCE, so I have modified the code to exclude the patch when that macro
4887 55. Some experimental protocols are using DNS PTR records for new purposes. The
4888 keys for these records are domain names, not reversed IP addresses. The
4889 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4890 leaves it alone. Component reversal etc. now happens only for IP addresses.
4893 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4895 57. Double the size of the debug message buffer (to 2048) so that more of very
4896 long debug lines gets shown.
4898 58. The exicyclog utility now does better if the number of log files to keep
4899 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4901 59. Two changes related to the smtp_active_hostname option:
4903 (1) $smtp_active_hostname is now available as a variable.
4904 (2) The default for smtp_banner uses $smtp_active_hostname instead
4905 of $primary_hostname.
4907 60. The host_aton() function is supposed to be passed a string that is known
4908 to be a valid IP address. However, in the case of IPv6 addresses, it was
4909 not checking this. This is a hostage to fortune. Exim now panics and dies
4910 if the condition is not met. A case was found where this could be provoked
4911 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4912 components; fortuitously, this particular loophole had already been fixed
4913 by change 4.50/55 above.
4915 If there are any other similar loopholes, the new check in host_aton()
4916 itself should stop them being exploited. The report I received stated that
4917 data on the command line could provoke the exploit when Exim was running as
4918 exim, but did not say which command line option was involved. All I could
4919 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4920 running as the user.
4923 61. There was a buffer overflow vulnerability in the SPA authentication code
4924 (which came originally from the Samba project). I have added a test to the
4925 spa_base64_to_bits() function which I hope fixes it.
4928 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4929 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4931 63. The daemon start-up calls getloadavg() while still root for those OS that
4932 need the first call to be done as root, but it missed one case: when
4933 deliver_queue_load_max is set with deliver_drop_privilege. This is
4934 necessary for the benefit of the queue runner, because there is no re-exec
4935 when deliver_drop_privilege is set.
4937 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4938 This has been fixed.
4940 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4941 was in use, was not putting the data itself into the right store pool;
4942 consequently, it could be overwritten for a subsequent message in the same
4943 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4946 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4948 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4949 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4950 the function were treating the return as a boolean value, which happened to
4951 work because 0=false and not-0=true, but is not correct code.
4953 68. The host_aton() function was not handling scoped IPv6 addresses (those
4954 with, for example, "%eth0" on the end) correctly.
4956 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4957 negated items (that is, ~something) in unsigned ints. Some compilers
4958 apparently mutter when there is no cast.
4960 70. If an address verification called from an ACL failed, and did not produce a
4961 user-specific message (i.e. there was only a "system" message), nothing was
4962 put in $acl_verify_message. In this situation, it now puts the system
4965 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4966 session; change 4.31/43 added the unwanted input to the log line - except
4967 that it did not do this in the start of session case. It now does.
4969 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4970 This is wrong and can cause the other end to generate a synchronization
4971 error if it is another Exim or anything else that does the synchronization
4972 check. A QUIT command is no longer sent after a timeout.
4974 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4975 during host lookups.
4977 74. The defer_ok option of callout verification was not working if it was used
4978 when verifying addresses in header lines, that is, for this case:
4980 verify = header_sender/callout=defer_ok
4982 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4983 those file descriptors could be used for SMTP connections. If anything
4984 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4985 could be sent to the SMTP client, causing chaos. The daemon now opens
4986 stdin, stdout, and stderr to /dev/null when it puts itself into the
4989 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4990 log by default. The user can override this with suitable Perl magic.
4992 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4993 the log message when discard triggers, was not working for the DATA ACL or
4994 for the non-SMTP ACL.
4996 78. Error message wording change in sieve.c.
4998 79. If smtp_accept_max_per_host was set, the number of connections could be
4999 restricted to fewer than expected, because the daemon was trying to set up
5000 a new connection before checking whether the processes handling previous
5001 connections had finished. The check for completed processes is now done
5002 earlier. On busy systems, this bug wouldn't be noticed because something
5003 else would have woken the daemon, and it would have reaped the completed
5006 80. If a message was submitted locally by a user whose login name contained one
5007 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5008 It caused a spool format error. I have fixed the spool reading code. A
5009 related problem was that the "from" clause in the Received: line became
5010 illegal because of the space(s). It is now covered by ${quote_local_part.
5012 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5015 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5017 83. Patch from Sieve maintainer to fix unterminated string problem in
5018 "vacation" handling.
5020 84. Some minor changes to the Linux configuration files to help with other
5021 OS variants using glibc.
5023 85. One more patch for Sieve to update vacation handling to latest spec.
5026 ----------------------------------------------------
5027 See the note above about the 4.44 and 4.50 releases.
5028 ----------------------------------------------------
5034 1. Change 4.43/35 introduced a bug that caused file counts to be
5035 incorrectly computed when quota_filecount was set in an appendfile
5038 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5039 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5042 3. Give more explanation in the error message when the command for a transport
5043 filter fails to execute.
5045 4. There are several places where Exim runs a non-Exim command in a
5046 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5047 was being done only for the command run by the queryprogram router. It is
5048 now done for all such subprocesses. The other cases are: ${run, transport
5049 filters, and the commands run by the lmtp and pipe transports.
5051 5. Some older OS have a limit of 256 on the maximum number of file
5052 descriptors. Exim was using setrlimit() to set 1000 as a large value
5053 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5054 systems. I've change it so that if it can't get 1000, it tries for 256.
5056 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5057 was an oversight, and furthermore, ever since the addition of extra
5058 controls (e.g. 4.43/32), the checks on when to allow different forms of
5059 "control" were broken. There should now be diagnostics for all cases when a
5060 control that does not make sense is encountered.
5062 7. $recipients is now available in the predata ACL (oversight).
5064 8. Tidy the search cache before the fork to do a delivery from a message
5065 received from the command line. Otherwise the child will trigger a lookup
5066 failure and thereby defer the delivery if it tries to use (for example) a
5067 cached ldap connection that the parent has called unbind on.
5069 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5070 of $address_data from the recipient verification was clobbered by the
5071 sender verification.
5073 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5074 was its contents. (It was OK if the option was not defined at all.)
5076 11. A "Completed" log line is now written for messages that are removed from
5077 the spool by the -Mrm option.
5079 12. $host_address is now set to the target address during the checking of
5080 ignore_target_hosts.
5082 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5083 being passed; this would have caused $sender_host_name to have been used if
5084 matching the list had actually called for a host name (not very likely,
5085 since this list is usually IP addresses). A host name is now passed as
5088 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5089 code that specifies a non-restarting handler (typically sigaction() in
5090 modern systems) in an attempt to fix a rare and obscure crash bug.
5092 15. Narrowed the window for a race in the daemon that could cause it to ignore
5093 SIGCHLD signals. This is not a major problem, because they are used only to
5094 wake it up if nothing else does.
5096 16. A malformed maildirsize file could cause Exim to calculate negative values
5097 for the mailbox size or file count. Odd effects could occur as a result.
5098 The maildirsize information is now recalculated if the size or filecount
5101 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5102 support for a long time. Removed HAVE_SYS_VFS_H.
5104 18. Updated exipick to current release from John Jetmore.
5106 19. Allow an empty sender to be matched against a lookup in an address list.
5107 Previously the only cases considered were a regular expression, or an
5110 20. Exim went into a mad DNS lookup loop when doing a callout where the
5111 host was specified on the transport, if the DNS lookup yielded more than
5114 21. The RFC2047 encoding function was originally intended for short strings
5115 such as real names; it was not keeping to the 75-character limit for
5116 encoded words that the RFC imposes. It now respects the limit, and
5117 generates multiple encoded words if necessary. To be on the safe side, I
5118 have increased the buffer size for the ${rfc2047: expansion operator from
5121 22. Failure to deliver a bounce message always caused it to be frozen, even if
5122 there was an errors_to setting on the router. The errors_to setting is now
5125 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5126 canonical form (fully expanded) before being placed in
5127 $sender_host_address.
5129 24. Updated eximstats to version 1.33
5131 25. Include certificate and key file names in error message when GnuTLS fails
5132 to set them up, because the GnuTLS error message doesn't include the name
5133 of the failing file when there is a problem reading it.
5135 26. Expand error message when OpenSSL has problems setting up cert/key files.
5138 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5139 (this can affect the format of dates).
5141 28. exim_tidydb, when checking for the continued existence of a message for
5142 which it has found a message-specific retry record, was not finding
5143 messages that were in split spool directories. Consequently, it was
5144 deleting retry records that should have stayed in existence.
5146 29. eximstats updated to version 1.35
5147 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5148 1.35 - bugfix such that pie charts by volume are generated correctly
5150 30. The SPA authentication driver was not abandoning authentication and moving
5151 on to the next authenticator when an expansion was forced to fail,
5152 contradicting the general specification for all authenticators. Instead it
5153 was generating a temporary error. It now behaves as specified.
5155 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5156 (the order specifies the preference for clients). The order is now AES256,
5157 AES128, 3DES, ARCFOUR128.
5159 31. Small patch to Sieve code - explicitly set From: when generating an
5162 32. Exim crashed if a remote delivery caused a very long error message to be
5163 recorded - for instance if somebody sent an entire SpamAssassin report back
5164 as a large number of 550 error lines. This bug was coincidentally fixed by
5165 increasing the size of one of Exim's internal buffers (big_buffer) that
5166 happened as part of the Exiscan merge. However, to be on the safe side, I
5167 have made the code more robust (and fixed the comments that describe what
5170 33. Some experimental protocols are using DNS PTR records for new purposes. The
5171 keys for these records are domain names, not reversed IP addresses. The
5172 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5173 leaves it alone. Component reversal etc. now happens only for IP addresses.
5176 34. The host_aton() function is supposed to be passed a string that is known
5177 to be a valid IP address. However, in the case of IPv6 addresses, it was
5178 not checking this. This is a hostage to fortune. Exim now panics and dies
5179 if the condition is not met. A case was found where this could be provoked
5180 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5181 components; fortuitously, this particular loophole had already been fixed
5182 by change 4.50/55 or 4.44/33 above.
5184 If there are any other similar loopholes, the new check in host_aton()
5185 itself should stop them being exploited. The report I received stated that
5186 data on the command line could provoke the exploit when Exim was running as
5187 exim, but did not say which command line option was involved. All I could
5188 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5189 running as the user.
5192 35. There was a buffer overflow vulnerability in the SPA authentication code
5193 (which came originally from the Samba project). I have added a test to the
5194 spa_base64_to_bits() function which I hope fixes it.
5197 36. The daemon start-up calls getloadavg() while still root for those OS that
5198 need the first call to be done as root, but it missed one case: when
5199 deliver_queue_load_max is set with deliver_drop_privilege. This is
5200 necessary for the benefit of the queue runner, because there is no re-exec
5201 when deliver_drop_privilege is set.
5203 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5204 was in use, was not putting the data itself into the right store pool;
5205 consequently, it could be overwritten for a subsequent message in the same
5206 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5209 38. Sometimes the final signoff response after QUIT could fail to get
5210 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5211 < 0 before doing a fflush(). This bug looks as though it goes back to the
5212 introduction of TLS in release 3.20, but "sometimes" must have been rare
5213 because the tests only now provoked it.
5219 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5220 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5221 Now it returns an integer. A number of calls were still expecting a T/F
5222 return. Fortuitously, in all cases, the tests worked in OK situations,
5223 which is the norm. However, things would have gone wrong on any write
5224 failures on the smtp file descriptor. This function is used when sending
5225 messages over SMTP and also when doing verify callouts.
5227 2. When Exim is called to do synchronous delivery of a locally submitted
5228 message (the -odf or -odi options), it no longer closes stderr before doing
5231 3. Implemented the mua_wrapper option.
5233 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5235 5. Implemented the functions header_remove(), header_testname(),
5236 header_add_at_position(), and receive_remove_recipient(), and exported them
5239 6. If an ACL "warn" statement specified the addition of headers, Exim already
5240 inserted X-ACL-Warn: at the start if there was no header name. However, it
5241 was not making this test for the second and subsequent header lines if
5242 there were newlines in the string. This meant that an invalid header could
5243 be inserted if Exim was badly configured.
5245 7. Allow an ACL "warn" statement to add header lines at the start or after all
5246 the Received: headers, as well as at the end.
5248 8. Added the rcpt_4xx retry error code.
5250 9. Added postmaster_mailfrom=xxx to callout verification option.
5252 10. Added mailfrom=xxxx to the callout verification option, for verify=
5255 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5256 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5258 12. Inserted some casts to stop certain compilers warning when using pointer
5259 differences as field lengths or precisions in printf-type calls (mostly
5260 affecting debugging statements).
5262 13. Added optional readline() support for -be (dynamically loaded).
5264 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5265 same clock tick as a message's arrival, so that its received time was the
5266 same as the "first fail" time on the retry record, and that message
5267 remained on the queue past the ultimate address timeout, every queue runner
5268 would try a delivery (because it was past the ultimate address timeout) but
5269 after another failure, the ultimate address timeout, which should have then
5270 bounced the address, did not kick in. This was a "< instead of <=" error;
5271 in most cases the first failure would have been in the next clock tick
5272 after the received time, and all would be well.
5274 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5275 being recognized when the domain list was tested by the match_domain
5276 condition in an expansion string.
5278 16. Added the ${str2b64: operator.
5280 17. Exim was always calling setrlimit() to set a large limit for the number of
5281 processes, without checking whether the existing limit was already
5282 adequate. (It did check for the limit on file descriptors.) Furthermore,
5283 errors from getrlimit() and setrlimit() were being ignored. Now they are
5284 logged to the main and panic logs, but Exim does carry on, to try to do its
5285 job under whatever limits there are.
5287 18. Imported PCRE 5.0.
5289 19. Trivial typo in log message " temporarily refused connection" (the leading
5292 20. If the log selector return_path_on_delivery was set and an address was
5293 redirected to /dev/null, the delivery process crashed because it assumed
5294 that a return path would always be set for a "successful" delivery. In this
5295 case, the whole delivery is bypassed as an optimization, and therefore no
5298 21. Internal re-arrangement: the function for sending a challenge and reading
5299 a response while authentication was assuming a zero-terminated challenge
5300 string. It's now changed to take a pointer and a length, to allow for
5301 binary data in such strings.
5303 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5305 23. Exim was not respecting finduser_retries when seeking the login of the
5306 uid under which it was called; it was always trying 10 times. (The default
5307 setting of finduser_retries is zero.) Also, it was sleeping after the final
5308 failure, which is pointless.
5310 24. Implemented tls_on_connect_ports.
5312 25. Implemented acl_smtp_predata.
5314 26. If the domain in control=submission is set empty, Exim assumes that the
5315 authenticated id is a complete email address when it generates From: or
5316 Sender: header lines.
5318 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5319 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5320 chown and chgrp in /bin and hostname in /usr/bin.
5322 28. Exim was keeping the "process log" file open after each use, just as it
5323 does for the main log. This opens the possibility of it remaining open for
5324 long periods when the USR1 signal hits a daemon. Occasional processlog
5325 errors were reported, that could have been caused by this. Anyway, it seems
5326 much more sensible not to leave this file open at all, so that is what now
5329 29. The long-running daemon process does not normally write to the log once it
5330 has entered its main loop, and it closes the log before doing so. This is
5331 so that log files can straightforwardly be renamed and moved. However,
5332 there are a couple of unusual error situations where the daemon does write
5333 log entries, and I had neglected to close the log afterwards.
5335 30. The text of an SMTP error response that was received during a remote
5336 delivery was being truncated at 512 bytes. This is too short for some of
5337 the long messages that one sometimes sees. I've increased the limit to
5340 31. It is now possible to make retry rules that apply only when a message has a
5341 specific sender, in particular, an empty sender.
5343 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5344 it possible to be selective about when SMTP synchronization is enforced.
5346 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5348 32. Implemented hosts_connection_nolog.
5350 33. Added an ACL for QUIT.
5352 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5355 35. Added mailbox_size and mailbox_filecount to appendfile.
5357 36. Added control = no_multiline_responses to ACLs.
5359 37. There was a bug in the logic of the code that waits for the clock to tick
5360 in the case where the clock went backwards by a substantial amount such
5361 that the microsecond fraction of "now" was more than the microsecond
5362 fraction of "then" (but the whole seconds number was less).
5364 38. Added support for the libradius Radius client library this is found on
5365 FreeBSD (previously only the radiusclient library was supported).
5371 1. When certain lookups returned multiple values in the form name=value, the
5372 quoting of the values was not always being done properly. Specifically:
5373 (a) If the value started with a double quote, but contained no whitespace,
5375 (b) If the value contained whitespace other than a space character (i.e.
5376 tabs or newlines or carriage returns) it was not quoted.
5377 This fix has been applied to the mysql and pgsql lookups by writing a
5378 separate quoting function and calling it from the lookup code. The fix
5379 should probably also be applied to nisplus, ibase and oracle lookups, but
5380 since I cannot test any of those, I have not disturbed their existing code.
5382 2. A hit in the callout cache for a specific address caused a log line with no
5383 reason for rejecting RCPT. Now it says "Previous (cached) callout
5384 verification failure".
5386 3. There was an off-by-one bug in the queryprogram router. An over-long
5387 return line was truncated at 256 instead of 255 characters, thereby
5388 overflowing its buffer with the terminating zero. As well as fixing this, I
5389 have increased the buffer size to 1024 (and made a note to document this).
5391 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5392 when Exim is waiting for an SMTP response from a remote server, Exim
5393 restarts its select() call on the socket, thereby resetting its timeout.
5394 This is not a problem when such interrupts are rare. Somebody set up a cron
5395 job to run exiwhat every 2 minutes, which is less than the normal select()
5396 timeout (5 or 10 minutes). This meant that the select() timeout never
5397 kicked in because it was always reset. I have fixed this by comparing the
5398 time when an interrupt arrives with the time at the start of the first call
5399 to select(). If more time than the timeout has elapsed, the interrupt is
5400 treated as a timeout.
5402 5. Some internal re-factoring in preparation for the addition of Sieve
5403 extensions (by MH). In particular, the "personal" test is moved to a
5404 separate function, and given an option for scanning Cc: and Bcc: (which is
5405 not set for Exim filters).
5407 6. When Exim created an email address using the login of the caller as the
5408 local part (e.g. when creating a From: or Sender: header line), it was not
5409 quoting the local part when it contained special characters such as @.
5411 7. Installed new OpenBSD configuration files.
5413 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5414 try to make them clearer.
5416 9. Callout options, other than the timeout value, were being ignored when
5417 verifying sender addresses in header lines. For example, when using
5419 verify = header_sender/callout=no_cache
5421 the cache was (incorrectly) being used.
5423 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5424 only the Cygwin environment.
5426 11. When return_path_on_delivery was set as a log selector, if different remote
5427 addresses in the same message used different return paths and parallel
5428 remote delivery occurred, the wrong values would sometimes be logged.
5429 (Whenever a remote delivery process finished, the return path value from
5430 the most recently started remote delivery process was logged.)
5432 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5433 lines when AUTH and/or TLS are in use. This is the "received protocol"
5434 field. Exim used to use "asmtp" for authenticated SMTP, without any
5435 indication (in the protocol name) for TLS use. Now it follows the RFC and
5436 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5437 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5438 names appear in log lines as well as in Received: header lines.
5440 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5441 extensions, and comparison tests, and to fix some bugs.
5443 14. Changes to the "personal" filter test:
5445 (1) The test was buggy in that it was just doing the equivalent of
5446 "contains" tests on header lines. For example, if a user's address was
5447 anne@some.where, the "personal" test would incorrectly be true for
5449 To: susanne@some.where
5451 This test is now done by extracting each address from the header in turn,
5452 and checking the entire address. Other tests that are part of "personal"
5453 are now done using regular expressions (for example, to check local parts
5454 of addresses in From: header lines).
5456 (2) The list of non-personal local parts in From: addresses has been
5457 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5458 taken from the Sieve specification recommendations.
5460 (3) If the message contains any header line starting with "List-" it is
5461 treated as non-personal.
5463 (4) The test for "circular" in the Subject: header line has been removed
5464 because it now seems ill-conceived.
5466 15. Minor typos in src/EDITME comments corrected.
5468 16. Installed latest exipick from John Jetmore.
5470 17. If headers_add on a router specified a text string that was too long for
5471 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5472 of string_sprintf() is now avoided.
5474 18. $message_body_size was not set (it was always zero) when running the DATA
5475 ACL and the local_scan() function.
5477 19. For the "mail" command in an Exim filter, no default was being set for
5478 the once_repeat time, causing a random time value to be used if "once" was
5479 specified. (If the value happened to be <= 0, no repeat happened.) The
5480 default is now 0s, meaning "never repeat". The "vacation" command was OK
5481 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5482 (I found it when inspecting the code).
5484 20. There is now an overall timeout for performing a callout verification. It
5485 defaults to 4 times the callout timeout, which applies to individual SMTP
5486 commands during the callout. The overall timeout applies when there is more
5487 than one host that can be tried. The timeout is checked before trying the
5488 next host. This prevents very long delays if there are a large number of
5489 hosts and all are timing out (e.g. when the network connections are timing
5490 out). The value of the overall timeout can be changed by specifying an
5491 additional sub-option for "callout", called "maxwait". For example:
5493 verify = sender/callout=5s,maxwait=20s
5495 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5496 to the end before writing, but this should make it even safer).
5498 22. Exim was forgetting that it had advertised PIPELINING for the second and
5499 subsequent messages on an SMTP connection. It was also not resetting its
5500 memory on STARTTLS and an internal HELO.
5502 23. When Exim logs an SMTP synchronization error within a session, it now
5503 records whether PIPELINING has been advertised or not.
5505 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5506 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5507 rather than long int.
5509 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5511 26. Added the never_mail option to autoreply.
5517 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5518 crash if the getsockname() call failed; this can happen if a connection is
5519 closed very soon after it is established. The problem was simply in the
5520 order in which certain operations were done, causing Exim to try to write
5521 to the SMTP stream before it had set up the file descriptor. The bug has
5522 been fixed by making things happen in the correct order.
5528 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5529 before the connection was closed, thus losing the rejection response.
5531 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5532 some early Solaris releases, but causes trouble in current releases where
5533 socklen_t is defined.
5535 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5538 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5541 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5542 that releasing the top of it at the end releases what was used for sub-
5543 expansions (unless the block got too big). However, discard this block if
5544 the first thing is a variable or header, so that we can use its block when
5545 it is dynamic (useful for very large $message_headers, for example).
5547 6. Lookups now cache *every* query, not just the most recent. A new, separate
5548 store pool is used for this. It can be recovered when all lookup caches are
5549 flushed. Lookups now release memory at the end of their result strings.
5550 This has involved some general refactoring of the lookup sources.
5552 7. Some code has been added to the store_xxx() functions to reduce the amount
5553 of flapping under certain conditions.
5555 8. log_incoming_interface used to affect only the <= reception log lines. Now
5556 it causes the local interface and port to be added to several more SMTP log
5557 lines, for example "SMTP connection from", and rejection lines.
5559 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5561 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5563 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5564 could be overwritten at the end of the current message (or the start of a
5565 new message if it was set in a HELO ACL). The value is now preserved for
5566 the duration of the SMTP connection.
5568 12. If a transport had a headers_rewrite setting, and a matching header line
5569 contained an unqualified address, that address was qualified, even if it
5570 did not match any rewriting rules. The underlying bug was that the values
5571 of the flags that permit the existence of unqualified sender and recipient
5572 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5573 non-local messages, and by -bnq for local messages) were not being
5574 preserved with the message after it was received.
5576 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5577 "next input=" as part of the text comprising the host identity instead of
5578 the correct text. The code was using the same buffer for two different
5579 strings. However, depending on which order the printing function evaluated
5580 its arguments, the bug did not always show up. Under Linux, for example, my
5581 test suite worked just fine.
5583 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5584 doesn't work with some older versions of Perl. It has been changed to "my",
5585 which in any case is probably the better facility to use.
5587 15. A really picky compiler found some instances of statements for creating
5588 error messages that either had too many or two few arguments for the format
5591 16. The size of the buffer for calls to the DNS resolver has been increased
5592 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5593 for addresses that have a lot of PTR records. This alleviates a problem; it
5594 does not fully solve it.
5596 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5597 buffer now truncates the list and logs the incident, which is the same
5598 action as happens when Exim is looking up a host name and its aliases.
5599 Previously in this situation something unpredictable would happen;
5600 sometimes it was "internal error: store_reset failed".
5602 18. If a server dropped the connection unexpectedly when an Exim client was
5603 using GnuTLS and trying to read a response, the client delivery process
5604 crashed while trying to generate an error log message.
5606 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5607 string, for example:
5609 warn message = H1: something\nH2: something
5611 the text was added as a single header line from Exim's point of view
5612 though it ended up OK in the delivered message. However, searching for the
5613 second and subsequent header lines using $h_h2: did not work. This has been
5614 fixed. Similarly, if a system filter added multiple headers in this way,
5615 the routers could not see them.
5617 20. Expanded the error message when iplsearch is called with an invalid key to
5618 suggest using net-iplsearch in a host list.
5620 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5621 ACLs are no longer actually imposed (and a message to that effect is
5624 22. If a "gecos" field in a passwd entry contained escaped characters, in
5625 particular, if it contained a \" sequence, Exim got it wrong when building
5626 a From: or a Sender: header from that name. A second bug also caused
5627 incorrect handling when an unquoted " was present following a character
5628 that needed quoting.
5630 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5631 was not being matched caselessly.
5633 24. Arranged for all hyphens in the exim.8 source to be escaped with
5636 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5637 properly. Recipient callout cache records were still being keyed to include
5638 the sender, even when use_sender was set false. This led to far more
5639 callouts that were necessary. The sender is no longer included in the key
5640 when use_sender is false.
5642 26. Added "control = submission" modifier to ACLs.
5644 27. Added the ${base62d: operator to decode base 62 numbers.
5646 28. dnsdb lookups can now access SRV records.
5648 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5649 the configuration file.
5651 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5652 (-v) mode. This makes the output for a verbose queue run more intelligible.
5654 31. Added a use_postmaster feature to recipient callouts.
5656 32. Added the $body_zerocount variable, containing the number of binary zero
5657 bytes in the message body.
5659 33. The time of last modification of the "new" subdirectory is now used as the
5660 "mailbox time last read" when there is a quota error for a maildir
5663 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5665 35. Added +ignore_unknown as a special item in host lists.
5667 36. Code for decoding IPv6 addresses in host lists is now included, even if
5668 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5669 address was recognized as an IP address, but was then not correctly decoded
5670 into binary, causing unexpected and incorrect effects when compared with
5677 1. Very minor rewording of debugging text in manualroute to say "list of
5678 hosts" instead of "hostlist".
5680 2. If verify=header_syntax was set, and a header line with an unqualified
5681 address (no domain) and a large number of spaces between the end of the
5682 name and the colon was received, the reception process suffered a buffer
5683 overflow, and (when I tested it) crashed. This was caused by some obsolete
5684 code that should have been removed. The fix is to remove it!
5686 3. When running in the test harness, delay a bit after writing a bounce
5687 message to get a bit more predictability in the log output.
5689 4. Added a call to search_tidyup() just before forking a reception process. In
5690 theory, someone could use a lookup in the expansion of smtp_accept_max_
5691 per_host which, without the tidyup, could leave open a database connection.
5693 5. Added the variables $recipient_data and $sender_data which get set from a
5694 lookup success in an ACL "recipients" or "senders" condition, or a router
5695 "senders" option, similar to $domain_data and $local_part_data.
5697 6. Moved the writing of debug_print from before to after the "senders" test
5700 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5701 problems for message scanning, either using a data ACL, or using
5702 local_scan() because the Received: header was not generated till after they
5703 were called (in order to set the time as the time of reception completion).
5704 I have revised the way this works. The header is now generated after the
5705 body is received, but before the ACL or local_scan() are called. After they
5706 are run, the timestamp in the header is updated.
5712 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5713 before starting a queue runner without re-exec. This happened only when
5714 deliver_drop_privilege was set or when the Exim user was set to root. The
5715 effect of the bug was that timeouts during subsequent deliveries caused
5716 crashes instead of being properly handled. The handler is now left at its
5717 default (and expected) setting.
5719 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5720 message, again when deliver_drop_privilege is set or Exim is run as root.
5721 The bug described in (1) was not present in this case, but the tidying up
5722 of the other signals was missing. I have made the two cases consistent.
5724 3. The ignore_target_hosts setting on a manualroute router was being ignored
5725 for hosts that were looked up using the /MX notation.
5727 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5730 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5731 operated on the sender address. After changing the $sender_address to <>
5732 for the sender address verify, Exim was re-instated it as the original
5733 (before rewriting) address, but remembering that it had rewritten it, so it
5734 wasn't rewriting it again. This bug also had the effect of breaking the
5735 sender address verification caching when the sender address was rewritten.
5737 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5738 This has been changed so that if the ip literal address matches
5739 ignore_target_hosts, the router declines.
5741 7. Added expansion conditions match_domain, match_address, and match_local_
5742 part (NOT match_host).
5744 8. The placeholder for the Received: header didn't have a length field set.
5746 9. Added code to Exim itself and to exim_lock to test for a specific race
5747 condition that could lead to file corruption when using MBX delivery. The
5748 issue is with the lockfile that is created in /tmp. If this file is removed
5749 after a process has opened it but before that process has acquired a lock,
5750 there is the potential for a second process to recreate the file and also
5751 acquire a lock. This could lead to two Exim processes writing to the file
5752 at the same time. The added code performs the same test as UW imapd; it
5753 checks after acquiring the lock that its file descriptor still refers to
5754 the same named file.
5756 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5757 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5760 11. Added the smtp_active_hostname option. If used, this will typically be made
5761 to depend on the incoming interface address. Because $interface_address is
5762 not set up until the daemon has forked a reception process, error responses
5763 that can happen earlier (such as "too many connections") no longer contain
5766 12. If an expansion in a condition on a "warn" statement fails because a lookup
5767 defers, the "warn" statement is abandoned, and the next ACL statement is
5768 processed. Previously this caused the whole ACL to be aborted.
5770 13. Added the iplsearch lookup type.
5772 14. Added ident_timeout as a log selector.
5774 15. Added tls_certificate_verified as a log selector.
5776 16. Added a global option tls_require_ciphers (compare the smtp transport
5777 option of the same name). This controls incoming TLS connections.
5779 17. I finally figured out how to make tls_require_ciphers do a similar thing
5780 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5781 before starting the TLS session.
5783 18. Tabs are now shown as \t in -bP output.
5785 19. If the log selector return_path_on_delivery was set, Exim crashed when
5786 bouncing a message because it had too many Received: header lines.
5788 20. If two routers both had headers_remove settings, and the first one included
5789 a superfluous trailing colon, the final name in the first list and the
5790 first name in the second list were incorrectly joined into one item (with a
5791 colon in the middle).
5797 1. Added -C and -D options to the exinext utility, mainly to make it easier
5798 to include in the automated testing, but these could be helpful when
5799 multiple configurations are in use.
5801 2. The exinext utility was not formatting the output nicely when there was
5802 an alternate port involved in the retry record key, nor when there was a
5803 message id as well (for retries that were specific to a specific message
5804 and a specific host). It was also confused by IPv6 addresses, because of
5805 the additional colons they contain. I have fixed the IPv4 problem, and
5806 patched it up to do a reasonable job for IPv6.
5808 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5809 delivery, the log line now contains "pipelined" if PIPELINING was used.
5811 4. An SMTP transport process used to panic and die if the bind() call to set
5812 an explicit outgoing interface failed. This has been changed; it is now
5813 treated in the same way as a connect() failure.
5815 5. A reference to $sender_host_name in the part of a conditional expansion
5816 that was being skipped was still causing a DNS lookup. This no longer
5819 6. The def: expansion condition was not recognizing references to header lines
5820 that used bh_ and bheader_.
5822 7. Added the _cache feature to named lists.
5824 8. The code for checking quota_filecount in the appendfile transport was
5825 allowing one more file than it should have been.
5827 9. For compatibility with Sendmail, the command line option
5835 and sets the incoming protocol and host name (for trusted callers). The
5836 host name and its colon can be omitted when only the protocol is to be set.
5837 Note the Exim already has two private options, -pd and -ps, that refer to
5838 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5839 "s", but I don't think that's a major issue.
5841 10. A number of refactoring changes to the code, none of which should affect
5844 (a) The number of logging options was getting close to filling up the
5845 32-bit word that was used as a bit map. I have split them into two classes:
5846 those that are passed in the argument to log_write(), and those that are
5847 only ever tested independently outside of that function. These are now in
5848 separate 32-bit words, so there is plenty of room for expansion again.
5849 There is no change in the user interface or the logging behaviour.
5851 (b) When building, for example, log lines, the code previously used a
5852 macro that called string_cat() twice, in order to add two strings. This is
5853 not really sufficiently general. Furthermore, there was one instance where
5854 it was actually wrong because one of the argument was used twice, and in
5855 one call a function was used. (As it happened, calling the function twice
5856 did not affect the overall behaviour.) The macro has been replaced by a
5857 function that can join an arbitrary number of extra strings onto a growing
5860 (c) The code for expansion conditions now uses a table and a binary chop
5861 instead of a serial search (which was left over from when there were very
5862 few conditions). Also, it now recognizes conditions like "pam" even when
5863 the relevant support is not compiled in: a suitably worded error message is
5864 given if an attempt is made to use such a condition.
5866 11. Added ${time_interval:xxxxx}.
5868 12. A bug was causing one of the ddress fields not to be passed back correctly
5869 from remote delivery subprocesses. The field in question was not being
5870 subsequently used, so this caused to problems in practice.
5872 13. Added new log selectors queue_time and deliver_time.
5874 14. Might have fixed a bug in maildirsizefile handling that threw up
5875 "unexpected character" debug warnings, and recalculated the data
5876 unnecessarily. In any case, I expanded the warning message to give more
5879 15. Added the message "Restricted characters in address" to the statements in
5880 the default ACL that block characters like @ and % in local parts.
5882 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5883 Three changes have been made:
5885 (a) There was a serious bug; a negative response to MAIL caused the whole
5886 recipient domain to be cached as invalid, thereby blocking all messages
5887 to all local parts at the same domain, from all senders. This bug has
5888 been fixed. The domain is no longer cached after a negative response to
5889 MAIL if the sender used is not empty.
5891 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5894 (c) A new callout option, "use_sender" has been added for people who want
5895 the modified behaviour.
5901 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5904 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5905 indeed breaks things for older releases.
5907 3. Added additional logging to the case where there is a problem reading data
5908 from a filter that is running in a subprocess using a pipe, in order to
5909 try to track down a specific problem.
5911 4. Testing facility fudge: when running in the test harness and attempting
5912 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5913 getting "No route to host". Convert this to a timeout.
5915 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5918 6. Some OS don't have socklen_t but use size_t instead. This affects the
5919 fifth argument of getsockopt() amongst other things. This is now
5920 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5921 can be set for individual OS. I have set it for SunOS5, OSF1, and
5922 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5923 some earlier ones do not.
5925 7. Change 4.30/15 was not doing the test caselessly.
5927 8. The standard form for an IPv6 address literal was being rejected by address
5928 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5929 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5930 this, as well as the form without the "IPv6" on the front (but only when
5931 address literals are enabled, of course).
5933 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5935 10. Exim crashed if a message with an empty sender address specified by -f
5936 encountered a router with an errors_to setting. This could be provoked only
5937 by a command such as
5941 where an empty string was supplied; "<>" did not hit this bug.
5943 11. Installed PCRE release 4.5.
5945 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5946 remained set. It is now erased.
5948 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5949 times from message ids (which are base 36 rather than the normal 62).
5951 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5952 were being counted as actual protocol errors, and logged if the log
5953 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5954 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5955 and DATA following a set of rejected RCPTs do not count as protocol errors.
5956 In other words, Exim assumes they were pipelined, though this may not
5957 actually be the case. Of course, in all cases the client gets an
5958 appropriate error code.
5960 15. If a lookup fails in an ACL condition, a message about the failure may
5961 be available; it is used if testing the ACL cannot continue, because most
5962 such messages specify what the cause of the deferral is. However, some
5963 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5964 that caused an old message to be retained and used if a later statement
5965 caused a defer, replacing the real cause of the deferral.
5967 16. If an IP address had so many PTR records that the DNS lookup buffer
5968 was not large enough to hold them, Exim could crash while trying to process
5969 the truncated data. It now detects and logs this case.
5971 17. Further to 4.21/58, another change has been made: if (and only if) the
5972 first line of a message (the first header line) ends with CRLF, a bare LF
5973 in a subsequent header line has a space inserted after it, so as not to
5974 terminate the header.
5976 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5977 unnecessarily, used atoi() instead of strtol(), and didn't check the
5978 termination when getting file sizes from file names by regex.
5980 19. Completely re-implemented the support for maildirsize files, in the light
5981 of a number of problems with the previous contributed implementation
5982 (4.30/29). In particular:
5984 . If the quota is zero, the maildirsize file is maintained, but no quota is
5987 . If the maildir directory does not exist, it is created before any attempt
5988 to write a maildirsize file.
5990 . The quota value in the file is just a cache; if the quota is changed in
5991 the transport, the new value overrides.
5993 . A regular expression is available for excluding directories from the
5996 20. The autoreply transport checks the characters in options that define the
5997 message's headers; it allows continued headers, but it was checking with
5998 isspace() after an embedded newline instead of explicitly looking for a
6001 21. If all the "regular" hosts to which an address was routed had passed their
6002 expiry times, and had not reached their retry times, the address was
6003 bounced, even if fallback hosts were defined. Now Exim should go on to try
6006 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6007 equivalent code in the SMTP transport. Some hosts send humungous responses
6008 to HELO/EHLO, more than 1024 it seems.
6010 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6011 gives compiler warnings in some environments. I've now done it "properly",
6014 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6015 (because of problems with the built-in one) was declared to return uschar *
6016 instead of char *, causing compiler failure.
6018 25. Fixed a file descriptor leak when processing alias/forward files.
6020 26. Fixed a minor format string issue in dbfn.c.
6022 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6024 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6025 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6028 29. When a sender address is verified, it is cached, to save repeating the test
6029 when there is more than one recipient in a message. However, when the
6030 verification involves a callout, it is possible for different callout
6031 options to be set for different recipients. It is too complicated to keep
6032 track of this in the cache, so now Exim always runs a verification when a
6033 callout is required, relying on the callout cache for the optimization.
6034 The overhead is duplication of the address routing, but this should not be
6037 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6038 to be verified with callout=postmaster, and the main callout worked but the
6039 postmaster check failed, the verification correctly failed. However, if a
6040 subsequent RCPT command asked for sender verification *without* the
6041 postmaster check, incorrect caching caused this verification also to fail,
6044 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6045 it was not caching the DNS options (qualify_single, search_parents) that
6046 were used when the lookup failed. A subsequent lookup with different
6047 options therefore always gave the same answer, though there were cases
6048 where it should not have. (Example: a "domains = !$mx_any" option on a
6049 dnslookup router: the "domains" option is always processed without any
6050 widening, but the router might have qualify_single set.) Now Exim uses the
6051 cached value only when the same options are set.
6053 32. Added John Jetmore's "exipick" utility to the distribution.
6055 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6056 than a timeout (e.g. a certificate is required, and is not provided), an
6057 Exim server now closes the connection immediately. Previously it waited for
6058 the client to close - but if the client is SSL, it seems that they each
6059 wait for each other, leading to a delay before one of them times out.
6061 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6062 maintained 0.8.x compatibility because I don't think many are using it, and
6063 it is clearly obsolete.
6065 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6068 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6069 client certificate was expired. A simple patch fixes this, though I don't
6070 understand the full logic of why the verify callback is called multiple
6073 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6074 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6075 which causes problems with some clients (such as the Certicom SSL Plus
6076 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6077 disables the coutermeasure allowing Eudora to connect."
6079 38. Exim was not checking that a write() to a log file succeeded. This could
6080 lead to Bad Things if a log got too big, in particular if it hit a file
6081 size limit. Exim now panics and dies if it cannot write to a log file, just
6082 as it does if it cannot open a log file.
6084 39. Modified OS/Makefile-Linux so that it now contains
6086 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6088 The two -D definitions ensure that Exim is compiled with large file
6089 support, which makes it possible to handle log files that are bigger than
6092 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6093 instance) a domain was checked against a named list that involved a lookup,
6094 causing $domain_data to be set, then another domain was checked against the
6095 same list, then the first domain was re-checked, the value of $domain_data
6096 after the final check could be wrong. In particular, if the second check
6097 failed, it could be set empty. This bug probably also applied to
6100 41. The strip_trailing_dot option was not being applied to the address given
6101 with the -f command-line option.
6103 42. The code for reading a message's header from the spool was incrementing
6104 $received_count, but never initializing it. This meant that the value was
6105 incorrect (doubled) while delivering a message in the same process in which
6106 it was received. In the most common configuration of Exim, this never
6107 happens - a fresh exec is done - but it can happen when
6108 deliver_drop_privilege is set.
6110 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6111 it now includes up to 150 characters of the unexpected data in the log
6114 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6115 and building data strings. The size of both of these buffers was 10 000
6116 bytes - far larger than anybody would *ever* want, thought I. Needless to
6117 say, somebody hit the limit. I have increased the maximum line length to
6118 20 000 and the maximum data length of concatenated lines to 100 000. I have
6119 also fixed two bugs, because there was no checking on these buffers. Tsk,
6120 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6121 buffer is too small.
6123 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6124 lsearch lookups. Now it does.
6126 46. When parsing a route_list item in a manualroute router, a fixed-length
6127 buffer was used for the list of hosts. I made this 1024 bytes long,
6128 thinking that nobody would ever have a list of hosts that long. Wrong.
6129 Somebody had a whole pile of complicated expansion conditions, and the
6130 string was silently truncated, leading to an expansion error. It turns out
6131 that it is easier to change to an unlimited length (owing to other changes
6132 that have happened since this code was originally written) than to build
6133 structure for giving a limitation error. The length of the item that
6134 expands into the list of hosts is now unlimited.
6136 47. The lsearch lookup could not handle data where the length of text line was
6137 more than 4095 characters. Such lines were truncated, leading to shortened
6138 data being returned. It should now handle lines of any length.
6140 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6141 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6144 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6145 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6146 no longer happens in comment lines. A list of replacements is now placed
6147 at the head of all of the source files, except those whose only change is
6148 to replace PERL_COMMAND in the very first #! line.
6150 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6151 messages on the queue, with a bottom-up merge sort, using code contributed
6152 by Michael Haardt. This should make operations like -bp somewhat faster on
6153 large queues. It won't affect queue runners, except when queue_run_in_order
6156 51. Installed eximstats 1.31 in the distribution.
6158 52. Added support for SRV lookups to the dnslookup router.
6160 53. If an ACL referred to $message_body or $message_body_end, the value was not
6161 reset for any messages that followed in the same SMTP session.
6163 54. The store-handling optimization for building very long strings was not
6164 differentiating between the different store pools. I don't think this
6165 actually made any difference in practice, but I've tidied it.
6167 55. While running the routers to verify a sender address, $sender_address
6168 was still set to the sender address. This is wrong, because when routing to
6169 send a bounce to the sender, it would be empty. Therefore, I have changed
6170 it so that, while verifying a sender address, $sender_address is set to <>.
6171 (There is no change to what happens when verifying a recipient address.)
6173 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6174 target A or AAAA records (if not already returned) without resetting the
6175 qualify_single or search_parents options of the DNS resolver. These are
6176 inappropriate in this case because the targets of MX and SRV records must
6177 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6178 target that, when qualified, matched something in the local domain. These
6179 two options are now turned off when doing these lookups.
6181 57. It seems that at least some releases of Reiserfs (which does not have the
6182 concept of a fixed number of inodes) returns zero and not -1 for the
6183 number of available inodes. This interacted badly with check_spool_inodes,
6184 which assumed that -1 was the "no such thing" setting. What I have done is
6185 to check that the total number of inodes is greater than zero before doing
6186 the test of how many are available.
6188 58. When a "warn" ACL statement has a log_message modifier, the message is
6189 remembered, and not repeated. This is to avoid a lot of repetition when a
6190 message has many recipients that cause the same warning to be written.
6191 However, Exim was preserving the list of already written lines for an
6192 entire SMTP session, which doesn't seem right. The memory is now reset if a
6193 new message is started.
6195 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6196 debugging output unless log_rewrite was also set.
6198 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6199 of (int)(handle) when we know that handle contains (void *)(-1).
6201 61. The Exim daemon panic-logs an error return when it closes the incoming
6202 connection. However "connection reset by peer" seems to be common, and
6203 isn't really an error worthy of noting specially, so that particular error
6206 62. When Exim is trying to find all the local interfaces, it used to panic and
6207 die if the ioctl to get the interface flags failed. However, it seems that
6208 on at least one OS (Solaris 9) it is possible to have an interface that is
6209 included in the list of interfaces, but for which you get a failure error
6210 for this call. This happens when the interface is not "plumbed" into a
6211 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6212 failure of the "get flags" call assumes that the interface is down.
6214 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6215 makes life easier for people who are doing arithmetic on fields extracted
6216 from dates, where you often get leading zeros that should not be
6217 interpreted as octal.
6219 64. Added qualify_domain to the redirect router, to override the global
6222 65. If a pathologically long header line contained very many addresses (the
6223 report of this problem mentioned 10 000) and each of them was rewritten,
6224 Exim could use up a very large amount of memory. (It kept on making new
6225 copies of the header line as it rewrote, and never released the old ones.)
6226 At the expense of a bit more processing, the header rewriting function has
6227 been changed so that it no longer eats memory in this way.
6229 66. The generation of the Received: header has been moved from the time that a
6230 message starts to be received, to the time that it finishes. The timestamp
6231 in the Received: header should now be very close to that of the <= log
6232 line. There are two side-effects of this change:
6234 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6235 logged header lines no longer include the local Received: line, because
6236 it has not yet been created. The same applies to a copy of the message
6237 that is returned to a non-SMTP sender when a message is rejected.
6239 (b) When a filter file is tested using -bf, no additional Received: header
6240 is added to the test message. After some thought, I decided that this
6243 This change does not affect the value of $received_for. It is still set
6244 after address rewriting, but before local_scan() is called.
6246 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6248 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6249 gave an unhelpful panic error message, and a defer error. I have managed to
6250 change this behaviour so that it now rejects any supplied certificate,
6251 which seems right, as the list of acceptable certificates is empty.
6253 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6254 gave an unhelpful defer error. I have not managed to make this reject any
6255 supplied certificates, but the error message it gives is "no certificate
6256 supplied", which is not helpful.
6258 70. exigrep's output now also includes lines that are not associated with any
6259 message, but which match the given pattern. Implemented by a patch from
6260 Martin Sluka, which also tidied up the Perl a bit.
6262 71. Recipient callout verification, like sender verification, was using <> in
6263 the MAIL FROM command. This isn't really the right thing, since the actual
6264 sender may affect whether the remote host accepts the recipient or not. I
6265 have changed it to use the actual sender in the callout; this means that
6266 the cache record is now keyed on a recipient/sender pair, not just the
6267 recipient address. There doesn't seem to be a real danger of callout loops,
6268 since a callout by the remote host to check the sender would use <>.
6269 [SEE ABOVE: changed after hitting problems.]
6271 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6272 temporary errors. However, in the case of such a code being given after
6273 the end of a data transmission (i.e. after ".") Exim was failing to write
6274 a retry record for the message. (Yes, there was some broken host that was
6275 actually sending 8xx at this point.)
6277 73. An unknown lookup type in a host list could cause Exim to panic-die when
6278 the list was checked. (An example that provoked this was putting <; in the
6279 middle of a list instead of at the start.) If this happened during a DATA
6280 ACL check, a -D file could be left lying around. This kind of configuration
6281 error no longer causes Exim to die; instead it causes a defer error. The
6282 incident is still logged to the main and panic logs.
6284 74. Buglet left over from Exim 3 conversion. The message "too many messages
6285 in one connection" was written to the rejectlog but not the mainlog, except
6286 when address rewriting (yes!) was being logged.
6288 75. Added write_rejectlog option.
6290 76. When a system filter was run not as root (that is, when system_filter_user
6291 was set), the values of the $n variables were not being returned to the
6292 main process; thus, they were not subsequently available in the $sn
6295 77. Added +return_path_on_delivery log selector.
6297 78. A connection timeout was being treated differently from recipients deferred
6298 when testing hosts_max_try with a message that was older than the host's
6299 retry timeout. (The host should not be counted, thus allowing all hosts to
6300 be tried at least once before bouncing.) This may have been the cause of an
6301 occasionally reported bug whereby a message would remain on the queue
6302 longer than the retry timeout, but would be bounced if a delivery was
6303 forced. I say "may" because I never totally pinned down the problem;
6304 setting up timeout/retry tests is difficult. See also the next item.
6306 79. The ultimate address timeout was not being applied to errors that involved
6307 a combination of host plus message (for example, a timeout on a MAIL
6308 command). When an address resolved to a number of possible hosts, and they
6309 were not all tried for each delivery (e.g. because of hosts_max_try), a
6310 message could remain on the queue longer than the retry timeout.
6312 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6315 81. Fixed an obscure SMTP outgoing bug which required at least the following
6316 conditions: (a) there was another message waiting for the same server;
6317 (b) the server returned 5xx to all RCPT commands in the first message so
6318 that the message was not completed; (c) the server dropped the connection
6319 or gave a negative response to the RSET that Exim sends to abort the
6320 transaction. The observed case was a dropped connection after DATA that had
6321 been sent in pipelining mode. That is, the server had advertised PIPELINING
6322 but was not implementing it correctly. The effect of the bug was incorrect
6323 behaviour, such as trying another host, and this could lead to a crash.
6329 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6330 and daemon.c were passed as pointers to ints; they should have been
6331 pointers to socklen_t variables (which are typically unsigned ints).
6333 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6336 3. Fixed a really odd bug that affected only the testing scheme; patching a
6337 certain fixed string in the binary changed the value of another string that
6338 happened to be identical to the end of the original first string.
6340 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6341 name", it returns that address as the IP address. On some operating
6342 systems (e.g. Solaris), it also passes back the IP address string as the
6343 "host name". However, on others (e.g. Linux), it passes back an empty
6344 string. Exim wasn't checking for this, and was changing the host name to an
6345 empty string, assuming it had been canonicalized.
6347 5. Although rare, it is permitted to have more than one PTR record for a given
6348 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6349 all the names associated with an address, because they do in Solaris.
6350 However, it seems that they do not in Linux for data that comes from the
6351 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6352 I found this out when I moved to a new Linux workstation and tried to run
6353 the Exim test suite.
6355 To get round this problem I have changed the code so that it now does its
6356 own call to the DNS to look up PTR records when searching for a host name.
6357 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6358 addresses that are only in /etc/hosts are still found.
6360 This behaviour is, however, controlled by an option called host_lookup_
6361 order, which defaults to "bydns:byaddr". If people want to use the other
6362 order, or indeed, just use one or the other means of lookup, they can
6363 specify it in this variable.
6365 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6366 some operating systems, this comes back from gethostbyaddr() as an empty
6367 string, and this is what Exim used to test for. However, it seems that in
6368 other systems, "." is yielded. Exim now tests for this case too.
6370 7. The values of check_spool_space and check_log_space are now held internally
6371 as a number of kilobytes instead of an absolute number of bytes. If a
6372 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6373 kilobyte. This means that much larger values can be stored.
6375 8. Exim monitor: an attempt to get the action menu when not actually pointing
6376 at a message produces an empty menu entitled "No message selected". This
6377 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6378 no entries in it ("Shell widget menu has zero width and/or height"). So I
6379 have added a single, blank menu entry in this case.
6381 9. Added ${quote_local_part.
6383 10. MIME decoding is now applied to the contents of Subject: header lines when
6386 11. Now that a reference to $sender_host_address automatically causes a reverse
6387 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6388 host lookup before query-style lookups in lists that might use this
6389 variable. This has therefore been abolished, and the "net-" prefix is no
6390 longer necessary for query-style lookups.
6392 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6393 have been a typo for LFLAGS, so it has been changed.
6395 13. The install script calls Exim with "-C /dev/null" in order to find the
6396 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6397 to be output. However, since Exim outputs its version number before the
6398 error, it didn't break the script. It just looked ugly. I fixed this by
6399 always allowing "-C /dev/null" if the caller is root.
6401 14. Ignore overlarge ACL variable number when reading spool file - insurance
6402 against a later release with more variables having written the file.
6404 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6405 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6406 this, as well as the form without the "IPv6" on the front.
6408 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6409 OS/Makefile-Darwin file.
6411 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6412 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6414 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6415 variable. However, some OS use ints for inodes. I've added cast to long int
6416 to get rid of the compiler warning.
6418 19. I had forgotten to lock out "/../" in configuration file names when
6419 ALT_CONFIG_PREFIX was set.
6421 20. Routers used for verification do not need to specify transports. However,
6422 if such a router generated a host list, and callout was configured, Exim
6423 crashed, because it could not find a port number from the (non-existent)
6424 transport. It now assumes port 25 in this circumstance.
6426 21. Added the -t option to exigrep.
6428 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6429 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6430 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6431 them is set, LOOKUP_LSEARCH is forced.
6433 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6435 24. Added sender and host information to the "rejected by local_scan()" log
6436 line; previously there was no indication of these.
6438 25. Added .include_if_exists.
6440 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6441 Linux. It turns out that not all file systems support this. Apparently some
6442 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6443 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6444 not supported on the file descriptor, is now ignored when Exim is trying to
6445 sync a directory. This applies only to Linux.
6447 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6449 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6450 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6453 29. Added support for maildirsize files from supplied patch (modified a bit).
6455 30. The use of :fail: followed by an empty string could lead Exim to respond to
6456 sender verification failures with (e.g.):
6458 550 Verification failed for <xxx>
6459 550 Sender verify failed
6461 where the first response line was missing the '-' that indicates it is not
6462 the final line of the response.
6464 31. The loop for finding the name of the user that called Exim had a hardwired
6465 limit of 10; it now uses the value of finduser_retries, which is used for
6466 all other user lookups.
6468 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6471 33. Exim was neglecting to zero errno before one call of strtol() when
6472 expanding a string and expecting an integer value. On some systems this
6473 resulted in spurious "integer overflow" errors. Also, it was casting the
6474 result into an int without checking.
6476 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6477 did not work. The code looks as if it has *never* worked, though it appears
6478 to have been documented since at least release 1.62. I have made it work.
6480 35. The "timeout_DNS" error in retry rules, also documented since at least
6481 1.62, also never worked. As it isn't clear exactly what this means, and
6482 clearly it isn't a major issue, I have abolished the feature by treating it
6483 as "timeout", and writing a warning to the main and panic logs.
6485 36. The display of retry rules for -brt wasn't always showing the error code
6488 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6489 timeout_connect_A, timeout_connect_MX.
6491 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6492 to the empty sender.
6494 39. The daemon was not analysing the content of -oX till after it had closed
6495 stderr and disconnected from the controlling terminal. This meant that any
6496 syntax errors were only noted on the panic log, and the return code from
6497 the command was 0. By re-arranging the code a little, I've made the
6498 decoding happen first, so such errors now appear on stderr, and the return
6499 code is 1. However, the actual setting up of the sockets still happens in
6500 the disconnected process, so errors there are still only recorded on the
6503 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6504 connections (as happens on some IP stacks) was logged at start up time as
6505 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6506 it from "IPv6 and IPv4", which means that two separate sockets are being
6509 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6510 says whether AF_INET or AF_INET6 was passed as an argument.
6512 42. Exiwhat output was messed up when time zones were included in log
6515 43. Exiwhat now gives more information about the daemon's listening ports,
6516 and whether -tls-on-connect was used.
6518 44. The "port" option of the smtp transport is now expanded.
6520 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6521 silently ignored. Now an error message is written to the main and panic
6524 46. There's a new ACL modifier called "logwrite" which writes to a log file
6525 as soon as it is encountered.
6527 47. Added $local_user_uid and $local_user_gid at routing time.
6529 48. Exim crashed when trying to verify a sender address that was being
6532 49. Exim was recognizing only a space character after ".include". It now also
6533 recognizes a tab character.
6535 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6536 extracting the relevant information from the specification. The man page no
6537 longer contains scrambled data for the -d option, and I've added a section
6538 at the front about calling Exim under different names.
6540 51. Added "extra_headers" argument to the "mail" command in filter files.
6542 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6545 53. Installed eximstats 1.29.
6547 54. Added transport_filter_timeout as a generic transport option.
6549 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6550 Cc: header lines. This was required by RFC 822, but it not required by RFC
6553 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6554 incoming messages that did not have them. Now it does so only if the
6555 message originates locally, that is, if there is no associated remote host
6556 address. When Resent- header lines are present, this applies to the Resent-
6557 lines rather than the non-Resent- lines.
6559 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6560 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6562 58. Messages for configuration errors now include the name of the main
6563 configuration file - useful now that there may be more than one file in a
6564 list (.included file names were always shown).
6566 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6567 for those rare installations that do not start the daemon as root or run it
6568 setuid root. I've cut out the call to initgroups() if the daemon is not
6571 60. The Exim user and group can now be bound into the binary as text strings
6572 that are looked up at the start of Exim's processing.
6574 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6576 62. Added $mailstore_basename variable.
6578 63. Installed patch to sieve.c from Michael Haardt.
6580 64. When Exim failed to open the panic log after failing to open the main log,
6581 the original message it was trying to log was written to stderr and debug
6582 output, but if they were not available (the usual case in production), it
6583 was lost. Now it is written to syslog before the two lines that record the
6584 failures to open the logs.
6586 65. Users' Exim filters run in subprocesses under the user's uid. It is
6587 possible for a "deliver" command or an alias in a "personal" command to
6588 provoke an address rewrite. If logging of address rewriting is configured,
6589 this fails because the process is not running as root or exim. There may be
6590 a better way of dealing with this, but for the moment (because 4.30 needs
6591 to be released), I have disabled address rewrite logging when running a
6592 filter in a non-root, non-exim process.
6598 1. The buildconfig auxiliary program wasn't quoting the value set for
6599 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6600 not defined. This bug was masked in 4.22 by the effect that was fixed in
6603 2. Some messages that were rejected after a message id was allocated were
6604 shown as "incomplete" by exigrep. It no longer does this for messages that
6605 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6607 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6608 have allow_domain_literals set, the ID did not get logged in the <= line.
6609 Domain literals are now always recognized in Message-ID: header lines.
6611 4. The first argument for a ${extract expansion item is the key name or field
6612 number. Leading and trailing spaces in this item were not being ignored,
6613 causing some misleading effects.
6615 5. When deliver_drop_privilege was set, single queue runner processes started
6616 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6617 same command in the process it spins off) were not dropping privilege.
6619 6. When the daemon running as "exim" started a queue runner, it always
6620 re-executed Exim in the spun-off process. This is a waste of effort when
6621 deliver_drop_privilege is set. The new process now just calls the
6622 queue-runner function directly.
6628 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6631 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6632 ignored. Though the use of -oP was forcing the writing of a pid file, it
6633 was always written to the default place.
6635 3. If the message "no IP address found for host xxxx" is generated during
6636 incoming verification, it is now followed by identification of the incoming
6637 connection (so you can more easily find what provoked it).
6639 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6641 5. Added some features to "harden" Exim a bit more against certain attacks:
6643 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6644 be put in Local/Makefile. This is like the never_users runtime option,
6645 but it cannot be overridden. The default setting is "root".
6647 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6648 prefix string with which any file named in a -C command line option
6651 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6652 is retained for -C and -D only if the caller of Exim is root. Without
6653 it, the exim user may also use -C and -D and retain privilege.
6655 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6656 command line option is disabled.
6658 6. Macro names set by the -D option must start with an upper case letter, just
6659 like macro names defined in the configuration file.
6661 7. Added "dereference=" facility to LDAP.
6663 8. Two instances of the typo "uknown" in the source files are fixed.
6665 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6666 the Configure-Makefile script screwed up while processing it.
6668 10. Incorporated PCRE 4.4.
6670 11. The SMTP synchronization check was not operating right at the start of an
6671 SMTP session. For example, it could not catch a HELO sent before the client
6672 waited for the greeting. There is now a check for outstanding input at the
6673 point when the greeting is written. Because of the duplex, asynchronous
6674 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6675 way, but not yet received, when the check is performed.
6677 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6678 on TCP/IP sockets, because this apparently causes some broken clients to
6681 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6682 unchanged) from the Cygwin maintainer.
6684 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6685 was supported instead of testing for mbx. Effectively a typo.
6687 15. The spa authenticator server code was not checking that the input it
6688 received was valid base64.
6690 16. The debug output line for the "set" modifier in ACLs was not showing the
6691 name of the variable that was being set.
6693 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6695 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6696 lookup on its own. Something else was needed to trigger the lookup. For
6697 example, a match in host_lookup or the need for a host name in a host list.
6698 Now, if $sender_host_name is referenced and the host name has not yet been
6699 looked up, a lookup is performed. If the lookup fails, the variable remains
6700 empty, and $host_lookup_failed is set to "1".
6702 19. Added "eqi" as a case-independent comparison operator.
6704 20. The saslauthd authentication condition could segfault if neither service
6705 nor realm was specified.
6707 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6708 error message that was logged was misleading, and incoming SMTP
6709 connections were dropped. The message is now more accurate, and temporary
6710 errors are given to SMTP connections.
6712 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6713 (except RSET) with a 421 error, until QUIT is received. However, it was
6714 failing to send a response to QUIT.
6716 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6717 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6718 ACL is now run after the helo_try_verify_hosts code.
6720 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6721 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6722 case-independent, so other case variants are also recognized). Apparently
6723 some systems use these upper case variants.
6725 25. If more than two messages were waiting for the same host, and a transport
6726 filter was specified for the transport, Exim sent two messages over the
6727 same TCP/IP connection, and then failed with "socket operation on non-
6728 socket" when it tried to send the third.
6730 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6732 27. The extern definition of crypt16() in expand.c was not being excluded when
6733 the OS had its own crypt16() function.
6735 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6736 as a preferred synonym for return_size_limit, both as an option and as an
6739 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6741 30. Changed the default configuration ACL to relax the local part checking rule
6742 for addresses that are not in any local domains. For these addresses,
6743 slashes and pipe symbols are allowed within local parts, but the sequence
6744 /../ is explicitly forbidden.
6746 31. SPA server authentication was not clearing the challenge buffer before
6749 32. log_message in a "warn" ACL statement was writing to the reject log as
6750 well as to the main log, which contradicts the documentation and doesn't
6751 seem right (because no rejection is happening). So I have stopped it.
6753 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6754 However, I am unable to do any testing of this.
6756 34. Fixed an infelicity in the appendfile transport. When checking directories
6757 for a mailbox, to see if any needed to be created, it was accidentally
6758 using path names with one or more superfluous leading slashes; tracing
6759 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6761 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6762 contents are added to the log line that is written for every discarded
6763 recipient. (Previously a log_message setting was ignored.)
6765 36. The ${quote: operator now quotes the string if it is empty.
6767 37. The install script runs exim in order to find its version number. If for
6768 some reason other than non-existence or emptiness, which it checks, it
6769 could not run './exim', it was installing it with an empty version number,
6770 i.e. as "exim-". This error state is now caught, and the installation is
6773 38. An argument was missing from the function that creates an error message
6774 when Exim fails to connect to the socket for saslauthd authentication.
6775 This could cause Exim to crash, or give a corrupted message.
6777 39. Added isip, isip4, and isip6 to ${if conditions.
6779 40. The ACL variables $acl_xx are now saved with the message, and can be
6780 accessed later in routers, transports, and filters.
6782 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6783 strings in the file are not string-expanded.
6785 42. If a MAIL command specified a SIZE value that was too large to fit into an
6786 int variable, the check against message_size_limit failed. Such values are
6787 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6788 day this will have to be increased, but I don't think I want to be around
6789 when emails are that large.
6796 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6797 iconv() is not standard in FreeBSD.
6799 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6800 IPv6 enabled. The observed symptom was a segmentation fault on return from
6801 the function os_common_find_running_interfaces() in src/os.c.
6803 3. In the check_special_case() function in daemon.c I had used "errno" as an
6804 argument name, which causes warnings on some systems. This was basically a
6805 typo, since it was named "eno" in the comments!
6807 4. The code that waits for the clock to tick (at a resolution of some fraction
6808 of a second) so as to ensure message-id uniqueness was always waiting for
6809 at least one whole tick, when it could have waited for less. [This is
6810 almost certainly not relevant at current processor speeds, where it is
6811 unlikely to ever wait at all. But we try to future-proof.]
6813 5. The function that sleeps for a time interval that includes fractions of a
6814 second contained a race. It did not block SIGALRM between setting the
6815 timer, and suspending (a couple of lines later). If the interval was short
6816 and the sigsuspend() was delayed until after it had expired, the suspension
6817 never ended. On busy systems this could lead to processes getting stuck for
6820 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6821 process, before it forks any delivery processes. The open lookup caching
6822 mechanism meant that the open file or database connection was passed into
6823 the delivery process. The problem was that delivery processes always tidy
6824 up cached lookup data. This could cause a problem for the next delivery
6825 process started by the queue runner, because the external queue runner
6826 process does not know about the closure. So the next delivery process
6827 still has data in the lookup cache. In the case of a file lookup, there was
6828 no problem because closing a file descriptor in a subprocess doesn't affect
6829 the parent. However, if the lookup was caching a connection to a database,
6830 the connection was closed, and the second delivery process was likely to
6831 see errors such as "PGSQL: query failed: server closed the connection
6832 unexpectedly". The problem has been fixed by closing all cached lookups
6833 in a queue runner before running a delivery process.
6835 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6836 seem to have the "const" qualifier which it has on other OS. I've
6839 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6840 *of the same type* (client or server) with the same public name that an
6841 error should be diagnosed.
6843 9. When Exim looked up a host name for an IP address, but failed to find the
6844 original IP address when looking up the host name (a safety check), it
6845 output the message "<ip address> does not match any IP for NULL", which was
6846 confusing, to say the least. The bug was that the host name should have
6847 appeared instead of "NULL".
6849 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6850 user that is built into the binary, and the -C or -D options is used, root
6851 privilege is dropped before the configuration file is read. In addition,
6852 logging is switched to stderr instead of the normal log files. If the
6853 configuration then re-defines the Exim user, the unprivileged environment
6854 is probably not what is expected, so Exim logs a panic warning message (but
6857 However, if deliver_drop_privilege is set, the unprivileged state may well
6858 be exactly what is intended, so the warning has been cut out in that case,
6859 and Exim is allowed to try to write to its normal log files.
6865 1. smtp_return_error_details was not giving details for temporary sender
6866 or receiver verification errors.
6868 2. Diagnose a configuration error if two authenticators have the same public
6871 3. Exim used not to create the message log file for a message until the first
6872 delivery attempt. This could be confusing when incoming messages were held
6873 for policy or load reasons. The message log file is now created at the time
6874 the message is received, and an initial "Received" line is written to it.
6876 4. The automatically generated man page for command line options had a minor
6877 bug that caused no ill effects; however, a more serious problem was that
6878 the procedure for building the man page automatically didn't always
6879 operate. Consequently, release 4.20 contains an out-of-date version. This
6880 shouldn't happen again.
6882 5. When building Exim with embedded Perl support, the script that builds the
6883 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6884 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6886 6. The freeze_tell option was not being used for messages that were frozen on
6887 arrival, either by an ACL or by local_scan().
6889 7. Added the smtp_incomplete_transaction log selector.
6891 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6892 was accepting AUTH without a new EHLO.
6894 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6895 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6898 10. Logging of TCP/IP connections (when configured) now happens in the main
6899 daemon process instead of the child process, so that the TCP/IP connection
6900 count is more accurate (but it can never be perfect).
6902 11. The use of "drop" in a nested ACL was not being handled correctly in the
6903 outer ACL. Now, if condition failure induced by the nested "drop" causes
6904 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6905 or "require"), the connection is dropped.
6907 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6908 that yield "discard" can now be used with an "accept" or a "discard" verb,
6909 but an error is generated for any others (because I can't see a useful way
6910 to define what should happen).
6912 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6913 are now processed in the same way as lines in the Exim configuration file.
6914 In particular, continuation lines are supported.
6916 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6918 15. Added -ti meaning -t -i.
6920 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6921 domains, and warn by logging if others are found.
6923 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6924 returned from GIFCONF when Exim is trying to find the list of interfaces on
6925 a host. The code in os.c has been modified to copy each ifreq to an aligned
6926 structure in all cases.
6928 Also, in some cases, the returned ifreq's were being copied to a 'struct
6929 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6930 means the last couple of bytes of an IPv6 address could be chopped if the
6931 ifreq contained only a normal sockaddr (14 bytes storage).
6933 18. Named domain lists were not supported in the hosts_treat_as_local option.
6934 An entry such as +xxxx was not recognized, and was treated as a literal
6937 19. Ensure that header lines added by a DATA ACL are included in the reject log
6938 if the ACL subsequently rejects the message.
6940 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6941 MD5 (which is deprecated).
6943 21. When testing a filter file using -bf, Exim was writing a message when it
6944 took the sender from a "From " line in the message, but it was not doing so
6945 when it took $return_path from a Return-Path: header line. It now does.
6947 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6948 with a valid header line field name (a series of printing characters
6949 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6951 23. Changed "disc" in the source to "disk" to conform to the documentation and
6952 the book and for uniformity.
6954 24. Ignore Sendmail's -Ooption=value command line item.
6956 25. When execve() failed while trying to run a command in a pipe transport,
6957 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6958 could be confused with a return value of 69 from the command itself. This
6959 has been changed to 127, the value the shell returns if it is asked to run
6960 a non-existent command. The wording for the related log line suggests a
6961 non-existent command as the problem.
6963 26. If received_header_text expands to an empty string, do not add a Received:
6964 header line to the message. (Well, it adds a token one on the spool, but
6965 marks it "old" so that it doesn't get used or transmitted.)
6967 27. Installed eximstats 1.28 (addition of -nt option).
6969 28. There was no check for failure on the call to getsockname() in the daemon
6970 code. This can fail if there is a shortage of resources on the system, with
6971 ENOMEM, for example. A temporary error is now given on failure.
6973 29. Contrary to the C standard, it seems that in some environments, the
6974 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6975 program. Exim now does this explicitly; it affects the formatting of
6976 timestamps using strftime().
6978 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6979 complaints. I've now initialized all the variables, to avoid this.
6981 32. Header lines added by a system filter were not being "seen" during
6982 transport-time rewrites.
6984 33. The info_callback() function passed to OpenSSL is set up with type void
6985 (*)(SSL *, int, int), as described somewhere. However, when calling the
6986 function (actually a macro) that sets it up, the type void(*)() is
6987 expected. I've put in a cast to prevent warnings from picky compilers.
6989 34. If a DNS black list lookup found a CNAME record, but there were no A
6990 records associated with the domain it pointed at, Exim crashed.
6992 35. If a DNS black list lookup returned more than one A record, Exim ignored
6993 all but the first. It now scans all returned addresses if a particular IP
6994 value is being sought. In this situation, the contents of the
6995 $dnslist_value variable are a list of all the addresses, separated by a
6998 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6999 to accept a host name and all its aliases if the forward lookup for any of
7000 them yielded the IP address of the incoming connection. Now it accepts only
7001 those names whose forward lookup yields the correct IP address. Any other
7002 names are discarded. This closes a loophole whereby a rogue DNS
7003 administrator could create reverse DNS records to break through a
7004 wildcarded host restriction in an ACL.
7006 37. If a user filter or a system filter that ran in a subprocess used any of
7007 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7008 the wrong values were passed to the pipe command ($thisaddress had the
7009 value of $0, $0 had the value of $1, etc). This bug was introduced by
7010 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7012 38. Improved the line breaking for long SMTP error messages from ACLs.
7013 Previously, if there was no break point between 40 and 75 characters, Exim
7014 left the rest of the message alone. Two changes have been made: (a) I've
7015 reduced the minimum length to 35 characters; (b) if it can't find a break
7016 point between 35 and 75 characters, it looks ahead and uses the first one
7017 that it finds. This may give the occasional overlong line, but at least the
7018 remaining text gets split now.
7020 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7021 file descriptors might be low, and that setting 1000 would always raise it.
7022 It turns out that in some environments, the limit is already over 1000 and
7023 that lowering it causes trouble. So now Exim takes care not to decrease it.
7025 40. When delivering a message, the value of $return_path is set to $sender_
7026 address at the start of routing (routers may change the value). By an
7027 oversight, this default was not being set up when an address was tested by
7028 -bt or -bv, which affected the outcome if any router or filter referred to
7031 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7032 log only when "message" or "log_message" are set. However, if one of the
7033 conditions was an address verification, or a call to a nested ACL, the
7034 messages generated by the underlying test were being passed through. This
7035 no longer happens. The underlying message is available in $acl_verify_
7036 message for both "message" and "log_message" expansions, so it can be
7037 passed through if needed.
7039 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7040 new expansion $bh_ to give the encoded byte string without charset
7041 translation. Translation happens only if iconv() is available; HAVE_ICONV
7042 indicates this at build time. HEADERS_CHARSET gives the charset to
7043 translate to; headers_charset can change it in the configuration, and
7044 "headers charset" can change it in an individual filter file.
7046 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7047 that creates RFC 2047 encoded "words" labels them as that charset instead
7048 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7049 expansion operator; (ii) when Exim creates a From: line for a local
7050 message; (iii) when a header line is rewritten to include a "phrase" part.
7052 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7053 buggy, causing it to skip the first lines of messages whose message ID
7054 ended in 'D'. This would not have bitten before Exim release 4.14, because
7055 message IDs were unlikely to end in 'D' before then. The effect was to have
7056 incorrect size information for certain domains.
7058 45. #include "config.h" was missing at the start of the crypt16.c module. This
7059 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7062 46. If there was a timeout during a "random" callout check, Exim treated it as
7063 a failure of the random address, and carried on sending RSET and the real
7064 address. If the delay was just some slowness somewhere, the response to the
7065 original RCPT would be taken as a response to RSET and so on, causing
7066 mayhem of various kinds.
7068 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7069 when I implemented it. It didn't allow for the fact that some option values
7070 may legitimately be negative (e.g. size_addition), and it didn't even do
7071 the right test for positive values.
7073 48. Domain names in DNS records are case-independent. Exim always looks them up
7074 in lower case. Some resolvers return domain names in exactly the case they
7075 appear in the zone file, that is, they may contain uppercase letters. Not
7076 all resolvers do this - some return always lower case. Exim was treating a
7077 change of case by a resolver as a change of domain, similar to a widening
7078 of a domain abbreviation. This triggered its re-routing code and so it was
7079 trying to route what was effectively the same domain again. This normally
7080 caused routing to fail (because the router wouldn't handle the domain
7081 twice). Now Exim checks for this case specially, and just changes the
7082 casing of the domain that it ultimately uses when it transmits the message
7085 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7088 50. If a filter generated a file delivery with a non-absolute name (possible if
7089 no home directory exists for the router), the forbid_file option was not
7092 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7093 the existing equality matching.
7095 52. Exim was using ints instead of ino_t variables in some places where it was
7096 dealing with inode numbers.
7098 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7099 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7100 called TMPDIR, and if it finds it is different, it changes its value.
7102 54. The smtp_printf() function is now made available to local_scan() so
7103 additional output lines can be written before returning. There is also an
7104 smtp_fflush() function to enable the detection of a dropped connection.
7105 The variables smtp_input and smtp_batched_input are exported to
7108 55. Changed the default runtime configuration: the message "Unknown user"
7109 has been removed from the ACL, and instead placed on the localuser router,
7110 using the cannot_route_message feature. This means that any verification
7111 failures that generate their own messages won't get overridden. Similarly,
7112 the "Unrouteable address" message that was in the ACL for unverifiable
7113 relay addresses has also been removed.
7115 56. Added hosts_avoid_esmtp to the smtp transport.
7117 57. The exicyclog script was not checking for the esoteric option
7118 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7119 will work only if exicyclog is run under the appropriate euid.
7121 58. Following a discussion on the list, the rules by which Exim recognises line
7122 endings on incoming messages have been changed. The -dropcr and drop_cr
7123 options are now no-ops, retained only for backwards compatibility. The
7124 following line terminators are recognized: LF CRLF CR. However, special
7125 processing applies to CR:
7127 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7128 nor a local message in the state where . is a terminator.
7130 (ii) If a bare CR is encountered in a header line, an extra space is added
7131 after the line terminator so as not to end the header. The reasoning
7132 behind this is that bare CRs in header lines are most likely either
7133 to be mistakes, or people trying to play silly games.
7135 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7136 was being incorrectly given as 18 bytes larger than it should have been.
7137 This is a VOB (very old bug).
7139 60. This may never have affected anything current, but just in case it has:
7140 When the local host is found other than at the start of a list of hosts,
7141 the local host, those with the same MX, and any that follow, are discarded.
7142 When the list in question was part of a longer list of hosts, the following
7143 hosts (not currently being processed) were also being discarded. This no
7144 longer happens. I'm not sure if this situation could ever has previously
7147 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7150 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7151 header. This is something that is recommended in a new Internet Draft, and
7152 is something that is documented as being done by Sendmail. There are two
7153 possible values. For messages generated by the autoreply transport, Exim
7156 Auto-Submitted: auto-replied
7158 whereas for all other generated messages (e.g. bounces) it adds
7160 Auto-Submitted: auto-generated
7162 63. The "personal" condition in filters now includes a test for the
7163 Auto-Submitted: header. If it contains the string "auto-" the message it
7164 not considered personal.
7166 64. Added rcpt_include_affixes as a generic transport option.
7168 65. Added queue_only_override (default true).
7170 66. Added the syslog_duplication option.
7172 67. If what should have been the first header line of a message consisted of
7173 a space followed by a colon, Exim was mis-interpreting it as a header line.
7174 It isn't of course - it is syntactically invalid and should therefore be
7175 treated as the start of the message body. The misbehaviour could have
7176 caused a number of strange effects, including loss of data in subsequent
7177 header lines, and spool format errors.
7179 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7180 client host had authenticated. This control can now be exercised by an ACL
7181 for more flexibility.
7183 69. By default, callouts do not happen when testing with -bh. There is now a
7184 variant, -bhc, which does actually run the callout code, including
7185 consulting and updating the callout cache.
7187 70. Added support for saslauthd authentication, courtesy of Alexander
7190 71. If statvfs() failed on the spool or log directories while checking their
7191 size for availability, Exim confusingly gave the error "space shortage".
7192 Furthermore, in debugging mode it crashed with a floating point exception.
7193 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7194 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7195 serious problem, Exim now writes to the main and panic logs when this
7196 happens, with details of the failure. It then refuses to accept the
7197 incoming message, giving the message "spool directory problem" or "log
7198 directory problem" with a 421 code for SMTP messages.
7200 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7201 0, 1, and 2 exist, because some OS complain for execs without them (see
7202 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7203 descriptors. However, the code omitted to check that the open succeeded,
7204 causing mysterious errors if for some reason the permissions on /dev/null
7205 got screwed. Now Exim writes a message to the main and panic logs, and
7206 bombs out if it can't open /dev/null.
7208 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7209 interact so that it is all more flexible. It is supposed to remain
7210 backwards compatible. Also added extra_local_interfaces.
7212 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7213 to bomb out with an assertion failure - to the client this appears as a
7214 connection drop. This problem occurs in the part of the code that was taken
7215 from the Samba project. Fortunately, the assertion is in a very simple
7216 function, so I have fixed this by reproducing the function inline in the
7217 one place where it is called, and arranging for authentication to fail
7218 instead of killing the process with assert().
7220 75. The SPA client code was not working when the server requested OEM rather
7221 than Unicode encoding.
7223 76. Added code to make require_files with a specific uid setting more usable in
7224 the case where statting the file as root fails - usually a non-root-mounted
7225 NFS file system. When this happens and the failure is EACCES, Exim now
7226 forks a subprocess and does the per-uid checking as the relevant uid.
7228 77. Added process_log_path.
7230 78. If log_file_path was not explicitly set, a setting of check_log_space or
7231 check_log_inodes was ignored.
7233 79. If a space check for the spool or log partitions fails, the incident is now
7234 logged. Of course, in the latter case the data may get lost...
7236 80. Added the %p formatting code to string_format() so that it can be used to
7237 print addresses in debug_print(). Adjusted all the address printing in the
7238 debugging in store.c to use %p rather than %d.
7240 81. There was a concern that a line of code in smtp_in.c could overflow a
7241 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7242 initially expressed, the concern was not well-founded, because trailing
7243 spaces are removed early. However, if the trailing spaces were followed by
7244 a NULL, they did not get removed, so the overflow was possible. Two fixes
7247 (a) I re-wrote the offending code in a cleaner fashion.
7248 (b) If an incoming SMTP command contains a NULL character, it is rejected
7251 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7252 runs initgroups(), so that if the Exim user is in any additional groups,
7253 they will be used during message reception.
7259 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob