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.
88 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
89 When redis_servers is set to a list of > 1 element, and the Redis servers
90 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
91 case of MOVED into a DEFER case instead, thus moving the query onto the
92 next server in the list. For a cluster of N elements, all N servers must
93 be defined in redis_servers.
95 GF/02 Catch and remove uninitialized value warning in exiqsumm
96 Check for existence of @ARGV before looking at $ARGV[0]
98 JH/01 Replace the store_release() internal interface with store_newblock(),
99 which internalises the check required to safely use the old one, plus
100 the allocate and data copy operations duplicated in both (!) of the
101 extant use locations.
103 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
104 modifier. This matches the restriction on the commandline.
106 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
107 Previously only the last row was returned.
109 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
110 we assumed that tags in the header were well-formed, and parsed the
111 element content after inspecting only the first char of the tag.
112 Assumptions at that stage could crash the receive process on malformed
115 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
116 While running the DKIM ACL we operate on the Permanent memory pool so that
117 variables created with "set" persist to the DATA ACL. Also (at any time)
118 DNS lookups that fail create cache records using the Permanent pool. But
119 expansions release any allocations made on the current pool - so a dnsdb
120 lookup expansion done in the DKIM ACL releases the memory used for the
121 DNS negative-cache, and bad things result. Solution is to switch to the
122 Main pool for expansions.
123 While we're in that code, add checks on the DNS cache during store_reset,
124 active in the testsuite.
125 Problem spotted, and debugging aided, by Wolfgang Breyha.
127 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
128 When none of the hosts presented to a transport match an already-open
129 connection, close it and proceed with the list. Previously we would
130 queue the message. Spotted by Lena with Yahoo, probably involving
133 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
134 Previously a spurious "250 OK id=" response was appended to the proper
137 JH/08 The "support for" informational output now, which built with Content
138 Scanning support, has a line for the malware scanner interfaces compiled
139 in. Interface can be individually included or not at build time.
141 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
142 by the template makefile "src/EDITME". The "STREAM" support for an older
143 ClamAV interface method is removed.
145 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
146 rows affected is given instead).
148 JH/11 The runtime Berkeley DB library version is now additionally output by
149 "exim -d -bV". Previously only the compile-time version was shown.
151 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
152 SMTP connection. Previously, when one had more receipients than the
153 first, an abortive onward connection was made. Move to full support for
154 multiple onward connections in sequence, handling cutthrough connection
155 for all multi-message initiating connections.
157 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
158 routers. Previously, a multi-recipient message would fail to match the
159 onward-connection opened for the first recipient, and cause its closure.
161 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
162 a timeout on read on a GnuTLS initiating connection, resulting in the
163 initiating connection being dropped. This mattered most when the callout
164 was marked defer_ok. Fix to keep the two timeout-detection methods
167 JH/15 Relax results from ACL control request to enable cutthrough, in
168 unsupported situations, from error to silently (except under debug)
169 ignoring. This covers use with PRDR, frozen messages, queue-only and
172 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
174 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
175 metadata, resulting in a crash in free().
177 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
178 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
179 Broken also in d185889f4, with init system revamp.
181 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
182 Previously we abruptly closed the connection after reading a malware-
183 found indication; now we go on to read the "scan ok" response line,
186 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
187 ACL. Previously, a crash would result.
189 JH/19 Speed up macro lookups during configuration file read, by skipping non-
190 macro text after a replacement (previously it was only once per line) and
191 by skipping builtin macros when searching for an uppercase lead character.
193 JH/20 DANE support moved from Experimental to mainline. The Makefile control
194 for the build is renamed.
196 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
197 was allocated for every new TLS startup, meaning one per message. Fix
198 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
200 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
201 reported the original. Fix to report (as far as possible) the ACL
202 result replacing the original.
204 JH/23 Fix memory leak during multi-message connections using STARTTLS under
205 OpenSSL. Certificate information is loaded for every new TLS startup,
206 and the resources needed to be freed.
208 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
210 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
213 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
214 DATA response info to the (existing) per-recipient response info for
215 the "C=" log element. It can have useful tracking info from the
216 destination system. Patch from Simon Arlott.
218 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
219 length value. Previously this would segfault.
221 HS/02 Support Avast multiline protoocol, this allows passing flags to
222 newer versions of the scanner.
224 JH/28 Ensure that variables possibly set during message acceptance are marked
225 dead before release of memory in the daemon loop. This stops complaints
226 about them when the debug_store option is enabled. Discovered specifically
227 for sender_rate_period, but applies to a whole set of variables.
228 Do the same for the queue-runner and queue-list loops, for variables set
229 from spool message files. Do the same for the SMTP per-message loop, for
230 certain variables indirectly set in ACL operations.
232 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
233 as a multi-recipient message from a mailinglist manager). The coding had
234 an arbitrary cutoff number of characters while checking for more input;
235 enforced by writing a NUL into the buffer. This corrupted long / fast
236 input. The problem was exposed more widely when more pipelineing of SMTP
237 responses was introduced, and one Exim system was feeding another.
238 The symptom is log complaints of SMTP syntax error (NUL chars) on the
239 receiving system, and refused recipients seen by the sending system
240 (propating to people being dropped from mailing lists).
241 Discovered and pinpointed by David Carter.
243 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
244 replaced by the ${authresults } expansion.
246 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
248 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
249 allows proper process termination in container environments.
251 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
252 Previously the "final dot" had a newline after it; ensure it is CR,LF.
254 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
255 and "err_perm", deprecated since 4.83 when the RFC-defined words
256 "temperror" and "permerror" were introduced.
258 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
259 transport-filters or DKIM-signing. The restriction was lost in the
260 consolidation of verify-callout and delivery SMTP handling.
261 Extend the restriction to also cover ARC-signing.
263 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
264 in defer=pass mode supply a 450 to the initiator. Previously the message
267 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
268 tls_require_ciphers is used as before.
270 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
271 "pass_unscanned". Only tmpfails from the scanner are written to
272 the paniclog, as they may require admin intervention (permission
273 denied, license issues). Other scanner errors (like decompression
274 bombs) do not cause a paniclog entry.
276 JH/36 Fix reinitialisation of DKIM logging variable between messages.
277 Previously it was possible to log spurious information in receive log
280 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
281 triggered odd behaviour from Outlook Express clients.
283 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
286 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
287 since the IETF WG has not yet settled on that versus the original
288 "bare" representation.
290 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
291 Previously the millisecond value corrupted the output.
292 Fix also for syslog_pid=no and log_selector +pid, for which the pid
293 corrupted the output.
299 JH/01 Rework error string handling in TLS interface so that the caller in
300 more cases is responsible for logging. This permits library-sourced
301 string to be attached to addresses during delivery, and collapses
302 pairs of long lines into single ones.
304 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
305 during configuration. Wildcards are allowed and expanded.
307 JH/02 Rework error string handling in DKIM to pass more info back to callers.
308 This permits better logging.
310 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
311 do not close it down and have the child transport start it up again on
312 the passed-on TCP connection. Instead, proxy the child (and any
313 subsequent ones) for TLS via a unix-domain socket channel. Logging is
314 affected: the continued delivery log lines do not have any DNSSEC, TLS
315 Certificate or OCSP information. TLS cipher information is still logged.
317 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
318 identical IP addresses on different listening ports. Will also affect
321 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
322 add noisy ifdef guards to special-case this sillyness.
323 Patch from Bernd Kuhls.
325 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
326 than 255 are no longer allowed.
328 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
329 Disable the session-cache too, which might reduce our load. Since we
330 currrectly use a new context for every connection, both as server and
331 client, there is no benefit for these.
332 GnuTLS appears to not support tickets server-side by default (we don't
333 call gnutls_session_ticket_enable_server()) but client side is enabled
334 by default on recent versions (3.1.3 +) unless the PFS priority string
337 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
338 <https://reproducible-builds.org/specs/source-date-epoch/>.
340 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
341 the check for any unsuccessful recipients did not notice the limit, and
342 erroneously found still-pending ones.
344 JH/08 Pipeline CHUNKING command and data together, on kernels that support
345 MSG_MORE. Only in-clear (not on TLS connections).
347 JH/09 Avoid using a temporary file during transport using dkim. Unless a
348 transport-filter is involved we can buffer the headers in memory for
349 creating the signature, and read the spool data file once for the
350 signature and again for transmission.
352 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
353 4.77 as the kernel support then wasn't solid, having issues in 64bit
354 mode. Now, it's been long enough. Add support for FreeBSD also.
356 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
357 case where the routing stage had gathered several addresses to send to
358 a host before calling the transport for the first, we previously failed
359 to close down TLS in the old transport process before passing the TCP
360 connection to the new process. The new one sent a STARTTLS command
361 which naturally failed, giving a failed delivery and bloating the retry
362 database. Investigation and fix prototype from Wolfgang Breyha.
364 JH/12 Fix check on SMTP command input synchronisation. Previously there were
365 false-negatives in the check that the sender had not preempted a response
366 or prompt from Exim (running as a server), due to that code's lack of
367 awareness of the SMTP input buffering.
369 PP/04 Add commandline_checks_require_admin option.
370 Exim drops privileges sanely, various checks such as -be aren't a
371 security problem, as long as you trust local users with access to their
372 own account. When invoked by services which pass untrusted data to
373 Exim, this might be an issue. Set this option in main configuration
374 AND make fixes to the calling application, such as using `--` to stop
377 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
378 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
379 responses to those, into a single TLS record each way (this usually means
380 a single packet). As a side issue, smtp_enforce_sync now works on TLS
383 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
384 affects you only if you're dancing at the edge of the param size limits.
385 If you are, and this message makes sense to you, then: raise the
386 configured limit or use OpenSSL 1.1. Nothing we can do for older
389 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
390 cmdline element to get the documented default one. Previously it was
393 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
394 are unsuable for non-message-specific reasons - eg. connection timeout,
395 banner-time rejection.
397 JH/16 Fix logging of delivery remote port, when specified by router, under
400 PP/06 Repair manualroute's ability to take options in any order, even if one
401 is the name of a transport.
404 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
406 JH/17 Change the list-building routines interface to use the expanding-string
407 triplet model, for better allocation and copying behaviour.
409 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
410 Previously it was constructed the first time a possibly-matching string
411 was met in the configuration file input during startup; now it is done
414 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
415 compatible one, to avoid the (poorly documented) possibility of a config
416 file in the working directory redirecting the DB files, possibly correpting
417 some existing file. CVE-2017-10140 assigned for BDB.
419 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
420 cache-hot. Previously, although the result was properly cached, the
421 initial verify call returned a defer.
423 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
424 the main verify for receipient in uncached-mode.
426 JH/22 Retire historical build files to an "unsupported" subdir. These are
427 defined as "ones for which we have no current evidence of testing".
429 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
430 if present. Previously it was ignored.
432 JH/24 Start using specified-initialisers in C structure init coding. This is
433 a C99 feature (it's 2017, so now considered safe).
435 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
436 if was a fixed-sized field and bitmask ops via macros; it is now more
439 PP/07 GitHub PR 56: Apply MariaDB build fix.
440 Patch provided by Jaroslav Škarvada.
442 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
443 during Coverity cleanups [4.87 JH/47]
444 Diagnosis and fix provided by Michael Fischer v. Mollard.
446 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
447 the right size to place the terminating semicolon on its own folded
448 line, the header hash was calculated to an incorrect value thanks to
449 the (relaxed) space the fold became.
451 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
452 and confused the parent.
454 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
455 which could crash as a result. This could lead to undeliverable messages.
457 JH/28 Logging: "next input sent too soon" now shows where input was truncated
460 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
461 matters on fast-turnover and PID-randomising systems, which were getting
462 out-of-order delivery.
464 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
465 a possibly-overlapping copy. The symptom was that "Remote host closed
466 connection in response to HELO" was logged instead of the actual 4xx
469 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
470 Previously only that bufferd was discarded, resulting in SYMTP command
473 JH/32 DKIM: when a message has multiple signatures matching an identity given
474 in dkim_verify_signers, run the dkim acl once for each. Previously only
475 one run was done. Bug 2189.
477 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
478 "panic the current process" to "deliberately defer". The panic log is
479 still written with the problem list name; the mail and reject logs now
480 get a temp-reject line for the message that was being handled, saying
481 something like "domains check lookup or other defer". The SMTP 451
482 message is still "Temporary local problem".
484 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
485 A crafted sequence of BDAT commands could result in in-use memory beeing
486 freed. CVE-2017-16943.
488 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
489 from SMTP input. Previously it was always done; now only done for DATA
490 and not BDAT commands. CVE-2017-16944.
492 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
493 to the message (such as an overlong header line). Previously this was
494 not done and we did not exit BDAT mode. Followon from the previous item
495 though a different problem.
501 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
502 than -2003 did; needs libidn2 in addition to libidn.
504 JH/02 The path option on a pipe transport is now expanded before use.
506 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
507 Patch provided by "Björn", documentation fix added too.
509 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
510 missing a wire-to-host endian conversion.
512 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
513 close after a BDAT command line could be taken as a following command,
514 giving a synch failure. Fix by only checking for synch immediately
515 before acknowledging the chunk.
517 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
518 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
519 macro. Patches provided by Josh Soref.
521 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
522 Previously we did not; the RFC seems ambiguous and VRFY is not listed
523 by IANA as a service extension. However, John Klensin suggests that we
526 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
527 the dkim code may be unix-mode line endings rather than smtp wire-format
528 CRLF, so prepend a CR to any bare LF.
530 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
531 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
533 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
534 the first were themselves being wrongly included in the feed into dkim
535 processing; with most chunk sizes in use this resulted in an incorrect
536 body hash calculated value.
538 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
539 DKIM signature block, for verification. Although advised against by
540 standards it is specifically not ruled illegal.
542 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
544 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
545 missing a body hash (the bh= tag).
547 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
548 It seems that HAProxy sends the Proxy Protocol information in clear and
549 only then does a TLS startup, so do the same.
551 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
552 TCP connections (such as for Spamd) unless the daemon successfully set
553 Fast Open mode on its listening sockets. This fixes breakage seen on
554 too-old kernels or those not configured for Fast Open, at the cost of
555 requiring both directions being enabled for TFO, and TFO never being used
556 by non-daemon-related Exim processes.
558 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
559 endings, at least on the first header line. Try to canonify any that get
560 past that check, despite the cost.
562 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
563 now limited to an arbitrary five deep, while parsing addresses with the
564 strip_excess_angle_brackets option enabled.
566 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
567 instead leave the unprompted TLS handshake in socket buffer for the
568 TLS library to consume.
570 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
572 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
574 JH/16 Drop variables when they go out of scope. Memory management drops a whole
575 region in one operation, for speed, and this leaves assigned pointers
576 dangling. Add checks run only under the testsuite which checks all
577 variables at a store-reset and panics on a dangling pointer; add code
578 explicitly nulling out all the variables discovered. Fixes one known
579 bug: a transport crash, where a dangling pointer for $sending_ip_address
580 originally assigned in a verify callout, is re-used.
582 PP/06 Drop '.' from @INC in various Perl scripts.
584 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
586 PP/08 Reduce a number of compilation warnings under clang; building with
587 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
588 should be warning-free.
590 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
592 HS/01 Fix portability problems introduced by PP/08 for platforms where
593 realloc(NULL) is not equivalent to malloc() [SunOS et al].
595 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
596 chunk. This allows us to accept broken chunked messages. We need a more
597 general solution here.
599 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
600 already-broken messages in the queue.
602 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
604 JH/19 Fix reference counting bug in routing-generated-address tracking.
610 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
611 supports it and a size is available (ie. the sending peer gave us one).
613 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
614 years of being deprecated). The replacements are the ACLs
615 acl_smtp_mime and acl_not_smtp_mime.
617 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
618 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
619 or even an in-clear connection were permitted. Now, if the host lookup
620 was dnssec and dane was requested then the host is only used if the
621 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
622 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
623 if one fails this test.
624 This means that a poorly-configured remote DNS will make it incommunicado;
625 but it protects against a DNS-interception attack on it.
627 JH/04 Bug 1810: make continued-use of an open smtp transport connection
628 non-noisy when a race steals the message being considered.
630 JH/05 If main configuration option tls_certificate is unset, generate a
631 self-signed certificate for inbound TLS connections.
633 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
634 in rewrites and routers.
636 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
637 and logged a warning sing 4.83; now they are a configuration file error.
639 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
640 (lacking @domain). Apply the same qualification processing as RCPT.
642 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
644 JH/10 Support ${sha256:} applied to a string (as well as the previous
647 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
648 a cutthrough deliver is pending, as we always want to make a connection.
649 This also avoids re-routing the message when later placing the cutthrough
650 connection after a verify cache hit.
651 Do not update it with the verify result either.
653 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
654 when routing results in more than one destination address.
656 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
657 signing (which inhibits the cutthrough capability). Previously only
658 the presence of an option was tested; now an expansion evaluating as
659 empty is permissible (obviously it should depend only on data available
660 when the cutthrough connection is made).
662 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
663 the relevant preceding SMTP command did not note the pipelining mode.
665 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
666 Previously they were not counted.
668 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
669 as one having no matching records. Previously we deferred the message
670 that needed the lookup.
672 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
673 distinguished as "(=".
675 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
676 for missing MX records. Previously it only worked for missing A records.
678 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
680 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
681 after the data-go-ahead and data-ack. Patch from Jason Betts.
683 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
684 even for a "none" policy. Patch from Tony Meyer.
686 JH/22 Fix continued use of a connection for further deliveries. If a port was
687 specified by a router, it must also match for the delivery to be
690 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
691 When one of the recipients of a message was unsuitable for the connection
692 (has no matching addresses), we lost track of needing to mark it
693 deferred. As a result mail would be lost.
695 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
697 JH/25 Decoding ACL controls is now done using a binary search; the source code
698 takes up less space and should be simpler to maintain. Merge the ACL
699 condition decode tables also, with similar effect.
701 JH/26 Fix problem with one_time used on a redirect router which returned the
702 parent address unchanged. A retry would see the parent address marked as
703 delivered, so not attempt the (identical) child. As a result mail would
706 JH/27 Fix a possible security hole, wherein a process operating with the Exim
707 UID can gain a root shell. Credit to http://www.halfdog.net/ for
708 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
711 JH/28 Enable {spool,log} filesystem space and inode checks as default.
712 Main config options check_{log,spool}_{inodes,space} are now
713 100 inodes, 10MB unless set otherwise in the configuration.
715 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
716 Previously it only applied to the main-section connection policy
719 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
721 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
722 by me. Added RFC7919 DH primes as an alternative.
724 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
725 are not in the system include path.
727 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
728 GnuTLS, when a session startup failed (eg because the client disconnected)
729 Exim did stdio operations after fclose. This was exposed by a recent
730 change which nulled out the file handle after the fclose.
732 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
733 signed directly by the cert-signing cert, rather than an intermediate
734 OCSP-signing cert. This is the model used by LetsEncrypt.
736 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
738 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
739 an incoming connection.
741 HS/02 Bug 1802: Do not half-close the connection after sending a request
744 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
745 fallback to "prime256v1".
747 JH/34 SECURITY: Use proper copy of DATA command in error message.
748 Could leak key material. Remotely exploitable. CVE-2016-9963.
754 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
755 and 3.4.4 - once the server is enabled to respond to an OCSP request
756 it does even when not requested, resulting in a stapling non-aware
757 client dropping the TLS connection.
759 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
760 support variable-length bit vectors. No functional change.
762 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
763 The I= interface field on outgoing lines is now after the H= remote
764 host field, same as incoming lines. There is a separate
765 outgoing_interface log selector which allows you to disable the
768 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
769 If not running log_selector +smtp_connection the mainlog would be held
770 open indefinitely after a "too many connections" event, including to a
771 deleted file after a log rotate. Leave the per net connection logging
772 leaving it open for efficiency as that will be quickly detected by the
773 check on the next write.
775 HS/01 Bug 1671: Fix post transport crash.
776 Processing the wait-<transport> messages could crash the delivery
777 process if the message IDs didn't exist for some reason. When
778 using 'split_spool_directory=yes' the construction of the spool
779 file name failed already, exposing the same netto behaviour.
781 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
782 mime_regex ACL conditions.
784 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
785 to DSN fail messages (bounces): remote IP, remote greeting, remote response
786 to HELO, local diagnostic string.
788 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
789 log line to debug. Even when configured with a tls authenticator many
790 client connections are expected to not authenticate in this way, so
791 an authenticate fail is not an error.
793 HS/02 Add the Exim version string to the process info. This way exiwhat
794 gives some more detail about the running daemon.
796 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
797 matter for fast-change records such as DNSBLs.
799 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
800 retry record, even if constant. There may be multiple transports with
801 different interface settings and the retry behaviour needs to be kept
804 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
806 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
808 JH/10 Bug 840: fix log_defer_output option of pipe transport
810 JH/11 Bug 830: use same host for all RCPTS of a message, even under
811 hosts_randomize. This matters a lot when combined with mua_wrapper.
813 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
814 ${quote_pgsql:<string>} operator.
816 JH/13 Bug 1708: avoid misaligned access in cached lookup.
818 JH/14 Change header file name for freeradius-client. Relevant if compiling
819 with Radius support; from the Gentoo tree and checked under Fedora.
821 JH/15 Bug 1712: Introduce $prdr_requested flag variable
823 JH/16 Bug 1714: Permit an empty string as expansion result for transport
824 option transport_filter, meaning no filtering.
826 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
828 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
829 defaults to "*" (all hosts). The variable is now available when not built
830 with TLS, default unset, mainly to enable keeping the testsuite sane.
831 If a server certificate is not supplied (via tls_certificate) an error is
832 logged, and clients will find TLS connections fail on startup. Presumably
833 they will retry in-clear.
834 Packagers of Exim are strongly encouraged to create a server certificate
835 at installation time.
837 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
838 with the $config_file variable.
840 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
841 in transport context, after the attempt, and per-recipient. The latter type
842 is per host attempted. The event data is the error message, and the errno
843 information encodes the lookup type (A vs. MX) used for the (first) host,
844 and the trailing two digits of the smtp 4xx response.
846 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
847 to write to mainlog (or rejectlog, paniclog) in the window between file
848 creation and permissions/ownership being changed. Particularly affects
849 installations where exicyclog is run as root, rather than exim user;
850 result is that the running daemon panics and dies.
852 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
854 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
855 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
856 "pri" and "weight". Note that the previous implicit priority given by the
857 list order is no longer honoured.
859 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
862 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
863 by defining SUPPORT_SOCKS.
865 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
866 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
867 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
868 variables are renamed to proxy_{local,external}_{address,port}.
870 JH/27 Move Internationalisation support from Experimental to mainline, enabled
871 for a build by defining SUPPORT_I18N
873 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
874 of the query string, and make ${quote_redis:} do that quoting.
876 JH/29 Move Events support from Experimental to mainline, enabled by default
877 and removable for a build by defining DISABLE_EVENT.
879 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
881 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
882 cached by the daemon.
884 JH/32 Move Redis support from Experimental to mainline, enabled for a build
885 by defining LOOKUP_REDIS. The libhiredis library is required.
887 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
888 keys are given for lookup.
890 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
891 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
892 only supported when built with TLS support. The PolarSSL SHA routines
893 are still used when the TLS library is too old for convenient support.
895 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
896 openssl_options), for security. OpenSSL forces this from version 1.1.0
897 server-side so match that on older versions.
899 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
900 allocation for $value could be released as the expansion processing
901 concluded, but leaving the global pointer active for it.
903 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
904 and to use the domains and local_parts ACL conditions.
906 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
907 incorrectly not doubled on cutthrough transmission, hence seen as a
908 body-termination at the receiving system - resulting in truncated mails.
909 Commonly the sender saw a TCP-level error, and retransmitted the message
910 via the normal store-and-forward channel. This could result in duplicates
911 received - but deduplicating mailstores were liable to retain only the
912 initial truncated version.
914 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
916 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
918 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
919 we're in there, support oversigning also; bug 1309.
921 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
923 HS/04 Add support for keep_environment and add_environment options.
925 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
926 either intentional arithmetic overflow during PRNG, or testing config-
929 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
930 delivery resulted in actual delivery. Cancel cutthrough before DATA
933 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
934 rejects a recipient: pass the reject to the originator.
936 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
937 Many were false-positives and ignorable, but it's worth fixing the
940 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
941 for the new environment-manipulation done at startup. Move the routines
942 from being local to tls.c to being global via the os.c file.
944 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
945 an extract embedded as result-arg for a map, the first arg for extract
946 is unavailable so we cannot tell if this is a numbered or keyed
947 extraction. Accept either.
953 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
956 JH/02 The smtp transport option "multi_domain" is now expanded.
958 JH/03 The smtp transport now requests PRDR by default, if the server offers
961 JH/04 Certificate name checking on server certificates, when exim is a client,
962 is now done by default. The transport option tls_verify_cert_hostnames
963 can be used to disable this per-host. The build option
964 EXPERIMENTAL_CERTNAMES is withdrawn.
966 JH/05 The value of the tls_verify_certificates smtp transport and main options
967 default to the word "system" to access the system default CA bundle.
968 For GnuTLS, only version 3.0.20 or later.
970 JH/06 Verification of the server certificate for a TLS connection is now tried
971 (but not required) by default. The verification status is now logged by
972 default, for both outbound TLS and client-certificate supplying inbound
975 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
978 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
979 Status Notification (bounce) messages are now MIME format per RFC 3464.
980 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
981 under the control of the dsn_advertise_hosts option, and routers may
982 have a dsn_lasthop option.
984 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
985 default, modifiable by a malware= option. The list separator for
986 the options can now be changed in the usual way. Bug 68.
988 JH/10 The smtp_receive_timeout main option is now expanded before use.
990 JH/11 The incoming_interface log option now also enables logging of the
991 local interface on delivery outgoing connections.
993 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
994 if the interface and destination host and port all match.
996 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
999 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1000 Patch from Andrew Lewis.
1002 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1003 now supports optional time-restrictions, weighting, and priority
1004 modifiers per server. Patch originally by <rommer@active.by>.
1006 JH/16 The spamd_address main option now supports a mixed list of local
1007 and remote servers. Remote servers can be IPv6 addresses, and
1008 specify a port-range.
1010 JH/17 Bug 68: The spamd_address main option now supports an optional
1011 timeout value per server.
1013 JH/18 Bug 1581: Router and transport options headers_add/remove can
1014 now have the list separator specified.
1016 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1019 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1022 JH/21 Support for the A6 type of dns record is withdrawn.
1024 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1025 rather than the verbs used.
1027 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1028 from 255 to 1024 chars.
1030 JH/24 Verification callouts now attempt to use TLS by default.
1032 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1033 are generic router options now. The defaults didn't change.
1035 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1036 Original patch from Alexander Shikoff, worked over by JH.
1038 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1039 files if ZCAT_COMMAND is not executable.
1041 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1043 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1045 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1046 Normally benign, it bites when the pair was led to by a CNAME;
1047 modern usage is to not canonicalize the domain to a CNAME target
1048 (and we were inconsistent anyway for A-only vs AAAA+A).
1050 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1052 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1053 when evaluating $sender_host_dnssec.
1055 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1056 $sender_helo_dnssec variable.
1058 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1060 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1062 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1064 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1065 documented as working, but never had. Support all but $spam_report.
1067 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1068 added for tls authenticator.
1070 HS/03 Add perl_taintmode main config option
1076 TL/01 When running the test suite, the README says that variables such as
1077 no_msglog_check are global and can be placed anywhere in a specific
1078 test's script, however it was observed that placement needed to be near
1079 the beginning for it to behave that way. Changed the runtest perl
1080 script to read through the entire script once to detect and set these
1081 variables, reset to the beginning of the script, and then run through
1082 the script parsing/test process like normal.
1084 TL/02 The BSD's have an arc4random API. One of the functions to induce
1085 adding randomness was arc4random_stir(), but it has been removed in
1086 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1087 function when detected.
1089 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1090 cause callback expansion.
1092 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1093 syntax errors in an expansion can be treated as a string instead of
1094 logging or causing an error, due to the internal use of bool_lax
1095 instead of bool when processing it.
1097 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1098 server certificates when making smtp deliveries.
1100 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1102 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1104 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1106 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1107 Merged patch from Sebastian Wiedenroth.
1109 JH/05 Fix results-pipe from transport process. Several recipients, combined
1110 with certificate use, exposed issues where response data items split
1111 over buffer boundaries were not parsed properly. This eventually
1112 resulted in duplicates being sent. This issue only became common enough
1113 to notice due to the introduction of connection certificate information,
1114 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1116 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1117 size buffer was used, resulting in syntax errors when an expansion
1120 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1121 version 3.3.6 or later.
1123 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1124 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1125 both become "event_action", the variables become $event_name, $event_data
1126 and $event_defer_errno. There is a new variable $verify_mode, usable in
1127 routers, transports and related events. The tls:cert event is now also
1128 raised for inbound connections, if the main configuration event_action
1131 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1132 early OCSP support, but no stapling (appears to be less than 1.0.0).
1134 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1135 server certificate names available under the smtp transport option
1136 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1139 JH/10 Time-related extraction expansions from certificates now use the main
1140 option "timezone" setting for output formatting, and are consistent
1141 between OpenSSL and GnuTLS compilations. Bug 1541.
1143 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1144 encoded parameter in the incoming message. Bug 1558.
1146 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1147 include certificate info, eximon was claiming there were spoolfile
1150 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1152 JH/14 Log delivery-related information more consistently, using the sequence
1153 "H=<name> [<ip>]" wherever possible.
1155 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1156 are problematic for Debian distribution, omit them from the release
1159 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1161 JH/16 Fix string representation of time values on 64bit time_t architectures.
1164 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1165 output list separator was used.
1170 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1171 checkers that were complaining about end of non-void function with no
1174 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1175 This was a regression introduced in 4.83 by another bugfix.
1177 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1179 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1180 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1186 TF/01 Correctly close the server side of TLS when forking for delivery.
1188 When a message was received over SMTP with TLS, Exim failed to clear up
1189 the incoming connection properly after forking off the child process to
1190 deliver the message. In some situations the subsequent outgoing
1191 delivery connection happened to have the same fd number as the incoming
1192 connection previously had. Exim would try to use TLS and fail, logging
1193 a "Bad file descriptor" error.
1195 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1196 utilities have not been installed.
1198 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1199 temporary space as the ACL may create new global variables.
1201 TL/01 LDAP support uses per connection or global context settings, depending
1202 upon the detected version of the libraries at build time.
1204 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1205 to extract and use the src ip:port in logging and expansions as if it
1206 were a direct connection from the outside internet. PPv2 support was
1207 updated based on HAProxy spec change in May 2014.
1209 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1211 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1212 Properly escape header and check for NULL return.
1214 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1217 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1219 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1220 characters in header names, implemented as a verify condition.
1221 Contributed by Michael Fischer v. Mollard.
1223 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1224 results permerror and temperror. Previous values are deprecated but
1225 still accepted. In a future release, err_perm and err_temp will be
1226 completely removed, which will be a backward incompatibility if the
1227 ACL tests for either of these two old results. Patch contributed by
1228 user bes-internal on the mailing list.
1230 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1232 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1233 selectors, in both main and reject logs.
1235 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1238 JH/07 Add malware type "sock" for talking to simple daemon.
1240 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1242 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1243 routers/transports under cutthrough routing.
1245 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1246 numbers. Touch up "bool" conditional to keep the same definition.
1248 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1250 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1252 JH/12 Expand items in router/transport headers_add or headers_remove lists
1253 individually rather than the list as a whole. Bug 1452.
1255 Required for reasonable handling of multiple headers_ options when
1256 they may be empty; requires that headers_remove items with embedded
1257 colons must have them doubled (or the list-separator changed).
1259 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1260 view the policy declared in the DMARC record. Currently, $dmarc_status
1261 is a combined value of both the record presence and the result of the
1264 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1266 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1267 dnslookup router and the smtp transport (applying to the forward
1270 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1271 of ldap servers used for a specific lookup. Patch provided by Heiko
1274 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1275 New variable $lookup_dnssec_authenticated for observability.
1277 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1278 Patch submitted by Lars Timman.
1280 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1282 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1283 Requires trusted mode and valid format message id, aborts otherwise.
1284 Patch contributed by Heiko Schlichting.
1286 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1287 certextract with support for various fields. Bug 1358.
1289 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1290 is requested by default, modifiable by smtp transport option
1293 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1294 operate on certificate variables to give certificate fingerprints
1295 Also new ${sha256:cert_variable}.
1297 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1299 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1300 Christian Aistleitner.
1302 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1304 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1305 file. Patch from Wolfgang Breyha.
1307 JH/25 Expand the coverage of the delivery $host and $host_address to
1308 client authenticators run in verify callout. Bug 1476.
1310 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1311 align with daemon_smtp_ports. Bug 72.
1313 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1314 support and error reporting did not work properly.
1316 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1317 and is readable. Patch from Andrew Colin Kissa.
1319 TL/14 Enhance documentation of ${run expansion and how it parses the
1320 commandline after expansion, particularly in the case when an
1321 unquoted variable expansion results in an empty value.
1323 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1325 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1326 by renaming away from it. Addresses GH issue 15, reported by
1329 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1331 TL/15 SECURITY: prevent double expansion in math comparison functions
1332 (can expand unsanitized data). Not remotely exploitable.
1339 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1341 PP/02 Make -n do something, by making it not do something.
1342 When combined with -bP, the name of an option is not output.
1344 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1347 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1348 $sender_host_name and config options to manage this, and basic check
1351 PP/05 DSCP support for outbound connections and control modifier for inbound.
1353 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1354 (Only plugin which currently uses this is kerberos4, which nobody should
1355 be using, but we should make it available and other future plugins might
1356 conceivably use it, even though it would break NAT; stuff *should* be
1357 using channel bindings instead).
1359 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1360 name; added for Sendmail compatibility; requires admin caller.
1361 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1362 just ignore it); requires trusted caller.
1363 Also parse but ignore: -Ac -Am -X<logfile>
1366 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1368 TL/02 Add +smtp_confirmation as a default logging option.
1370 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1371 Patch by Magnus Holmgren from 2007-02-20.
1373 TL/04 Bugzilla 1281 - Spec typo.
1374 Bugzilla 1283 - Spec typo.
1375 Bugzilla 1290 - Spec grammar fixes.
1377 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1379 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1381 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1382 mailing list by by Dmitry Isaikin.
1384 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1386 JH/02 Support "G" suffix to numbers in ${if comparisons.
1388 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1390 NM/01 Bugzilla 1197 - Spec typo
1391 Bugzilla 1196 - Spec examples corrections
1393 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1395 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1396 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1399 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1400 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1402 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1403 "acl {{name}{arg}...}", and optional args on acl condition
1406 JH/05 Permit multiple router/transport headers_add/remove lines.
1408 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1410 JH/07 Avoid using a waiting database for a single-message-only transport.
1411 Performance patch from Paul Fisher. Bugzilla 1262.
1413 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1416 JH/09 Add $headers_added variable, with content from use of ACL modifier
1417 add_header (but not yet added to the message). Bugzilla 199.
1419 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1420 Pulled from Bugzilla 817 by Wolfgang Breyha.
1422 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1424 (nb: this is the same fix as in Exim 4.80.1)
1426 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1429 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1432 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1434 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1435 advertise SMTP AUTH mechanism to us, instead of a generic
1436 protocol violation error. Also, make Exim more robust to bad
1437 data from the Dovecot auth socket.
1439 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1441 When a queue runner is handling a message, Exim first routes the
1442 recipient addresses, during which it prunes them based on the retry
1443 hints database. After that it attempts to deliver the message to
1444 any remaining recipients. It then updates the hints database using
1447 So if a recipient address works intermittently, it can get repeatedly
1448 deferred at routing time. The retry hints record remains fresh so the
1449 address never reaches the final cutoff time.
1451 This is a fairly common occurrence when a user is bumping up against
1452 their storage quota. Exim had some logic in its local delivery code
1453 to deal with this. However it did not apply to per-recipient defers
1454 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1456 This change adds a proper retry rule check during routing so that the
1457 final cutoff time is checked against the message's age. We only do
1458 this check if there is an address retry record and there is not a
1459 domain retry record; this implies that previous attempts to handle
1460 the address had the retry_use_local_parts option turned on. We use
1461 this as an approximation for the destination being like a local
1462 delivery, as in LMTP.
1464 I suspect this new check makes the old local delivery cutoff check
1465 redundant, but I have not verified this so I left the code in place.
1467 TF/02 Correct gecos expansion when From: is a prefix of the username.
1469 Test 0254 submits a message to Exim with the header
1473 When I ran the test suite under the user fanf2, Exim expanded
1474 the header to contain my full name, whereas it should have added
1475 a Resent-Sender: header. It erroneously treats any prefix of the
1476 username as equal to the username.
1478 This change corrects that bug.
1480 GF/01 DCC debug and logging tidyup
1481 Error conditions log to paniclog rather than rejectlog.
1482 Debug lines prefixed by "DCC: " to remove any ambiguity.
1484 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1486 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1487 Bug spotted by Jeremy Harris; was flawed since initial commit.
1488 Would have resulted in OCSP responses post-SNI triggering an Exim
1489 NULL dereference and crash.
1491 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1493 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1494 Bug detection, analysis and fix by Samuel Thibault.
1495 Bugzilla 1331, Debian bug #698092.
1497 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1499 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1500 Server implementation by Todd Lyons, client by JH.
1501 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1502 config variable "prdr_enable" controls whether the server
1503 advertises the facility. If the client requests PRDR a new
1504 acl_data_smtp_prdr ACL is called once for each recipient, after
1505 the body content is received and before the acl_smtp_data ACL.
1506 The client is controlled by both of: a hosts_try_prdr option
1507 on the smtp transport, and the server advertisement.
1508 Default client logging of deliveries and rejections involving
1509 PRDR are flagged with the string "PRDR".
1511 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1512 fclose(). Diagnosis by Todd Lyons.
1514 PP/17 Update configure.default to handle IPv6 localhost better.
1515 Patch by Alain Williams (plus minor tweaks).
1518 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1519 This is now consistent with GnuTLS, and is now documented: the
1520 previous undocumented portable approach to treating the option as
1521 unset was to force an expansion failure. That still works, and
1522 an empty string is now equivalent.
1524 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1525 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1526 not performing validation itself.
1528 PP/20 Added force_command boolean option to pipe transport.
1529 Patch from Nick Koston, of cPanel Inc.
1531 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1534 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1536 PP/21 Fix eximon continuous updating with timestamped log-files.
1537 Broken in a format-string cleanup in 4.80, missed when I repaired the
1538 other false fix of the same issue.
1539 Report and fix from Heiko Schlichting.
1542 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1543 Report from Prashanth Katuri.
1545 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1546 It's SecureTransport, so affects any MacOS clients which use the
1547 system-integrated TLS libraries, including email clients.
1549 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1550 using a MIME ACL for non-SMTP local injection.
1551 Report and assistance in diagnosis by Warren Baker.
1553 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1555 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1557 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1558 last id that failed so it may be referenced in subsequent ACL's.
1560 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1563 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1564 ldap library initialization, allowing self-signed CA's to be
1565 used. Also properly sets require_cert option later in code by
1566 using NULL (global ldap config) instead of ldap handle (per
1567 session). Bug diagnosis and testing by alxgomz.
1569 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1570 the src/util/ subdirectory.
1572 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1573 renamed to Transport Post Delivery Action by Jeremy Harris, as
1576 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1577 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1578 redis_servers = needs to be configured which will be used by the redis
1579 lookup. Patch from Warren Baker, of The Packet Hub.
1581 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1583 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1584 hostname or reverse DNS when processing a host list. Used suggestions
1585 from multiple comments on this bug.
1587 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1589 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1590 Missed a few lines, added it to make the runtest require no keyboard
1593 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1594 contains upper case chars. Make router use caseful_local_part.
1596 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1597 support when GnuTLS has been built with p11-kit.
1603 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1605 This, or similar/improved, will also be change PP/11 of 4.82.
1611 PP/01 Handle short writes when writing local log-files.
1612 In practice, only affects FreeBSD (8 onwards).
1613 Bugzilla 1053, with thanks to Dmitry Isaikin.
1615 NM/01 Bugzilla 949 - Documentation tweak
1617 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1620 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1622 PP/02 Implemented gsasl authenticator.
1624 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1626 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1627 `pkg-config foo` for cflags/libs.
1629 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1630 with rest of GSASL and with heimdal_gssapi.
1632 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1633 `pkg-config foo` for cflags/libs for the TLS implementation.
1635 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1636 properties get this fed in as external SSF. A number of robustness
1637 and debugging improvements to the cyrus_sasl authenticator.
1639 PP/08 cyrus_sasl server now expands the server_realm option.
1641 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1642 Patch by Jeremy Harris.
1644 PP/10 Added dbmjz lookup type.
1646 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1648 PP/12 MAIL args handles TAB as well as SP, for better interop with
1649 non-compliant senders.
1650 Analysis and variant patch by Todd Lyons.
1652 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1653 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1654 Patch from Dirk Mueller <dmueller@suse.com>
1656 PP/13 tls_peerdn now print-escaped for spool files.
1657 Observed some $tls_peerdn in wild which contained \n, which resulted
1658 in spool file corruption.
1660 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1661 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1662 or write after TLS renegotiation, which otherwise led to messages
1665 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1666 as a tracking header (ie: a signed header comes before the signature).
1667 Patch from Wolfgang Breyha.
1669 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1670 comma-sep list; embedded commas doubled.
1672 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1674 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1676 Report and patch from Dmitry Banschikov.
1678 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1679 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1680 failures. We appear to now support TLS1.1+ with Exim.
1682 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1683 lets Exim select keys and certificates based upon TLS SNI from client.
1684 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1685 before an outbound SMTP session. New log_selector, +tls_sni.
1687 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1688 NULL dereference. Report and patch from Alun Jones.
1690 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1691 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1692 Not seeing resolver debug output on NetBSD, but suspect this is a
1693 resolver implementation change.
1695 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1696 Left warnings. Added "eximon gdb" invocation mode.
1698 PP/21 Defaulting "accept_8bitmime" to true, not false.
1700 PP/22 Added -bw for inetd wait mode support.
1702 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1703 locate the relevant includes and libraries. Made this the default.
1705 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1706 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1708 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1709 This may cause build issues on older platforms.
1711 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1712 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1713 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1714 Added SNI support via GnuTLS too.
1715 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1717 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1719 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1720 Applied second patch from Janne, implementing suggestion to default
1721 multiple-strings-in-record handling to match SPF spec.
1723 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1725 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1726 read-only, out of scope).
1727 Patch from Wolfgang Breyha, report from Stuart Northfield.
1729 PP/29 Fix three issues highlighted by clang analyser static analysis.
1730 Only crash-plausible issue would require the Cambridge-specific
1731 iplookup router and a misconfiguration.
1732 Report from Marcin Mirosław.
1734 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1736 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1737 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1738 As part of this, removing so much warning spew let me fix some minor
1739 real issues in debug logging.
1741 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1742 assignment on my part. Fixed.
1744 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1745 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1746 Janne Snabb (who went above and beyond: thank you).
1748 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1749 string otherwise requires a connection and a bunch more work and it's
1750 relatively easy to get wrong. Should also expose TLS library linkage
1753 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1754 64-bit ${eval} (JH/03).
1756 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1757 GNU libc to support some of the 64-bit stuff, should not lead to
1758 conflicts. Defined before os.h is pulled in, so if a given platform
1759 needs to override this, it can.
1761 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1762 protection layer was required, which is not implemented.
1763 Bugzilla 1254, patch from Wolfgang Breyha.
1765 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1766 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1767 tls_dhparam take prime identifiers. Also unbreak combination of
1768 OpenSSL+DH_params+TLSSNI.
1770 PP/39 Disable SSLv2 by default in OpenSSL support.
1776 PP/01 Solaris build fix for Oracle's LDAP libraries.
1777 Bugzilla 1109, patch from Stephen Usher.
1779 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1781 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1784 TF/02 Fix a couple more cases where we did not log the error message
1785 when unlink() failed. See also change 4.74-TF/03.
1787 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1788 lock up or crash if it happened to be inside a call to libc when it
1789 got a SIGUSR1 from exiwhat.
1791 The SIGUSR1 handler appends the current process status to the process
1792 log which is later printed by exiwhat. It used to use the general
1793 purpose logging code to do this, but several functions it calls are
1794 not safe for signals.
1796 The new output code in the SIGUSR1 handler is specific to the process
1797 log, and simple enough that it's easy to inspect for signal safety.
1798 Removing some special cases also simplifies the general logging code.
1799 Removing the spurious timestamps from the process log simplifies
1802 TF/04 Improved ratelimit ACL condition.
1804 The /noupdate option has been deprecated in favour of /readonly which
1805 has clearer semantics. The /leaky, /strict, and /readonly update modes
1806 are mutually exclusive. The update mode is no longer included in the
1807 database key; it just determines when the database is updated. (This
1808 means that when you upgrade Exim will forget old rate measurements.)
1810 Exim now checks that the per_* options are used with an update mode that
1811 makes sense for the current ACL. For example, when Exim is processing a
1812 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1813 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1814 must specify per_mail/readonly. If you omit the update mode it defaults to
1815 /leaky where that makes sense (as before) or /readonly where required.
1817 The /noupdate option is now undocumented but still supported for
1818 backwards compatibility. It is equivalent to /readonly except that in
1819 ACLs where /readonly is required you may specify /leaky/noupdate or
1820 /strict/noupdate which are treated the same as /readonly.
1822 A useful new feature is the /count= option. This is a generalization
1823 of the per_byte option, so that you can measure the throughput of other
1824 aggregate values. For example, the per_byte option is now equivalent
1825 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1827 The per_rcpt option has been generalized using the /count= mechanism
1828 (though it's more complicated than the per_byte equivalence). When it is
1829 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1830 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1831 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1832 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1833 Note that using per_rcpt with a non-readonly update mode in more than
1834 one ACL will cause the recipients to be double-counted. (The per_mail
1835 and per_byte options don't have this problem.)
1837 The handling of very low rates has changed slightly. If the computed rate
1838 is less than the event's count (usually one) then this event is the first
1839 after a long gap. In this case the rate is set to the same as this event's
1840 count, so that the first message of a spam run is counted properly.
1842 The major new feature is a mechanism for counting the rate of unique
1843 events. The new per_addr option counts the number of different
1844 recipients that someone has sent messages to in the last time period. It
1845 behaves like per_rcpt if all the recipient addresses are different, but
1846 duplicate recipient addresses do not increase the measured rate. Like
1847 the /count= option this is a general mechanism, so the per_addr option
1848 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1849 example, measure the rate that a client uses different sender addresses
1850 with the options per_mail/unique=$sender_address. There are further
1851 details in the main documentation.
1853 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1855 TF/06 Removed a few PCRE remnants.
1857 TF/07 Automatically extract Exim's version number from tags in the git
1858 repository when doing development or release builds.
1860 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1861 Bugzilla 879. Patch from Paul Fisher.
1863 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1864 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1867 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1869 PP/05 Variable $av_failed, true if the AV scanner deferred.
1870 Bugzilla 1078. Patch from John Horne.
1872 PP/06 Stop make process more reliably on build failure.
1873 Bugzilla 1087. Patch from Heiko Schlittermann.
1875 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1876 Bugzilla 1089. Patch from Heiko Schlittermann.
1878 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1879 Bugzilla 1131. Patch from Holger Weiß.
1881 PP/09 Handle IPv6 addresses with SPF.
1882 Bugzilla 860. Patch from Wolfgang Breyha.
1884 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1886 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1889 PP/11 match_* no longer expand right-hand-side by default.
1890 New compile-time build option, EXPAND_LISTMATCH_RHS.
1891 New expansion conditions, "inlist", "inlisti".
1893 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1895 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1897 PP/14 fix log_write() format string regression from TF/03.
1898 Bugzilla 1152. Patch from Dmitry Isaikin.
1904 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1906 PP/02 Harmonised TLS library version reporting; only show if debugging.
1907 Layout now matches that introduced for other libraries in 4.74 PP/03.
1909 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1911 PP/04 New "dns_use_edns0" global option.
1913 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1916 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1917 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1919 TK/01 Updated PolarSSL code to 0.14.2.
1920 Bugzilla 1097. Patch from Andreas Metzler.
1922 PP/07 Catch divide-by-zero in ${eval:...}.
1923 Fixes bugzilla 1102.
1925 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1928 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1929 format-string attack -- SECURITY: remote arbitrary code execution.
1931 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1932 time unintentionally subject to list matching rules, letting the header
1933 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1934 arbitrary string expansion). This allowed for information disclosure.
1936 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1937 INT_MIN/-1 -- value coerced to INT_MAX.
1943 NM/01 Workaround for PCRE version dependency in version reporting
1946 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1947 This fixes portability to compilers other than gcc, notably
1948 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1950 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1951 makefiles for portability to HP-UX and POSIX correctness.
1953 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1954 Also via indented variable definition in the Makefile.
1955 (Debugging by Oliver Heesakkers).
1957 PP/02 Restore caching of spamd results with expanded spamd_address.
1958 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1960 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1961 Improves build reliability. Fix from: Frank Elsner
1963 NM/02 Fix wide character breakage in the rfc2047 coding
1964 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1966 NM/03 Allow underscore in dnslist lookups
1967 Fixes bug 1026. Patch from Graeme Fowler
1969 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1970 Code patches from Adam Ciarcinski of NetBSD.
1972 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1975 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1976 is logged, to avoid truncation. Patch from John Horne.
1978 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1979 Patch from Jakob Hirsch.
1981 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1982 SQL string expansion failure details.
1983 Patch from Andrey Oktyabrski.
1985 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1986 Patch from Simon Arlott.
1988 PP/09 New lookups functionality failed to compile on old gcc which rejects
1989 extern declarations in function scope.
1990 Patch from Oliver Fleischmann
1992 PP/10 Use sig_atomic_t for flags set from signal handlers.
1993 Check getgroups() return and improve debugging.
1994 Fixed developed for diagnosis in bug 927 (which turned out to be
1997 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1998 Patch from Mark Zealey.
2000 PP/12 Bugzilla 1056: Improved spamd server selection.
2001 Patch from Mark Zealey.
2003 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2004 Based on patch from Heiko Schlittermann.
2006 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2007 Patch from Uwe Doering, sign-off by Michael Haardt.
2009 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2010 variable declaration deep within a block. Bug and patch from
2013 PP/15 lookups-Makefile IRIX compatibility coercion.
2015 PP/16 Make DISABLE_DKIM build knob functional.
2017 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2018 Patch by Simon Arlott
2020 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2021 variable argument macros. Our copy now differs from upstream.
2027 TF/01 Failure to get a lock on a hints database can have serious
2028 consequences so log it to the panic log.
2030 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2031 controlled using the smtp_confirmation log selector.
2033 TF/03 Include the error message when we fail to unlink a spool file.
2035 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2036 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2037 for maintaining out-of-tree patches for some time.
2039 PP/01 Bugzilla 139: Documentation and portability issues.
2040 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2041 Handle per-OS dynamic-module compilation flags.
2043 PP/02 Let /dev/null have normal permissions.
2044 The 4.73 fixes were a little too stringent and complained about the
2045 permissions on /dev/null. Exempt it from some checks.
2046 Reported by Andreas M. Kirchwitz.
2048 PP/03 Report version information for many libraries, including
2049 Exim version information for dynamically loaded libraries. Created
2050 version.h, now support a version extension string for distributors
2051 who patch heavily. Dynamic module ABI change.
2053 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2054 privilege escalation vulnerability whereby the Exim run-time user
2055 can cause root to append content of the attacker's choosing to
2058 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2061 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2062 If dropping privileges for untrusted macros, we disabled normal logging
2063 on the basis that it would fail; for the Exim run-time user, this is not
2064 the case, and it resulted in successful deliveries going unlogged.
2065 Fixed. Reported by Andreas Metzler.
2071 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2072 only prepend for the Resent-* case. Fixes regression introduced in
2073 Exim 4.70 by NM/22 for Bugzilla 607.
2075 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2076 increasing numbers of administrators be bitten by this.
2078 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2080 PP/03 Bugzilla 994: added openssl_options main configuration option.
2082 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2084 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2086 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2088 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2089 without a peer certificate, leading to a segfault because of an
2090 assumption that peers always have certificates. Be a little more
2091 paranoid. Problem reported by Martin Tscholak.
2093 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2094 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2095 NB: ClamAV planning to remove STREAM in "middle of 2010".
2096 CL also introduces -bmalware, various -d+acl logging additions and
2097 more caution in buffer sizes.
2099 PP/09 Implemented reverse_ip expansion operator.
2101 PP/10 Bugzilla 937: provide a "debug" ACL control.
2103 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2105 PP/12 Bugzilla 973: Implement --version.
2107 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2109 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2111 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2113 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2114 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2115 ignore trailing whitespace.
2117 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2119 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2122 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2123 Notification from Dr Andrew Aitchison.
2125 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2126 ExtendedDetectionInfo response format.
2127 Notification from John Horne.
2129 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2132 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2133 XSL and documented dependency on system catalogs, with examples of how
2136 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2139 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2140 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2141 configuration file which is writeable by the Exim user or group.
2143 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2144 of configuration files to cover files specified with the -C option if
2145 they are going to be used with root privileges, not just the default
2148 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2149 option (effectively making it always true).
2151 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2152 files to be used while preserving root privileges.
2154 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2155 that rogue child processes cannot use them.
2157 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2158 run-time user, instead of root.
2160 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2161 Exim run-time user without dropping privileges.
2163 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2164 result string, instead of calling string_vformat() twice with the same
2167 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2168 for other users. Others should always drop root privileges if they use
2169 -C on the command line, even for a whitelisted configure file.
2171 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2173 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2179 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2180 $data_path, and $header_path variables; fixed documentation bugs and
2183 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2184 exipick to access non-standard spools, including the "frozen" queue
2187 NM/01 Bugzilla 965: Support mysql stored procedures.
2188 Patch from Alain Williams
2190 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2192 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2193 Patch from Andreas Metzler
2195 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2196 Patch from Kirill Miazine
2198 NM/05 Bugzilla 671: Added umask to procmail example.
2200 JJ/03 installed exipick 20100323.0, fixing doc bug
2202 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2203 directory. Notification and patch from Dan Rosenberg.
2205 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2207 TK/02 Improve log output when DKIM signing operation fails.
2209 MH/01 Treat the transport option dkim_domain as a colon separated
2210 list, not as a single string, and sign the message with each element,
2211 omitting multiple occurences of the same signer.
2213 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2214 Bugzilla 985, 986. Patch by Simon Arlott
2216 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2217 Patch by Simon Arlott
2219 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2220 MBX locking. Notification from Dan Rosenberg.
2226 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2228 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2230 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2232 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2234 NM/04 Bugzilla 915: Flag broken perl installation during build.
2240 TK/01 Added patch by Johannes Berg that expands the main option
2241 "spamd_address" if it starts with a dollar sign.
2243 TK/02 Write list of recipients to X-Envelope-Sender header when building
2244 the mbox-format spool file for content scanning (suggested by Jakob
2247 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2248 (http://www.dcc-servers.net/) support via dccifd. Activated by
2249 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2251 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2252 by Mark Daniel Reidel <mr@df.eu>.
2254 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2255 When building exim an external PCRE library is now needed -
2256 PCRE is a system library on the majority of modern systems.
2257 See entry on PCRE_LIBS in EDITME file.
2259 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2260 conversation. Added nologin parameter to request.
2261 Patch contributed by Kirill Miazine.
2263 TF/01 Do not log submission mode rewrites if they do not change the address.
2265 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2267 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2268 log files in place. Contributed by Roberto Lima.
2270 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2272 TF/03 Bugzilla 615: When checking the local_parts router precondition
2273 after a local_part_suffix or local_part_prefix option, Exim now
2274 does not use the address's named list lookup cache, since this
2275 contains cached lookups for the whole local part.
2277 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2278 Robert Millan. Documentation is in experimental-spec.txt.
2280 TF/04 Bugzilla 668: Fix parallel build (make -j).
2282 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2284 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2285 Patch provided by Jan Srzednicki.
2287 TF/05 Leading white space used to be stripped from $spam_report which
2288 wrecked the formatting. Now it is preserved.
2290 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2291 that they are available at delivery time.
2293 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2295 TF/08 TLS error reporting now respects the incoming_interface and
2296 incoming_port log selectors.
2298 TF/09 Produce a more useful error message if an SMTP transport's hosts
2299 setting expands to an empty string.
2301 NM/06 Bugzilla 744: EXPN did not work under TLS.
2302 Patch provided by Phil Pennock.
2304 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2305 Patch provided by Richard Godbee.
2307 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2308 acl_smtp_notquit, added index entry.
2310 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2311 Patch provided by Eugene Bujak.
2313 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2314 accept(). Patch provided by Maxim Dounin.
2316 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2317 Patch provided by Phil Pennock.
2319 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2321 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2322 Patch provided by Brad "anomie" Jorsch.
2324 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2325 Patch provided by Dean Brooks.
2327 TK/05 Add native DKIM support (does not depend on external libraries).
2329 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2330 Patch provided by Graeme Fowler.
2332 NM/16 Bugzilla 851: Documentation example syntax fix.
2334 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2336 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2339 NM/19 Bugzilla 745: TLS version reporting.
2340 Patch provided by Phil Pennock.
2342 NM/20 Bugzilla 167: bool: condition support.
2343 Patch provided by Phil Pennock.
2345 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2346 clients. Patch provided by Phil Pennock.
2348 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2349 Patch provided by Brad "anomie" Jorsch.
2351 NM/23 Bugzilla 687: Fix misparses in eximstats.
2352 Patch provided by Heiko Schlittermann.
2354 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2355 Patch provided by Heiko Schlittermann.
2357 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2358 plus update to original patch.
2360 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2362 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2363 Patch provided by David Brownlee.
2365 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2367 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2369 NM/30 Bugzilla 888: TLS documentation bugfixes.
2371 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2373 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2374 Unlike the original bugzilla I have changed all shell scripts in src tree.
2376 NM/33 Bugzilla 898: Transport filter timeout fix.
2377 Patch by Todd Rinaldo.
2379 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2380 Patch by Serge Demonchaux.
2382 NM/35 Bugzilla 39: Base64 decode bug fixes.
2383 Patch by Jakob Hirsch.
2385 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2387 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2389 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2391 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2397 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2398 ALT-N's libdkim that I have put here:
2399 http://duncanthrax.net/exim-experimental/
2401 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2402 were called 'true' and it seems that C99 defines that as a reserved
2403 keyword to be used with 'bool' variable types. That means you could
2404 not include C99-style headers which use bools without triggering
2405 build errors in sieve.c.
2407 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2408 as mailq or other aliases. Changed the --help handling significantly
2409 to do whats expected. exim_usage() emits usage/help information.
2411 SC/01 Added the -bylocaldomain option to eximstats.
2413 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2415 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2417 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2423 PH/01 Another patch from the Sieve maintainer.
2425 PH/02 When an IPv6 address is converted to a string for single-key lookup
2426 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2427 dots are used instead of colons so that keys in lsearch files need not
2428 contain colons. This was done some time before quoting was made available
2429 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2430 (notated using the quote facility) so as to distinguish them from IPv4
2431 keys. This meant that lookups for IP addresses in host lists did not work
2432 for iplsearch lookups.
2434 This has been fixed by arranging for IPv6 addresses to be expressed with
2435 colons if the lookup type is iplsearch. This is not incompatible, because
2436 previously such lookups could never work.
2438 The situation is now rather anomalous, since one *can* have colons in
2439 ordinary lsearch keys. However, making the change in all cases is
2440 incompatible and would probably break a number of configurations.
2442 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2445 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2446 conversion specification without a maximum field width, thereby enabling
2447 a rogue spamd server to cause a buffer overflow. While nobody in their
2448 right mind would setup Exim to query an untrusted spamd server, an
2449 attacker that gains access to a server running spamd could potentially
2450 exploit this vulnerability to run arbitrary code as the Exim user.
2452 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2453 $primary_hostname instead of what libspf2 thinks the hosts name is.
2455 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2456 a directory entry by the name of the lookup key. Previously, if a
2457 symlink pointed to a non-existing file or a file in a directory that
2458 Exim lacked permissions to read, a lookup for a key matching that
2459 symlink would fail. Now it is enough that a matching directory entry
2460 exists, symlink or not. (Bugzilla 503.)
2462 PH/03 The body_linecount and body_zerocount variables are now exported in the
2465 PH/04 Added the $dnslist_matched variable.
2467 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2468 This means they are set thereafter only if the connection becomes
2471 PH/06 Added the client_condition to authenticators so that some can be skipped
2472 by clients under certain conditions.
2474 PH/07 The error message for a badly-placed control=no_multiline_responses left
2475 "_responses" off the end of the name.
2477 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2479 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2480 (without spaces) instead of just copying the configuration text.
2482 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2484 PH/11 Added $max_received_linelength.
2486 PH/12 Added +ignore_defer and +include_defer to host lists.
2488 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2489 way in which PCRE > 7.0 is built.
2491 PH/14 Implemented queue_only_load_latch.
2493 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2494 MAIL command. The effect was to mangle the value on 64-bit systems.
2496 PH/16 Another patch from the Sieve maintainer.
2498 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2500 PH/18 If a system quota error occurred while trying to create the file for
2501 a maildir delivery, the message "Mailbox is full" was not appended to the
2502 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2503 only to a quota excession during the actual writing of the file.
2505 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2506 characters?) which causes problems in log lines. The DN values are now
2507 passed through string_printing() before being added to log lines.
2509 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2510 and InterBase are left for another time.)
2512 PH/21 Added message_body_newlines option.
2514 PH/22 Guard against possible overflow in moan_check_errorcopy().
2516 PH/23 POSIX allows open() to be a macro; guard against that.
2518 PH/24 If the recipient of an error message contained an @ in the local part
2519 (suitably quoted, of course), incorrect values were put in $domain and
2520 $local_part during the evaluation of errors_copy.
2526 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2527 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2530 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2531 issue a MAIL command.
2533 PH/02 In an ACL statement such as
2535 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2537 if a client was not listed at all, or was listed with a value other than
2538 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2539 the condition was not true (as it should be), so access was not denied.
2540 The bug was that the ! inversion was incorrectly passed on to the second
2541 item. This has been fixed.
2543 PH/03 Added additional dnslists conditions == and =& which are different from
2544 = and & when the dns lookup returns more than one IP address.
2546 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2547 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2549 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2550 FSYNC, which compiles an option called disable_fsync that allows for
2551 bypassing fsync(). The documentation is heavily laced with warnings.
2553 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2555 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2556 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2557 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2558 including adding "make clean"; (3) Added -fPIC when compiling the test
2559 dynamically loaded module, to get rid of a warning.
2561 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2562 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2563 = 0s. The bug is otherwise harmless.
2565 PH/07 There was a bug in the dovecot authenticator such that the value of
2566 $auth1 could be overwritten, and so not correctly preserved, after a
2567 successful authentication. This usually meant that the value preserved by
2568 the server_setid option was incorrect.
2570 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2572 PH/09 Installed PCRE release 7.0.
2574 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2575 run for batched SMTP input. It is now run at the start of every message
2576 in the batch. While fixing this I discovered that the process information
2577 (output by running exiwhat) was not always getting set for -bs and -bS
2578 input. This is fixed, and it now also says "batched" for BSMTP.
2580 PH/11 Added control=no_pipelining.
2582 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2583 patch, slightly modified), and move the expansion of helo_data till after
2584 the connection is made in the smtp transport (so it can use these
2587 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2589 PH/14 Added log_selector = +pid.
2591 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2593 PH/16 Add ${if forany and ${if forall.
2595 PH/17 Added dsn_from option to vary the From: line in DSNs.
2597 PH/18 Flush SMTP output before performing a callout, unless control =
2598 no_callout_flush is set.
2600 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2601 was true (the default) a successful delivery failed to delete the retry
2602 item, thus causing premature timeout of the address. The bug is now
2605 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2607 PH/21 Long custom messages for fakedefer and fakereject are now split up
2608 into multiline responses in the same way that messages for "deny" and
2609 other ACL rejections are.
2611 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2612 with slight modification.
2614 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2615 draft, changing the syntax and factoring some duplicate code".
2617 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2618 for deliveries of the second and subsequent messages over the same SMTP
2621 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2622 ${reduce, with only minor "tidies".
2624 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2626 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2627 expansion side effects.
2629 PH/27 When a message times out after an over-quota error from an Exim-imposed
2630 quota, the bounce message says "mailbox is full". This message was not
2631 being given when it was a system quota that was exceeded. It now should
2634 MH/03 Made $recipients available in local_scan(). local_scan() already has
2635 better access to the recipient list through recipients_list[], but
2636 $recipients can be useful in postmaster-provided expansion strings.
2638 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2639 in the case of a MAIL command with additional options following the
2640 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2641 were accidentally chopped off.
2643 PH/29 SMTP synchronization checks are implemented when a command is read -
2644 there is a check that no more input is waiting when there shouldn't be
2645 any. However, for some commands, a delay in an ACL can mean that it is
2646 some time before the response is written. In this time, more input might
2647 arrive, invalidly. So now there are extra checks after an ACL has run for
2648 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2649 pipelining has not been advertised.
2651 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2653 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2654 returned in the SMTP response when smtp_return_error_details was false.
2655 This has been fixed.
2657 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2658 instead of the C I/O that was originally supplied, because problems were
2659 reported on Solaris.
2661 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2662 Exim which did not show up earlier: it was assuming that a call to
2663 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2664 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2665 was a macro that became an assignment, so it seemed to work. This has
2666 changed to a proper function call with a void return, hence the compile
2667 error. Exim's code has been fixed.
2669 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2672 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2674 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2676 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2677 because it turns out to be too big or there is a timeout) there is no
2678 "Completed" line in the log. When some messages of this type were
2679 selected by exigrep, they were listed as "not completed". Others were
2680 picked up by some special patterns. I have improved the selection
2681 criteria to be more general.
2683 PH/38 The host_find_failed option in the manualroute router can now be set
2684 to "ignore", to completely ignore a host whose IP address cannot be
2685 found. If all hosts are ignored, the behaviour is controlled by the new
2686 host_all_ignored option.
2688 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2689 homing or because of multiple MX records with /mx) generated more than
2690 one IP address, and the following item turned out to be the local host,
2691 all the secondary addresses of the first item were incorrectly removed
2692 from the list, along with the local host and any following hosts (which
2693 is what is supposed to happen).
2695 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2696 whoever called Exim into the -H file. In the case of the daemon it was
2697 behaving confusingly. When first started, it used values for whoever
2698 started the daemon, but after a SIGHUP it used the Exim user (because it
2699 calls itself on a restart). I have changed the code so that it now always
2702 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2703 message are rejected with the same error (e.g. no authentication or bad
2704 sender address), and a DATA command is nevertheless sent (as can happen
2705 with PIPELINING or a stupid MUA), the error message that was given to the
2706 RCPT commands is included in the rejection of the DATA command. This is
2707 intended to be helpful for MUAs that show only the final error to their
2710 PH/42 Another patch from the Sieve maintainer.
2712 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2713 Eximstats - Fixed some broken HTML links and added missing column headers
2715 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2716 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2718 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2720 PH/43 Yet another patch from the Sieve maintainer.
2722 PH/44 I found a way to check for a TCP/IP connection going away before sending
2723 the response to the final '.' that terminates a message, but only in the
2724 case where the client has not sent further data following the '.'
2725 (unfortunately, this is allowed). However, in many cases there won't be
2726 any further data because there won't be any more messages to send. A call
2727 to select() can be used: if it shows that the input is "ready", there is
2728 either input waiting, or the socket has been closed. An attempt to read
2729 the next input character can distinguish the two cases. Previously, Exim
2730 would have sent an OK response which the client would never have see.
2731 This could lead to message repetition. This fix should cure that, at
2732 least in a lot of common cases.
2734 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2735 advertised in response to EHLO.
2741 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2742 fixed by 4.65/MH/01 (is this a record?) are fixed:
2744 (i) An empty string was always treated as zero by the numeric comparison
2745 operators. This behaviour has been restored.
2747 (ii) It is documented that the numeric comparison operators always treat
2748 their arguments as decimal numbers. This was broken in that numbers
2749 starting with 0 were being interpreted as octal.
2751 While fixing these problems I realized that there was another issue that
2752 hadn't been noticed. Values of message_size_limit (both the global option
2753 and the transport option) were treated as octal if they started with 0.
2754 The documentation was vague. These values are now always treated as
2755 decimal, and I will make that clear in the documentation.
2761 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2762 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2765 MH/01 Don't check that the operands of numeric comparison operators are
2766 integers when their expansion is in "skipping" mode (fixes bug
2767 introduced by 4.64-PH/07).
2769 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2770 child addresses, Exim now panics and dies. Previously, because the count
2771 is held in a short int, deliveries were likely to be lost. As such a
2772 large number of recipients for a single message is ridiculous
2773 (performance will be very, very poor), I have chosen to impose a limit
2774 rather than extend the field.
2780 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2781 leftover -K file (the existence of which was triggered by #402).
2782 While we were at it, introduced process PID as part of the -K
2783 filename. This should rule out race conditions when creating
2786 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2787 processing considerably. Previous code took too long for large mails,
2788 triggering a timeout which in turn triggers #401.
2790 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2791 in the DK code in transports.c. sendfile() is not really portable,
2792 hence the _LINUX specificness.
2794 TF/01 In the add_headers option to the mail command in an Exim filter,
2795 there was a bug that Exim would claim a syntax error in any
2796 header after the first one which had an odd number of characters
2799 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2800 callout verification, Exim cached a "reject" for the entire domain. This
2801 is correct for most verifications, but it is not correct for a recipient
2802 verification with use_sender or use_postmaster set, because in that case
2803 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2804 case of MAIL FROM:<> rejection from other early rejections (e.g.
2805 rejection of HELO). When verifying a recipient using a non-null MAIL
2806 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2807 Whatever the result of the callout, the value of the domain cache is
2808 left unchanged (for any other kind of callout, getting as far as trying
2809 RCPT means that the domain itself is ok).
2811 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2814 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2815 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2816 session, the daemon ignores ECONNECT errors and logs others; it now
2817 ignores EPIPE as well.
2819 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2820 (quoted-printable decoding).
2822 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2823 later the small subsequent patch to fix an introduced bug.
2825 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2827 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2829 PH/08 An error is now given if message_size_limit is specified negative.
2831 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2832 to be given (somewhat) arbitrary names.
2834 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2837 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2838 miscellaneous code fixes
2840 PH/10 Added the log_reject_target ACL modifier to specify where to log
2843 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2844 hostname. This is wrong, because it relates to the incoming message (and
2845 probably the interface on which it is arriving) and not to the outgoing
2846 callout (which could be using a different interface). This has been
2847 changed to use the value of the helo_data option from the smtp transport
2848 instead - this is what is used when a message is actually being sent. If
2849 there is no remote transport (possible with a router that sets up host
2850 addresses), $smtp_active_hostname is used.
2852 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2853 tweaks were necessary in order to get it to work (see also 21 below):
2854 (a) The code assumed that strncpy() returns a negative number on buffer
2855 overflow, which isn't the case. Replaced with Exim's string_format()
2857 (b) There were several signed/unsigned issues. I just did the minimum
2858 hacking in of casts. There is scope for a larger refactoring.
2859 (c) The code used strcasecmp() which is not a standard C function.
2860 Replaced with Exim's strcmpic() function.
2861 (d) The code set only $1; it now sets $auth1 as well.
2862 (e) A simple test gave the error "authentication client didn't specify
2863 service in request". It would seem that Dovecot has changed its
2864 interface. Fortunately there's a specification; I followed it and
2865 changed what the client sends and it appears to be working now.
2867 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2870 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2871 address A is aliased to B and C, where B exists and C does not. Without
2872 -v the output is "A verified" because verification stops after a
2873 successful redirection if more than one address is generated. However,
2874 with -v the child addresses are also verified. Exim was outputting "A
2875 failed to verify" and then showing the successful verification for C,
2876 with its parentage. It now outputs "B failed to verify", showing B's
2877 parentage before showing the successful verification of C.
2879 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2880 look up a TXT record in a specific list after matching in a combined
2883 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2884 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2885 they consult the DNS. I had assumed they would set it the way they
2886 wanted; and indeed my experiments on Linux seem to show that in some
2887 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2888 To be on the safe side, however, I have now made the interface to
2889 host_find_byname() similar to host_find_bydns(), with an argument
2890 containing the DNS resolver options. The host_find_byname() function now
2891 sets these options at its start, just as host_find_bydns() does. The smtp
2892 transport options dns_qualify_single and dns_search_parents are passed to
2893 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2894 of host_find_byname() use the default settings of RES_DEFNAMES
2895 (qualify_single) but not RES_DNSRCH (search_parents).
2897 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2898 spool_read_header() do less string testing, by means of a preliminary
2899 switch on the second character of optional "-foo" lines. (This is
2900 overdue, caused by the large number of possibilities that now exist.
2901 Originally there were few.) While I was there, I also converted the
2902 str(n)cmp tests so they don't re-test the leading "-" and the first
2903 character, in the hope this might squeeze out yet more improvement.
2905 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2906 flag allowing group syntax was set by the header_syntax check but not
2907 turned off, possible causing trouble later; (2) The flag was not being
2908 set at all for the header_verify test, causing "group"-style headers to
2909 be rejected. I have now set it in this case, and also caused header_
2910 verify to ignore an empty address taken from a group. While doing this, I
2911 came across some other cases where the code for allowing group syntax
2912 while scanning a header line wasn't quite right (mostly, not resetting
2913 the flag correctly in the right place). These bugs could have caused
2914 trouble for malformed header lines. I hope it is now all correct.
2916 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2917 with the "reply" argument non-NULL. The code, however (which originally
2918 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2919 but it didn't always do it. This confused somebody who was copying the
2920 code for some other use. I have removed all the tests.
2922 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2923 feature that was used to support insecure browsers during the U.S. crypto
2924 embargo. It requires special client support, and Exim is probably the
2925 only MTA that supported it -- and would never use it because real RSA is
2926 always available. This code has been removed, because it had the bad
2927 effect of slowing Exim down by computing (never used) parameters for the
2928 RSA_EXPORT functionality.
2930 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2931 authenticator to fail if there's a tab character in the incoming data
2932 (there should never be unless someone is messing about, as it's supposed
2933 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2934 if the connection is using TLS or if the remote IP is the same as the
2935 local IP, and the "valid-client-cert option" if a client certificate has
2938 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2939 authenticators. This can be used for authorization after authentication
2940 succeeds. (In the case of plaintext, it servers for both authentication
2943 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2944 if any retry times were supplied.
2946 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2947 connection, where there is no client IP address to check. In this
2948 situation, the verify now always succeeds.
2950 PH/25 Applied John Jetmore's -Mset patch.
2952 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2954 PH/27 In a string expansion for a processed (not raw) header when multiple
2955 headers of the same name were present, leading whitespace was being
2956 removed from all of them, but trailing whitespace was being removed only
2957 from the last one. Now trailing whitespace is removed from each header
2958 before concatenation. Completely empty headers in a concatenation (as
2959 before) are ignored.
2961 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2962 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2964 PH/29 [Removed. This was a change that I later backed out, and forgot to
2965 correct the ChangeLog entry (that I had efficiently created) before
2966 committing the later change.]
2968 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2969 address errors (4xx response to RCPT) over the same connection as other
2970 messages routed to the same hosts. Such deliveries are always "forced",
2971 so retry times are not inspected. This resulted in far too many retries
2972 for the affected addresses. The effect occurred only when there were more
2973 hosts than the hosts_max_try setting in the smtp transport when it had
2974 the 4xx errors. Those hosts that it had tried were not added to the list
2975 of hosts for which the message was waiting, so if all were tried, there
2976 was no problem. Two fixes have been applied:
2978 (i) If there are any address or message errors in an SMTP delivery, none
2979 of the hosts (tried or untried) are now added to the list of hosts
2980 for which the message is waiting, so the message should not be a
2981 candidate for sending over the same connection that was used for a
2982 successful delivery of some other message. This seems entirely
2983 reasonable: after all the message is NOT "waiting for some host".
2984 This is so "obvious" that I'm not sure why it wasn't done
2985 previously. Hope I haven't missed anything, but it can't do any
2986 harm, as the worst effect is to miss an optimization.
2988 (ii) If, despite (i), such a delivery is accidentally attempted, the
2989 routing retry time is respected, so at least it doesn't keep
2990 hammering the server.
2992 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2993 in ${readsocket because some servers need this prod.
2995 PH/32 Added some extra debug output when updating a wait-xxx database.
2997 PH/33 The hint "could be header name not terminated by colon", which has been
2998 given for certain expansion errors for a long time, was not being given
2999 for the ${if def:h_colon_omitted{... case.
3001 PH/34 The spec says: "With one important exception, whenever a domain list is
3002 being scanned, $domain contains the subject domain." There was at least
3003 one case where this was not true.
3005 PH/35 The error "getsockname() failed: connection reset by peer" was being
3006 written to the panic log as well as the main log, but it isn't really
3007 panic-worthy as it just means the connection died rather early on. I have
3008 removed the panic log writing for the ECONNRESET error when getsockname()
3011 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3012 runs only) independently of the message's sender address. This meant
3013 that, if the 4xx error was in fact related to the sender, a different
3014 message to the same recipient with a different sender could confuse
3015 things. In particular, this can happen when sending to a greylisting
3016 server, but other circumstances could also provoke similar problems.
3017 I have changed the default so that the retry time for these errors is now
3018 based a combination of the sender and recipient addresses. This change
3019 can be overridden by setting address_retry_include_sender=false in the
3022 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3023 remote server are returned as part of bounce messages. This was not
3024 happening for LMTP over a pipe (the lmtp transport), but now it is the
3025 same for both kinds of LMTP.
3027 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3028 in header lines that were in fact CNAMEs. This is no longer the case.
3030 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3031 and queue runs started by the daemon processed all messages. This has
3032 been fixed so that -R and -S can now usefully be given with -q<time>.
3034 PH/40 Import PCRE release 6.7 (fixes some bugs).
3036 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3038 PH/42 Give an error if -q is specified more than once.
3040 PH/43 Renamed the variables $interface_address and $interface_port as
3041 $received_ip_address and $received_port, to make it clear that these
3042 values apply to message reception, and not to the outgoing interface when
3043 a message is delivered. (The old names remain recognized, of course.)
3045 PH/44 There was no timeout on the connect() call when using a Unix domain
3046 socket in the ${readsocket expansion. There now is.
3048 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3049 be meaningful with "accept".
3051 SC/01 Eximstats V1.43
3052 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3054 SC/02 Eximstats V1.44
3055 Use a glob alias rather than an array ref in the generated
3056 parser. This improves both readability and performance.
3058 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3059 Collect SpamAssassin and rejection statistics.
3060 Don't display local sender or destination tables unless
3061 there is data to show.
3062 Added average volumes into the top table text output.
3064 SC/04 Eximstats V1.46
3065 Collect data on the number of addresses (recipients)
3066 as well as the number of messages.
3068 SC/05 Eximstats V1.47
3069 Added 'Message too big' to the list of mail rejection
3070 reasons (thanks to Marco Gaiarin).
3072 SC/06 Eximstats V1.48
3073 Mainlog lines which have GMT offsets and are too short to
3074 have a flag are now skipped.
3076 SC/07 Eximstats V1.49 (Alain Williams)
3077 Added the -emptyok flag.
3079 SC/08 Eximstats V1.50
3080 Fixes for obtaining the IP address from reject messages.
3082 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3083 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3084 whitespace changes from 4.64-PH/27
3086 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3089 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3090 are found, allow negative numbers in numeric criteria)
3092 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3094 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3095 to match changes made in 4.64-PH/43
3097 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3099 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3100 "domain" is actually an IP address, and force a failure. This locks out
3101 those revolvers/nameservers that support "A-for-A" lookups, in
3102 contravention of the specifications.
3104 PH/48 When a host name was looked up from an IP address, and the subsequent
3105 forward lookup of the name timed out, the host name was left in
3106 $sender_host_name, contrary to the specification.
3108 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3109 restricted to single-key lookups, Exim was not diagnosing an error if
3110 * or *@ was used with a query-style lookup.
3112 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3114 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3115 long ago, but noone interested enough thought of it. Let's just say that
3116 the "1.1" means that there are some new functions that weren't there at
3117 some point in the past.
3119 PH/51 Error processing for expansion failure of helo_data from an smtp
3120 transport during callout processing was broken.
3122 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3123 tested/used via the -bh/-bhc/-bs options.
3125 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3126 bug, fixed in subsequent PCRE releases).
3128 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3129 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3131 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3137 SC/01 Use a glob alias rather than an array ref in eximstats generated
3138 parser. This improves both readability and performance.
3140 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3141 Don't display local sender or destination tables in eximstats unless
3142 there is data to show.
3143 Added average volumes into the eximstats top table text output.
3145 SC/03 Collect data on the number of addresses (recipients) as well
3146 as the number of messages in eximstats.
3148 TF/01 Correct an error in the documentation for the redirect router. Exim
3149 does (usually) call initgroups() when daemonizing.
3151 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3152 with consistent privilege compared to when running as a daemon.
3154 TF/03 Note in the spec that $authenticated_id is not set for local
3155 submissions from trusted users.
3157 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3158 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3160 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3161 by adding some example configuration directives to the default
3162 configuration file. A little bit of work is required to uncomment the
3163 directives and define how usernames and passwords are checked, but
3164 there is now a framework to start from.
3166 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3167 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3168 without this. I don't know how relevant this is to other LDAP libraries.
3170 PH/02 Add the verb name to the "unknown ACL verb" error.
3172 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3174 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3176 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3177 directory not expanded when it should be if an expanded home directory
3178 was set for the address (which is overridden by the transport).
3180 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3183 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3184 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3185 because it is too late at that time, and has no effect.
3187 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3188 security issue with \' (bugzilla #107). I could not use the
3189 PQescapeStringConn() function, because it needs a PGconn value as one of
3192 PH/08 When testing addresses using -bt, indicate those final addresses that
3193 are duplicates that would not cause an additional delivery. At least one
3194 person was confused, thinking that -bt output corresponded to deliveries.
3195 (Suppressing duplicates isn't a good idea as you lose the information
3196 about possibly different redirections that led to the duplicates.)
3198 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3199 systems where poll() doesn't work, in particular OS X.
3201 PH/10 Added more information to debugging output for retry time not reached.
3203 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3204 operations in malware.c.
3206 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3209 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3210 syslog_duplication set false, log lines that would normally be written
3211 both the the main log and to the reject log were not written to syslog at
3214 PH/14 In the default configuration, change the use of "message" in ACL warn
3215 statements to "add_header".
3217 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3218 not followed by a command (e.g. "seen endif").
3220 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3221 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3224 PH/17 Added extra conditions to the default value of delay_warning_condition
3228 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3229 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3230 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3233 The Auto-Submitted: and various List- headers are standardised, whereas I
3234 don't think Precedence: ever was.
3236 PH/18 Refactored debugging code in route_finduser() to show more information,
3237 in particular, the error code if getpwnam() issues one.
3239 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3240 This is apparently needed in addition to the PH/07 change above to avoid
3241 any possible encoding problems.
3243 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3244 but not after initializing Perl.
3246 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3247 output them only if debugging. By default they are written stderr,
3248 apparently, which is not desirable.
3250 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3253 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3256 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3258 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3259 authenticated or an ident call has been made. Suppress the default
3260 values for $authenticated_id and $authenticated_sender (but permit -oMai
3261 and -oMas) when testing with -bh.
3263 PH/24 Re-jigged the order of the tests in the default configuration so that the
3264 tests for valid domains and recipients precede the DNS black list and CSA
3265 tests, on the grounds that those ones are more expensive.
3267 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3268 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3269 This bug exists in every version of Exim that I still have, right back to
3272 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3273 However, an attempt to turn on case-sensitivity in a regex key by
3274 including (?-i) didn't work because the subject string was already
3275 lowercased, and the effects were non-intuitive. It turns out that a
3276 one-line patch can be used to allow (?-i) to work as expected.
3282 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3283 other effects) broke the use of negated acl sub-conditions.
3285 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3288 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3289 "Deny" causes Exim to reject the incoming connection with a 554 error.
3290 Unfortunately, if there is a major crisis, such as a disk failure,
3291 tcp-wrappers gives "deny", whereas what one would like would be some
3292 kind of temporary error. A kludge has been added to help with this.
3293 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3294 554 error is used if errno is still zero or contains ENOENT (which occurs
3295 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3298 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3300 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3301 errors. Otherwise a message that provokes a temporary error (when other
3302 messages do not) can cause a whole host to time out.
3304 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3305 addresses were routed directly to files or pipes from a redirect router.
3306 File deliveries just didn't batch; pipe deliveries might have suffered
3309 PH/06 A failure to get a lock for a hints database would erroneously always say
3310 "Failed to get write lock", even when it was really a read lock.
3312 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3313 of 0600. This has been changed to use the value of the lockfile_mode
3314 option (which defaults to 0600).
3316 PH/08 Applied small patch from the Sieve maintainer.
3318 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3319 folder from quota calculations, a direct delivery into this folder messed
3320 up the contents of the maildirsize file. This was because the regex was
3321 used only to exclude .Trash (or whatever) when the size of the mailbox
3322 was calculated. There was no check that a delivery was happening into an
3323 excluded directory. This bug has been fixed by ignoring all quota
3324 processing for deliveries into excluded directories.
3326 PH/10 Added the maildirfolder_create_regex option to appendfile.
3332 PH/01 The code for finding all the local interface addresses on a FreeBSD
3333 system running IPv6 was broken. This may well have applied to all BSD
3334 systems, as well as to others that have similar system calls. The broken
3335 code found IPv4 interfaces correctly, but gave incorrect values for the
3336 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3337 that it would not match correctly against @[] and not recognize the IPv6
3340 PH/02 The ipliteral router was not recognizing addresses of the form user@
3341 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3343 PH/03 Added disable_ipv6.
3345 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3346 decoded form, because it is most often used to construct To: headers
3347 lines in autoreplies, and the decoded form may well be syntactically
3348 invalid. However, $reply_address has leading white space removed, and all
3349 newlines turned into spaces so that the autoreply transport does not
3352 PH/05 If group was specified without a user on a router, and no group or user
3353 was specified on a transport, the group from the router was ignored.
3355 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3356 for visible compile-time settings that can be used to change these
3357 numbers, for those that want even more. Backwards compatibility with old
3358 spool files has been maintained. However, going back to a previous Exim
3359 release will lost any variables that are in spool files.
3361 PH/07 Two small changes when running in the test harness: increase delay when
3362 passing a TCP/IP connection to a new process, in case the original
3363 process has to generate a bounce, and remove special handling of
3364 127.0.0.2 (sic), which is no longer necessary.
3366 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3367 be the same on different OS.
3369 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3372 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3373 whether --show-vars was specified or not
3375 JJ/02 exipick: Added support for new ACL variable spool format introduced
3378 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3379 syntactically invalid From: or Reply-to: line, and a filter used this to
3380 generate an autoreply, and therefore failed to obtain an address for the
3381 autoreply, Exim could try to deliver to a non-existent relative file
3382 name, causing unrelated and misleading errors. What now happens is that
3383 it logs this as a hard delivery error, but does not attempt to create a
3386 PH/11 The exinext utility has a -C option for testing purposes, but although
3387 the given file was scanned by exinext itself; it wasn't being passed on
3388 when Exim was called.
3390 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3391 an end-of-file indication when reading a command response.
3393 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3394 compiled. In many other places in Exim, IPv6 addresses are always
3395 recognized, so I have changed this. It also means that IPv4 domain
3396 literals of the form [IPV4:n.n.n.n] are now always recognized.
3398 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3399 used if the router is not running as root, for example, when verifying at
3400 ACL time, or when using -bh. The debugging output from this situation was
3401 non-existent - all you got was a failure to exec. I have made two
3404 (a) Failures to set uid/gid, the current directory, or a process leader
3405 in a subprocess such as that created by queryprogram now generate
3406 suitable debugging output when -d is set.
3408 (b) The queryprogram router detects when it is not running as root,
3409 outputs suitable debugging information if -d is set, and then runs
3410 the subprocess without attempting to change uid/gid.
3412 PH/15 Minor change to Makefile for building test_host (undocumented testing
3415 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3416 additional section of a DNS packet that returns MX or SRV records.
3417 Instead, it always explicitly searches for A/AAAA records. This avoids
3418 major problems that occur when a DNS server includes only records of one
3419 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3420 fixed another bug: if SRV records were looked up and the corresponding
3421 address records were *not* found in the additional section, the port
3422 values from the SRV records were lost.
3424 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3425 using the correct key (the original address) when searching the retry
3426 rules in order to find which one to use for generating the retry hint.
3428 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3429 adding the default one. Similarly, if it contains a Reply-To: header, the
3430 errors_reply_to option, if set, is not used.
3432 PH/19 When calculating a retry time, Exim used to measure the "time since
3433 failure" by looking at the "first failed" field in the retry record. Now
3434 it does not use this if it is later than than the arrival time of the
3435 message. Instead it uses the arrival time. This makes for better
3436 behaviour in cases where some deliveries succeed, thus re-setting the
3437 "first failed" field. An example is a quota failure for a huge message
3438 when small messages continue to be delivered. Without this change, the
3439 "time since failure" will always be short, possible causing more frequent
3440 delivery attempts for the huge message than are intended.
3441 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3443 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3444 $1, $2, $3) because the numerical variables can be reset during some
3445 expansion items (e.g. "match"), thereby losing the authentication data.
3447 PH/21 Make -bV show the size of off_t variables so that the test suite can
3448 decide whether to run tests for quotas > 2G.
3450 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3451 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3452 filecount value is greater than 2G or if a quota value is greater than 2G
3453 on a system where the size of off_t is not greater than 4, a panic error
3456 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3457 never match. The debug and -bh output now contains an explicit error
3458 message indicating a malformed IPv4 address or mask.
3460 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3461 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3462 PH/23 above applies.
3464 PH/25 Do not write to syslog when running in the test harness. The only
3465 occasion when this arises is a failure to open the main or panic logs
3466 (for which there is an explicit test).
3468 PH/26 Added the /no_tell option to "control=freeze".
3470 PH/27 If a host name lookup failed very early in a connection, for example, if
3471 the IP address matched host_lookup and the reverse lookup yielded a name
3472 that did not have a forward lookup, an error message of the form "no IP
3473 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3474 could be logged. Now it outputs the IP address instead of "NULL".
3476 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3477 allows the sender and the authenticated sender to be set when
3478 submitting a message from within Exim. Since child_open_exim() is
3479 documented for local_scan(), the new function should be too.
3481 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3482 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3483 results in an empty string is now treated as unset.
3485 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3487 PH/31 Added sender_verify_fail logging option.
3489 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3490 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3491 tidied the source and removed it altogether.
3493 PH/33 When a queue run was abandoned because the load average was too high, a
3494 log line was always written; now it is written only if the queue_run log
3495 selector is set. In addition, the log line for abandonment now contains
3496 information about the queue run such as the pid. This is always present
3497 in "start" and "stop" lines but was omitted from the "abandon" line.
3499 PH/34 Omit spaces between a header name and the colon in the error message that
3500 is given when verify = headers_syntax fails (if there are lots of them,
3501 the message gets confusing).
3503 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3504 names, as there are now some PTR records that contain slashes. This check
3505 is only to protect against broken name servers that fall over on strange
3506 characters, so the fact that it applies to all lookups doesn't matter.
3508 PH/36 Now that the new test suite is complete, we can remove some of the
3509 special code in Exim that was needed for the old test suite. For example,
3510 sorting DNS records because real resolvers return them in an arbitrary
3511 order. The new test suite's fake resolver always returns records in the
3514 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3515 bounces) except when queue_only is set, to avoid logging races between
3516 the different processes.
3518 PH/38 Panic-die if .include specifies a non-absolute path.
3520 PH/39 A tweak to the "H" retry rule from its user.
3522 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3523 a label. They prevented compilation on older perls.
3525 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3526 a warning to be raised on newish perls.
3528 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3529 on queue. Changes to match documented behaviour of showing count of
3530 messages matching specified criteria.
3532 PH/40 Changed the default ident timeout from 30s to 5s.
3534 PH/41 Added support for the use of login_cap features, on those BSD systems
3535 that have them, for controlling the resources used by pipe deliveries.
3537 PH/42 The content-scanning code uses fopen() to create files in which to put
3538 message data. Previously it was not paying any attention to the mode of
3539 the files. Exim runs with umask(0) because the rest of the code creates
3540 files with open(), and sets the required mode explicitly. Thus, these
3541 files were ending up world-writeable. This was not a big issue, because,
3542 being within the spool directory, they were not world-accessible. I have
3543 created a function called modefopen, which takes an additional mode
3544 argument. It sets umask(777), creates the file, chmods it to the required
3545 mode, then resets the umask. All the relevant calls to fopen() in the
3546 content scanning code have been changed to use this function.
3548 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3549 to 24 hours. This avoids potential overflow problems when processing G
3550 and H retry rules. I suspect nobody ever tinkers with this value.
3552 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3554 PH/45 When the plaintext authenticator is running as a client, the server's
3555 challenges are checked to ensure they are valid base64 strings. By
3556 default, the authentication attempt is cancelled if an invalid string is
3557 received. Setting client_ignore_invalid_base64 true ignores these errors.
3558 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3559 they are received. Thus, the responses can be made to depend on the
3560 challenges. If an invalid string is ignored, an empty string is placed in
3563 PH/46 Messages that are created by the autoreply transport now contains a
3564 References: header, in accordance with RFCs 2822 and 3834.
3566 PH/47 Added authenticated_sender_force to the smtp transport.
3568 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3570 PH/49 Installed latest patch from the Sieve maintainer.
3572 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3573 was also set, the appendfile transport was unnecessarily scanning a
3574 directory of message files (e.g. for maildir delivery) to find the count
3575 of files (along with the size), even though it did not need this
3576 information. It now does the scan only if it needs to find either the
3577 size of the count of files.
3579 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3581 PH/52 Two bugs concerned with error handling when the smtp transport is
3584 (i) Exim was not creating retry information for temporary errors given
3585 for individual recipients after the DATA command when the smtp transport
3586 was used in LMTP mode. This meant that they could be retried too
3587 frequently, and not timed out correctly.
3589 (ii) Exim was setting the flag that allows error details to be returned
3590 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3591 recipients that were returned after the DATA command.
3593 PH/53 This is related to PH/52, but is more general: for any failing address,
3594 when detailed error information was permitted to be returned to the
3595 sender, but the error was temporary, then after the final timeout, only
3596 "retry timeout exceeded" was returned. Now it returns the full error as
3597 well as "retry timeout exceeded".
3599 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3600 do this, and (what is worse) MTAs that accept it.
3602 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3603 will now be deprecated.
3605 PH/56 New os.c-cygwin from the Cygwin maintainer.
3607 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3608 formats (previously only available in exim formats via -bpr, -bpru,
3609 and -bpra. Now also available in native and exiqgrep formats)
3611 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3612 with very large, slow to parse queues
3614 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3616 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3618 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3619 responses to authentication challenges, though it was showing the
3620 challenges; (ii) I've removed the CR characters from the debug output for
3623 PH/58 Allow for the insertion of a newline as well as a space when a string
3624 is turned into more than one encoded-word during RFC 2047 encoding. The
3625 Sieve code now uses this.
3627 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3628 data_4xx, lost_connection, tls_required.
3630 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3631 message was being sent as an SMTP response.
3633 PH/61 Add -l and -k options to exicyclog.
3635 PH/62 When verifying, if an address was redirected to one new address, so that
3636 verification continued, and the new address failed or deferred after
3637 having set something in $address_data, the value of $address_data was not
3638 passed back to the ACL. This was different to the case when no
3639 redirection occurred. The value is now passed back in both cases.
3641 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3642 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3643 use login_cap.h, so on its own it isn't the distinguishing feature. The
3644 new name refers directly to the setclassresources() function.
3646 PH/65 Added configuration files for NetBSD3.
3648 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3650 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3651 is preferred over IPv4.
3653 PH/68 The bounce_return_message and bounce_return_body options were not being
3654 honoured for bounces generated during the reception of non-SMTP messages.
3655 In particular, this applied to messages rejected by the ACL. This bug has
3656 been fixed. However, if bounce_return_message is true and bounce_return_
3657 body is false, the headers that are returned for a non-SMTP message
3658 include only those that have been read before the error was detected.
3659 (In the case of an ACL rejection, they have all been read.)
3661 PH/69 The HTML version of the specification is now built in a directory called
3662 spec_html instead of spec.html, because the latter looks like a path with
3663 a MIME-type, and this confuses some software.
3665 PH/70 Catch two compiler warnings in sieve.c.
3667 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3668 function verify_get_ident() calls ip_connect() to connect a socket, but
3669 if the "connect()" function timed out, ip_connect() used to close the
3670 socket. However, verify_get_ident() also closes the socket later, and in
3671 between Exim writes to the log, which may get opened at this point. When
3672 the socket was closed in ip_connect(), the log could get the same file
3673 descriptor number as the socket. This naturally causes chaos. The fix is
3674 not to close the socket in ip_connect(); the socket should be closed by
3675 the function that creates it. There was only one place in the code where
3676 this was missing, in the iplookup router, which I don't think anybody now
3677 uses, but I've fixed it anyway.
3679 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3680 well as to direct DNS lookups. Otherwise the handling of names in host
3681 lists is inconsistent and therefore confusing.
3687 PH/01 Two changes to the default runtime configuration:
3689 (1) Move the checks for relay_from_hosts and authenticated clients from
3690 after to before the (commented out) DNS black list checks.
3692 (2) Add control=submission to the relay_from_hosts and authenticated
3693 clients checks, on the grounds that messages accepted by these
3694 statements are most likely to be submissions.
3696 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3698 (1) Generate an error if the third argument for the ${prvs expansion is
3701 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3704 (3) Reset the variables that are obtained from the first argument of
3705 ${prvscheck and used in the second argument before leaving the code,
3706 because their memory is reclaimed, so using them afterwards may do
3709 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3710 one (it's much easier than Tom thought :-).
3712 (5) Because of (4), we can now allow for the use of $prvscheck_result
3713 inside the third argument.
3715 PH/03 For some reason, the default setting of PATH when running a command from
3716 a pipe transport was just "/usr/bin". I have changed it to
3719 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3720 anything to be listed in the output from -bV.
3722 PH/05 When a filter generated an autoreply, the entire To: header line was
3723 quoted in the delivery log line, like this:
3725 => >A.N.Other <ano@some.domain> <original@ddress> ...
3727 This has been changed so that it extracts the operative address. There
3728 may be more than one such address. If so, they are comma-separated, like
3731 => >ano@some.domain,ona@other.domain <original@ddress> ...
3733 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3734 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3735 not being looked up in the rDNS to get a host name, Exim was showing the
3736 IP address twice in Received: lines, even though the IP addresses were
3737 identical. For example:
3739 Received: from [1.2.3.4] (helo=[1.2.3.4])
3741 However, if the real host name was known, it was omitting the HELO data
3742 if it matched the actual IP address. This has been tidied up so that it
3743 doesn't show the same IP address twice.
3745 PH/07 When both +timestamp and +memory debugging was on, the value given by
3746 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3747 called by the debug printing, thereby overwriting the timestamp buffer.
3748 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3750 PH/08 When the original message was included in an autoreply transport, it
3751 always said "this is a copy of the message, including all the headers",
3752 even if body_only or headers_only was set. It now gives an appropriate
3755 PH/09 Applied a patch from the Sieve maintainer which:
3757 o fixes some comments
3758 o adds the (disabled) notify extension core
3759 o adds some debug output for the result of if/elsif tests
3760 o points to the current vacation draft in the documentation
3761 and documents the missing references header update
3765 o fixes a bug in processing the envelope test (when testing
3766 multiple envelope elements, the last element determined the
3769 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3770 Electronic Mail") by including:
3772 Auto-submitted: auto-generated
3774 in the messages that it generates (bounce messages and others, such as
3775 warnings). In the case of bounce messages for non-SMTP messages, there was
3776 also a typo: it was using "Auto_submitted" (underscore instead of
3777 hyphen). Since every message generated by Exim is necessarily in response
3778 to another message, thes have all been changed to:
3780 Auto-Submitted: auto-replied
3782 in accordance with these statements in the RFC:
3784 The auto-replied keyword:
3786 - SHOULD be used on messages sent in direct response to another
3787 message by an automatic process,
3789 - MUST NOT be used on manually-generated messages,
3791 - MAY be used on Delivery Status Notifications (DSNs) and Message
3792 Disposition Notifications (MDNs),
3794 - MUST NOT be used on messages generated by automatic or periodic
3795 processes, except for messages which are automatic responses to
3798 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3799 to the default Received: header definition.
3801 PH/12 Added log selector acl_warn_skipped (default on).
3803 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3804 variables because (a) they are in the wrong storage pool and (b) even if
3805 they were copied, it wouldn't work properly because of the caching.
3807 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3808 checking when decoding. Apparently there are clients that generate
3809 overlong encoded strings. Why am I not surprised?
3811 PH/15 If the first argument of "${if match_address" was not empty, but did not
3812 contain an "@" character, Exim crashed. Now it writes a panic log message
3813 and treats the condition as false.
3815 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3817 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3818 extension "envelope-auth". The code is finished and in agreement with
3819 other implementations, but there is no documentation so far and in fact,
3820 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3821 not changing the active code.
3823 Print executed "if" and "elsif" statements when debugging is used. This
3824 helps a great deal to understand what a filter does.
3826 Document more things not specified clearly in RFC3028. I had all this
3827 sorted out, when out of a sudden new issues came to my mind. Oops."
3829 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3832 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3833 canonical form (as documented). However, after a host name lookup from
3834 the IP address, check_host() was doing a simple string comparison with
3835 addresses acquired from the DNS when checking that the found name did
3836 have the original IP as one of its addresses. Since any found IPv6
3837 addresses are likely to be in abbreviated form, the comparison could
3838 fail. Luckily, there already exists a function for doing the comparison
3839 by converting both addresses to binary, so now that is used instead of
3840 the text comparison.
3842 PH/20 There was another similar case to PH/19, when a complete host name was
3843 given in a host list; looking up its IP address could give an abbreviated
3844 form, whereas the current host's name might or might not be abbreviated.
3845 The same fix has been applied.
3851 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3852 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3855 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3856 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3858 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3860 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3861 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3862 Also, for "Auto-Submitted", treat anything other than "no" as
3863 non-personal, in accordance with RFC 3834. (Previously it treated
3864 anything starting "auto-" as non-personal.)
3866 TF/01 The control=submission/name=... option had a problem with syntax
3867 errors if the name included a slash character. The /name= option
3868 now slurps the rest of the string, so it can include any characters
3869 but it must come last in the list of options (after /sender_retain
3872 PH/05 Some modifications to the interface to the fake nameserver for the new
3880 TK/01 Added the "success_on_redirect" address verification option. See
3881 NewStuff for rationale and an example.
3883 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3885 PH/02 Patch to exigrep to allow it to work on syslog lines.
3887 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3888 fread() to skip over the body file's header line, because in Cygwin the
3889 header line is locked and is inaccessible.
3891 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3892 co-exist for some time) to make it clear that it is the Exim ID that is
3893 referenced, not the Message-ID: header line.
3895 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3896 string_format() function, because snprintf() does not exist on all
3899 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3900 ${stat: expansion item.
3902 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3903 protocol synchronization error", to keep the pedants happy.
3905 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3906 well as for IRIX systems, when gcc is being used. See the host.c source
3909 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3911 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3914 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3915 transport and to the smtp transport in LMTP mode.
3917 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3919 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3920 run a filter in a subprocess. This could lead to confusion in subsequent
3921 lookups in the parent process. There should also be a search_tidyup() at
3922 the end of the subprocess.
3924 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3925 only if the host matched helo_try_verify_hosts, which caused the
3926 verification to occur when the EHLO/HELO command was issued. The ACL just
3927 tested the remembered result. Now, if a previous verification attempt has
3928 not happened, "verify = helo" does it there and then.
3930 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3932 TK/03 Fix log output including CR from clamd.
3934 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3935 exist provoked a memory error which could cause a segfault.
3937 PH/15 Installed PCRE 6.2
3939 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3941 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3942 of the problem. Specifically, suggested +O2 rather than +O1 for the
3945 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3947 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3948 record, the port was not correctly logged when the outgoing_port log
3949 selector was set (it logged the transort's default port).
3951 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3952 fallback_hosts, and "hosts" in the smtp transport.
3954 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3955 host errors such as "Connection refused".
3957 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3958 authentication with radiusclient 0.4.9:
3960 - Error returned from rc_read_config was caught wrongly
3961 - Username/password not passed on to radius server due to wrong length.
3963 The presumption is that some radiusclient API changes for 4.51/PH/17
3964 were not taken care of correctly. The code is still untested by me (my
3965 Linux distribution still has 0.3.2 of radiusclient), but it was
3966 contributed by a Radius user.
3968 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3969 expanding the "port" option of the smtp transport.
3971 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3972 while reading a MIME header. Thanks to Tom Hughes for a patch.
3974 PH/24 Include config.h inside local_scan.h so that configuration settings are
3977 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3978 that in an ACL for RCPT (for example), you can examine exactly what was
3981 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3982 commands, but it was not correctly comparing the address with the actual
3983 client host address. Thus, it would show the EHLO address in Received:
3984 header lines when this was not necessary.
3986 PH/27 Added the % operator to ${eval:}.
3988 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3989 it should be ignoring failures (because with -C, for example, it has lost
3990 privilege). It wasn't ignoring creation failures other than "already
3993 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3996 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3997 because an input line was too long, either on its own, or by virtue of
3998 too many continuations, the temporary file was not being removed, and the
3999 return code was incorrect.
4001 PH/31 Missing "BOOL" in function definition in filtertest.c.
4003 PH/32 Applied Sieve patches from the maintainer.
4005 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4007 PH/33 Added "verify = not_blind".
4009 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4010 Local/Makefile (with some defaults set). These are used in built scripts
4011 such as exicyclog, but they have never been used in the exim_install
4012 script (though there are many overriding facilities there). I have
4013 arranged that the exim_install script now takes note of these two
4016 PH/35 Installed configuration files for Dragonfly.
4018 PH/36 When a locally submitted message by a trusted user did not contain a
4019 From: header, and the sender address was obtained from -f or from an SMTP
4020 MAIL command, and the trusted user did not use -F to supply a sender
4021 name, $originator_name was incorrectly used when constructing a From:
4022 header. Furthermore, $originator_name was used for submission mode
4023 messages from external hosts without From: headers in a similar way,
4024 which is clearly wrong.
4026 PH/37 Added control=suppress_local_fixups.
4028 PH/38 When log_selector = +received_sender was set, and the addition of the
4029 sender made the log line's construction buffer exactly full, or one byte
4030 less than full, an overflow happened when the terminating "\n" was
4033 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4034 when the result of a list match is failure because a DNS lookup failed.
4036 PH/40 RM_COMMAND is now used in the building process.
4038 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4039 the "build-* directories that it finds.
4041 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4042 address in a domain literal was a prefix of an interface address.
4044 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4045 when verifying a sender address, unless rewrite_headers is false.
4047 PH/44 Wrote a long comment about why errors_to addresses are verified as
4048 recipients, not senders.
4050 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4051 the ratelimit ACL was added.
4053 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4055 PH/46 Added extra information about PostgreSQL errors to the error string.
4057 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4058 avoiding the need to install special zones in a real server. This is
4059 backwards compatible; if it can't find the fake resolver, it drops back.
4060 Thus, both old and new test suites can be run.
4062 TF/02 Added util/ratelimit.pl
4064 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4065 clock is set back in time.
4067 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4068 Candler <B.Candler@pobox.com>.
4070 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4071 OK for addresses that are the result of redirections.
4073 PH/48 A number of further additions for the benefit of the new test suite,
4074 including a fake gethostbyname() that interfaces to the fake DNS resolver
4077 TF/06 The fix for widen_domains has also been applied to qualify_single and
4078 search_parents which are the other dnslookup options that can cause
4081 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4084 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4086 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4087 service" if it encountered an MX record with an empty target hostname.
4088 The message is now "an MX or SRV record indicated no SMTP service".
4090 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4091 if the DNS of the sending site is misconfigured. This is quite a
4092 common situation. This change restores the behaviour of treating a
4093 helo verification defer as a failure.
4095 PH/51 If self=fail was set on a router, the bounce message did not include the
4096 actual error message.
4102 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4104 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4105 command itself ended in error, the underlying message about the transport
4106 filter timeout was being overwritten with the pipe command error. Now the
4107 underlying error message should be appended to the second error message.
4109 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4111 PH/02 Reduce the amount of output that "make" produces by default. Full output
4112 can still be requested.
4114 PH/03 The warning log line about a condition test deferring for a "warn" verb
4115 was being output only once per connection, rather than after each
4116 occurrence (because it was using the same function as for successful
4117 "warn" verbs). This seems wrong, so I have changed it.
4119 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4120 it should not have, which might have caused a crash in the right
4121 circumstances, but probably never did.
4123 PH/04 Installed a modified version of Tony Finch's patch to make submission
4124 mode fix the return path as well as the Sender: header line, and to
4125 add a /name= option so that you can make the user's friendly name appear
4128 TF/03 Added the control = fakedefer ACL modifier.
4130 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4131 Mark Lowes for thorough testing.
4133 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4135 TK/03 Merged latest SRS patch from Miles Wilton.
4137 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4138 with the definition in sysexits.h (which is #included earlier).
4139 Fortunately, Exim does not actually use EX_OK. The code used to try to
4140 preserve the sysexits.h value, by assuming that macro definitions were
4141 scanned for macro replacements. I have been disabused of this notion,
4142 so now the code just undefines EX_OK before #including unistd.h.
4144 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4145 in the smtp transport. When a block could not be written in a single
4146 write() function, the timeout was being re-applied to each part-write.
4147 This seems wrong - if the receiver was accepting one byte at a time it
4148 would take for ever. The timeout is now adjusted when this happens. It
4149 doesn't have to be particularly precise.
4151 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4152 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4154 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4155 without a domain if the check to <postmaster@domain> fails.
4157 SC/01 Eximstats: added -xls and the ability to specify output files
4158 (patch written by Frank Heydlauf).
4160 SC/02 Eximstats: use FileHandles for outputting results.
4162 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4164 SC/04 Eximstats: fixed display of large numbers with -nvr option
4166 SC/05 Eximstats: fixed merging of reports with empty tables.
4168 SC/06 Eximstats: added the -include_original_destination flag
4170 SC/07 Eximstats: removed tabs and trailing whitespace.
4172 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4174 TK/06 MBOX spool code: Add real "From " MBOX separator line
4175 so the .eml file is really in mbox format (even though
4176 most programs do not really care). Patch from Alex Miller.
4178 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4179 The latter is generated from $received_to and is only set if the
4180 message has one envelope recipient. SA can use these headers,
4181 obviously out-of-the-box. Patch from Alex Miller.
4183 PH/08 The ${def test on a variable was returning false if the variable's
4184 value was "0", contrary to what the specification has always said!
4185 The result should be true unless the variable is empty.
4187 PH/09 The syntax error of a character other than { following "${if
4188 def:variable_name" (after optional whitespace) was not being diagnosed.
4189 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4190 accidental colon was present, for example, could give incorrect results.
4192 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4193 result is used (not including appendfile, where other changes are about
4196 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4197 This involved changing a lot of size variables from int to off_t. It
4198 should work with maildirs and everything.
4200 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4201 spamd dying while we are connected to it.
4203 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4206 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4207 function for BDB 4.3.
4209 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4211 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4212 the sources for PCRE have been re-arranged and more files are now
4215 PH/15 The code I had for printing potentially long long variables in PH/11
4216 above was not the best (it lost precision). The length of off_t variables
4217 is now inspected at build time, and an appropriate printing format (%ld
4218 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4219 to be "long long int" or "long int". This is needed for the internal
4220 formatting function string_vformat().
4222 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4223 the configuration file to be ":syslog", then the script "guesses" where
4224 the logs files are, rather than using the compiled in default. In our
4225 case the guess is not the same as the compiled default, so the script
4226 suddenly stopped working when I started to use syslog. The patch checks
4227 to see if log_file_path is "". If so, it attempts to read it from exim
4228 with no configuration file to get the compiled in version, before it
4229 falls back to the previous guessing code."
4231 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4232 implementing BATV in an Exim configuration. See NewStuff for the gory
4235 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4236 Makefile that are specific to HP-UX.
4238 PH/18 If the "use_postmaster" option was set for a recipient callout together
4239 with the "random" option, the postmaster address was used as the MAIL
4240 FROM address for the random test, but not for the subsequent recipient
4241 test. It is now used for both.
4243 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4244 patch removes a few documentation additions to RFC 3028, because the
4245 latest draft now contains them. It adds the new en;ascii-case comparator
4246 and a new error check for 8bit text in MIME parts. Comparator and
4247 require names are now matched exactly. I enabled the subaddress
4248 extension, but it is not well tested yet (read: it works for me)."
4250 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4251 rework some of the code of TK/09 above to avoid the hardwired use of
4252 "%lld" and "long long". Replaced the call to snprintf() with a call to
4255 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4256 records point to non-existent hosts", "retry timeout exceeded", and
4257 "retry time not reached for any host after a long failure period".
4259 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4260 experimental DomainKeys support:
4262 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4263 (2) On an error such as an illegally used "control", the wrong name for
4264 the control was given.
4266 These problems did NOT occur unless DomainKeys support was compiled.
4268 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4270 PH/24 Added ${if match_ip condition.
4272 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4273 databases so that it will be absolutely obvious if a crash occurs in the
4274 DB library. This is a regular occurrence (often caused by mis-matched
4277 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4278 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4279 picked up on a user's system that detects such things. There doesn't seem
4280 to be a gcc warning option for this - only an attribute that has to be
4281 put on the function's prototype. It seems that in Fedora Core 4 they have
4282 set this on a number of new functions. No doubt there will be more in due
4285 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4286 specify a transport. However, if an address that was verified by such a
4287 router was the subject of a callout, Exim crashed because it tried to
4288 read the rcpt_include_affixes from the non-existent transport. Now it
4289 just assumes that the setting of that option is false. This bug was
4290 introduced by 4.51/PH/31.
4292 PH/28 Changed -d+all to exclude +memory, because that information is very
4293 rarely of interest, but it makes the output a lot bigger. People tend to
4294 do -d+all out of habit.
4296 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4297 code in os-type was giving problems when libc.so lives in lib64, like on
4300 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4301 aren't the modern standard, and it seems that some systems' include files
4302 don't always have them. Exim was already checking for some of the newer
4303 ones like T_AAAA, and defining it itself. I've added checks for all the
4304 record types that Exim uses.
4306 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4307 not automatically generating a new one, as it is supposed to. This
4308 prevented TLS from working. If the file did exist, but contained invalid
4309 data, a new version was generated, as expected. It was only the case of a
4310 non-existent file that was broken.
4312 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4313 with a change in libdomainkeys > 0.64.
4315 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4316 from DNS. If the selector record carries the flag, it now has
4317 precedence over the domain-wide flag.
4319 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4321 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4322 the use of a port name that isn't defined in /etc/services) occurred, the
4323 message was deferred as in a normal delivery, and thus remained on the
4324 spool, instead of being failed because of the mua_wrapper setting. This
4325 is now fixed, and I tidied up some of the mua_wrapper messages at the
4328 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4329 the messages in a hash of arrays rather than using individual hashes.
4330 This is a bit cleaner and results in dramatic memory savings, albeit
4331 at a slight CPU cost.
4333 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4334 as requested by Marc Sherman.
4336 SC/10 Eximstats: added histograms for user specified patterns as requested
4339 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4341 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4342 fopen() in the content-scanning modules that did not already have it.
4348 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4349 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4351 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4353 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4355 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4356 to test Sieve filters that use "vacation".
4358 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4359 that changes the way the GnuTLS parameters are stored in the cache file.
4360 The new format can be generated externally. For backward compatibility,
4361 if the data in the cache doesn't make sense, Exim assumes it has read an
4362 old-format file, and it generates new data and writes a new file. This
4363 means that you can't go back to an older release without removing the
4366 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4367 work if there are any delivery delays because "one_time" forces the
4368 parent to be marked "delivered", so its unseen clone is never tried
4369 again. For this reason, Exim now forbids the simultaneous setting of
4372 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4373 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4374 entry if you want to know the details. The fix, however, neglected to
4375 consider the case when local delivery batching is involved. The test for
4376 "previously delivered" was not happening when checking to see if an
4377 address could be batched with a previous (undelivered) one; under
4378 certain circumstances this could lead to multiple deliveries to the same
4381 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4382 in its include files, and this causes problems building Exim.
4384 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4385 header_syntax) but Exim was just ignoring anything given after a slash.
4386 In particular, this caused confusion with an attempt to use "verify =
4387 reverse_host_lookup/defer_ok". An error is now given when options are
4388 supplied for verify items that do not have them. (Maybe reverse_host_
4389 lookup should have a defer_ok option, but that's a different point.)
4391 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4392 defined by RFC 821) to 2048, because there were problems with some AUTH
4393 commands, and RFC 1869 says the size should be increased for extended
4394 SMTP commands that take arguments.
4396 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4399 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4400 "unknown" error; now it says that the functionality isn't in the binary.
4402 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4403 an address' error message when a string expansion fails (syntax or
4404 whatever). Otherwise the password may appear in the log. Following change
4405 PH/42 below, there is no longer a chance of it appearing in a bounce
4408 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4410 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4411 of its addresses was ever tried. (Bugzilla bug #2.)
4413 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4414 the result incorrectly in the debug output. (It correctly added a newline
4415 to what was transported.)
4417 TF/01 Added $received_time.
4419 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4420 commented out examples of how to interface to a virus scanner and to
4421 SpamAssassin. Also added commented examples of av_scanner and
4422 spamd_address settings.
4424 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4425 and controls are allowed in which ACLs. There were a couple of minor
4426 errors. Some of the entries in the conditions table (which is a table of
4427 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4428 negation of where the condition IS allowed.
4430 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4432 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4433 header file does not have a version number, so I've had to invent a new
4434 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4435 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4436 radiusclient), but it was contributed by a Radius user.
4438 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4439 files or directories, for OpenSSL.
4441 PH/19 When an Exim process that is running as root has to create an Exim log
4442 file, it does so in a subprocess that runs as exim:exim so as to get the
4443 ownership right at creation (otherwise, other Exim processes might see
4444 the file with the wrong ownership). There was no test for failure of this
4445 fork() call, which would lead to the process getting stuck as it waited
4446 for a non-existent subprocess. Forks do occasionally fail when resources
4447 run out. I reviewed all the other calls to fork(); they all seem to check
4450 PH/20 When checking for unexpected SMTP input at connect time (before writing
4451 the banner), Exim was not dealing correctly with a non-positive return
4452 from the read() function. If the client had disconnected by this time,
4453 the result was a log entry for a synchronization error with an empty
4454 string after "input=" when read() returned zero. If read() returned -1
4455 (an event I could not check), uninitialized data bytes were printed.
4456 There were reports of junk text (parts of files, etc) appearing after
4459 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4461 PH/22 Added support for macro redefinition, and (re)definition in between
4462 driver and ACL definitions.
4464 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4465 forgetting to use the resulting value; it was using the unexpanded value.
4467 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4468 hadn't been configured. The fix is from Juergen Kreileder, who
4469 understands it better than I do:
4471 "Here's what I see happening with three configured cyrus_sasl
4472 authenticators configured (plain, login, cram-md5):
4474 On startup auth_cyrus_sasl_init() gets called for each of these.
4475 This means three calls to sasl_listmech() without a specified mech_list.
4476 => SASL tests which mechs of all available mechs actually work
4477 => three warnings about OTP not working
4478 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4480 With the patch, sasl_listmech() also gets called three times. But now
4481 SASL's mech_list option is set to the server_mech specified in the the
4482 authenticator. Or in other words, the answer from sasl_listmech()
4483 gets limited to just the mech you're testing for (which is different
4485 => the return list contains just 'plain' or 'login', 'cram-md5' or
4486 nothing depending on the value of ob->server_mech.
4488 I've just tested the patch: Authentication still works fine,
4489 unavailable mechs specified in the exim configuration are still
4490 caught, and the auth.log warnings about OTP are gone."
4492 PH/25 When debugging is enabled, the contents of the command line are added
4493 to the debugging output, even when log_selector=+arguments is not
4496 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4497 answer is "GNU", and only if the return is "GNU/something" is the answer
4500 PH/27 $acl_verify_message is now set immediately after the failure of a
4501 verification in an ACL, and so is available in subsequent modifiers. In
4502 particular, the message can be preserved by coding like this:
4504 warn !verify = sender
4505 set acl_m0 = $acl_verify_message
4507 Previously, $acl_verify_message was set only while expanding "message"
4508 and "log_message" when a very denied access.
4510 PH/28 Modified OS/os.c-Linux with
4512 -#ifndef OS_LOAD_AVERAGE
4513 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4515 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4516 nomenclature these days.)
4518 PH/29 Installed patch from the Sieve maintainer that adds the options
4519 sieve_useraddress and sieve_subaddress to the redirect router.
4521 PH/30 In these circumstances:
4522 . Two addresses routed to the same list of hosts;
4523 . First host does not offer TLS;
4524 . First host accepts first address;
4525 . First host gives temporary error to second address;
4526 . Second host offers TLS and a TLS session is established;
4527 . Second host accepts second address.
4528 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4529 and peerdn, if requested) that were in fact used only for the second
4532 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4533 attention to any local part prefix or suffix that was matched by the
4534 router that accepted the address. It now behaves in the same way as it
4535 does for delivery: the affixes are removed from the local part unless
4536 rcpt_include_affixes is set on the transport.
4538 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4539 timeout during the DATA phase of an incoming message.
4541 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4542 applied a patch sanctioned by the Sieve maintainer.
4544 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4545 the uid or gid is negative. A case of a negative gid caused this to be
4546 noticed. The fix allows for either to be negative.
4548 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4549 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4552 PH/36 The change PH/12 above was broken. Fixed it.
4554 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4555 the grounds that routing the same address twice would always produce the
4556 same answer. This might have been true once, but it is certainly no
4557 longer true now. Routing a child address may depend on the previous
4558 routing that produced that child. Some complicated redirection strategies
4559 went wrong when messages had multiple recipients, and made Exim's
4560 behaviour dependent on the order in which the addresses were given.
4562 I have moved the duplicate checking until after the routing is complete.
4563 Exim scans the addresses that are assigned to local and remote
4564 transports, and removes any duplicates. This means that more work will be
4565 done, as duplicates will always all be routed, but duplicates are
4566 presumably rare, so I don't expect this is of any significance.
4568 For deliveries to pipes, files, and autoreplies, the duplicate checking
4569 still happens during the routing process, since they are not going to be
4572 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4573 It corrects a timeout issue with spamd. This is Ian's comment: "The
4574 background is that sometimes spamd either never reads data from a
4575 connection it has accepted, or it never writes response data. The exiscan
4576 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4577 blindly assumes that writes won't block so it may never time out."
4579 PH/39 Allow G after quota size as well as K and M.
4581 PH/40 The value set for $authenticated_id in an authenticator may not contain
4582 binary zeroes or newlines because the value is written to log lines and
4583 to spool files. There was no check on this. Now the value is run through
4584 the string_printing() function so that such characters are converted to
4585 printable escape sequences.
4587 PH/41 $message_linecount is a new variable that contains the total number of
4588 lines in the message. Compare $body_linecount, which is the count for the
4591 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4592 bounce and delay warning messages, except in certain special cases, which
4595 (a) An SMTP error message from a remote host;
4596 (b) A message specified in a :fail: redirection;
4597 (c) A message specified in a "fail" command in a system filter;
4598 (d) A message specified in a FAIL return from the queryprogram router;
4599 (e) A message specified by the cannot_route_message router option.
4601 In these cases only, Exim does include the error details in bounce and
4602 warning messages. There are also a few cases where bland messages such
4603 as "unrouteable address" or "local delivery error" are given.
4605 PH/43 $value is now also set for the "else" part of a ${run expansion.
4607 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4608 being worked on, but at least Exim now implements the latest version to
4611 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4612 process to complete was treated as a delivery failure, a timeout while
4613 writing the message to the pipe was logged, but erroneously treated as a
4614 successful delivery. Such timeouts include transport filter timeouts. For
4615 consistency with the overall process timeout, these timeouts are now
4616 treated as errors, giving rise to delivery failures by default. However,
4617 there is now a new Boolean option for the pipe transport called
4618 timeout_defer, which, if set TRUE, converts the failures into defers for
4619 both kinds of timeout. A transport filter timeout is now identified in
4622 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4623 systems where "make" and "gmake" are different, calling "gmake" at top
4624 level broke things. I've arranged for the value of $(MAKE) to be passed
4625 from the Makefile to this script so that it can call the same version of
4629 A note about Exim versions 4.44 and 4.50
4630 ----------------------------------------
4632 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4633 changes of various kinds. As a consequence, a big documentation update was
4634 needed. This delayed the release for rather longer than seemed good, especially
4635 in the light of a couple of (minor) security issues. Therefore, the changes
4636 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4637 release. So 4.44 and 4.50 are in effect two different branches that both start
4640 I have left the 4.50 change log unchanged; it contains all the changes since
4641 4.43. The change log for 4.44 is below; many of its items are identical to
4642 those for 4.50. This seems to be the most sensible way to preserve the
4643 historical information.
4649 1. Minor wording change to the doc/README.SIEVE file.
4651 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4652 computation of the current number of files was incorrect.
4654 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4655 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4658 4. Give more explanation in the error message when the command for a transport
4659 filter fails to execute.
4661 5. There are several places where Exim runs a non-Exim command in a
4662 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4663 was being done only for the command run by the queryprogram router. It is
4664 now done for all such subprocesses. The other cases are: ${run, transport
4665 filters, and the commands run by the lmtp and pipe transports.
4667 6. Added CONFIGURE_GROUP build-time option.
4669 7. Some older OS have a limit of 256 on the maximum number of file
4670 descriptors. Exim was using setrlimit() to set 1000 as a large value
4671 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4672 systems. I've change it so that if it can't get 1000, it tries for 256.
4674 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4675 was an oversight, and furthermore, ever since the addition of extra
4676 controls (e.g. 4.43/32), the checks on when to allow different forms of
4677 "control" were broken. There should now be diagnostics for all cases when a
4678 control that does not make sense is encountered.
4680 9. Added the /retain_sender option to "control=submission".
4682 10. $recipients is now available in the predata ACL (oversight).
4684 11. Tidy the search cache before the fork to do a delivery from a message
4685 received from the command line. Otherwise the child will trigger a lookup
4686 failure and thereby defer the delivery if it tries to use (for example) a
4687 cached ldap connection that the parent has called unbind on.
4689 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4690 of $address_data from the recipient verification was clobbered by the
4691 sender verification.
4693 13. The value of address_data from a sender verification is now available in
4694 $sender_address_data in subsequent conditions in the ACL statement.
4696 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4698 15. Added a new option "connect=<time>" to callout options, to set a different
4701 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4702 was its contents. (It was OK if the option was not defined at all.)
4704 17. A "Completed" log line is now written for messages that are removed from
4705 the spool by the -Mrm option.
4707 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4708 information about exactly what failed.
4710 19. Added -dd to debug only the daemon process.
4712 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4713 handles timeouts, both on the server side and network timeouts. Renamed the
4714 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4716 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4717 It is now set to "smtps".
4719 22. $host_address is now set to the target address during the checking of
4720 ignore_target_hosts.
4722 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4723 being passed; this would have caused $sender_host_name to have been used if
4724 matching the list had actually called for a host name (not very likely,
4725 since this list is usually IP addresses). A host name is now passed as
4728 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4729 code that specifies a non-restarting handler (typically sigaction() in
4730 modern systems) in an attempt to fix a rare and obscure crash bug.
4732 25. Narrowed the window for a race in the daemon that could cause it to ignore
4733 SIGCHLD signals. This is not a major problem, because they are used only to
4734 wake it up if nothing else does.
4736 26. A malformed maildirsize file could cause Exim to calculate negative values
4737 for the mailbox size or file count. Odd effects could occur as a result.
4738 The maildirsize information is now recalculated if the size or filecount
4741 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4742 support for a long time. Removed HAVE_SYS_VFS_H.
4744 28. Installed the latest version of exipick from John Jetmore.
4746 29. In an address list, if the pattern was not a regular expression, an empty
4747 subject address (from a bounce message) matched only if the pattern was an
4748 empty string. Non-empty patterns were not even tested. This was the wrong
4749 because it is perfectly reasonable to use an empty address as part of a
4750 database query. An empty address is now tested by patterns that are
4751 lookups. However, all the other forms of pattern expect the subject to
4752 contain a local part and a domain, and therefore, for them, an empty
4753 address still always fails if the pattern is not itself empty.
4755 30. Exim went into a mad DNS loop when attempting to do a callout where the
4756 host was specified on an smtp transport, and looking it up yielded more
4757 than one IP address.
4759 31. Re-factored the code for checking spool and log partition space into a
4760 function that finds that data and another that does the check. The former
4761 is then used to implement four new variables: $spool_space, $log_space,
4762 $spool_inodes, and $log_inodes.
4764 32. The RFC2047 encoding function was originally intended for short strings
4765 such as real names; it was not keeping to the 75-character limit for
4766 encoded words that the RFC imposes. It now respects the limit, and
4767 generates multiple encoded words if necessary. To be on the safe side, I
4768 have increased the buffer size for the ${rfc2047: expansion operator from
4771 33. It is now permitted to omit both strings after an "if" condition; if the
4772 condition is true, the result is "true". As before, when the second string
4773 is omitted, a false condition yields an empty string. This makes it less
4774 cumbersome to write custom ACL and router conditions.
4776 34. Failure to deliver a bounce message always caused it to be frozen, even if
4777 there was an errors_to setting on the router. The errors_to setting is now
4780 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4781 canonical form (fully expanded) before being placed in
4782 $sender_host_address.
4784 36. The table in the code that translates DNS record types into text (T_A to
4785 "A" for instance) was missing entries for NS and CNAME. It is just possible
4786 that this could have caused confusion if both these types were looked up
4787 for the same domain, because the text type is used as part of Exim's
4788 per-process caching. But the chance of anyone hitting this buglet seems
4791 37. The dnsdb lookup has been extended in a number of ways.
4793 (1) There is a new type, "zns", which walks up the domain tree until it
4794 finds some nameserver records. It should be used with care.
4796 (2) There is a new type, "mxh", which is like "mx" except that it returns
4797 just the host names, not the priorities.
4799 (3) It is now possible to give a list of domains (or IP addresses) to be
4800 looked up. The behaviour when one of the lookups defers can be
4801 controlled by a keyword.
4803 (4) It is now possible to specify the separator character for use when
4804 multiple records are returned.
4806 38. The dnslists ACL condition has been extended: it is now possible to supply
4807 a list of IP addresses and/or domains to be looked up in a particular DNS
4810 39. Added log_selector=+queue_time_overall.
4812 40. When running the queue in the test harness, wait just a tad after forking a
4813 delivery process, to get repeatability of debugging output.
4815 41. Include certificate and key file names in error message when GnuTLS fails
4816 to set them up, because the GnuTLS error message doesn't include the name
4817 of the failing file when there is a problem reading it.
4819 42. Allow both -bf and -bF in the same test run.
4821 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4823 44. The "Exiscan patch" is now merged into the mainline Exim source.
4825 45. Sometimes the final signoff response after QUIT could fail to get
4826 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4827 < 0 before doing a fflush(). This bug looks as though it goes back to the
4828 introduction of TLS in release 3.20, but "sometimes" must have been rare
4829 because the tests only now provoked it.
4831 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4832 (this can affect the format of dates).
4834 47. exim_tidydb, when checking for the continued existence of a message for
4835 which it has found a message-specific retry record, was not finding
4836 messages that were in split spool directories. Consequently, it was
4837 deleting retry records that should have stayed in existence.
4839 48. Steve fixed some bugs in eximstats.
4841 49. The SPA authentication driver was not abandoning authentication and moving
4842 on to the next authenticator when an expansion was forced to fail,
4843 contradicting the general specification for all authenticators. Instead it
4844 was generating a temporary error. It now behaves as specified.
4846 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4847 (the order specifies the preference for clients). The order is now AES256,
4848 AES128, 3DES, ARCFOUR128.
4850 51. Small patch to Sieve code - explicitly set From: when generating an
4853 52. Exim crashed if a remote delivery caused a very long error message to be
4854 recorded - for instance if somebody sent an entire SpamAssassin report back
4855 as a large number of 550 error lines. This bug was coincidentally fixed by
4856 increasing the size of one of Exim's internal buffers (big_buffer) that
4857 happened as part of the Exiscan merge. However, to be on the safe side, I
4858 have made the code more robust (and fixed the comments that describe what
4861 53. Now that there can be additional text after "Completed" in log lines (if
4862 the queue_time_overall log selector is set), a one-byte patch to exigrep
4863 was needed to allow it to recognize "Completed" as not the last thing in
4866 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4867 patch that reportedly fixes this has been added. I am not expert enough to
4868 create a test for it. This is what the patch creator wrote:
4870 "I found a little strange behaviour of ldap code when working with
4871 Windows 2003 AD Domain, where users was placed in more than one
4872 Organization Units. When I tried to give exim partial DN, the exit code
4873 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4874 But simultaneously result of request was absolutely normal ldap result,
4875 so I produce this patch..."
4877 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4878 REFERENCE, so I have modified the code to exclude the patch when that macro
4881 55. Some experimental protocols are using DNS PTR records for new purposes. The
4882 keys for these records are domain names, not reversed IP addresses. The
4883 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4884 leaves it alone. Component reversal etc. now happens only for IP addresses.
4887 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4889 57. Double the size of the debug message buffer (to 2048) so that more of very
4890 long debug lines gets shown.
4892 58. The exicyclog utility now does better if the number of log files to keep
4893 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4895 59. Two changes related to the smtp_active_hostname option:
4897 (1) $smtp_active_hostname is now available as a variable.
4898 (2) The default for smtp_banner uses $smtp_active_hostname instead
4899 of $primary_hostname.
4901 60. The host_aton() function is supposed to be passed a string that is known
4902 to be a valid IP address. However, in the case of IPv6 addresses, it was
4903 not checking this. This is a hostage to fortune. Exim now panics and dies
4904 if the condition is not met. A case was found where this could be provoked
4905 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4906 components; fortuitously, this particular loophole had already been fixed
4907 by change 4.50/55 above.
4909 If there are any other similar loopholes, the new check in host_aton()
4910 itself should stop them being exploited. The report I received stated that
4911 data on the command line could provoke the exploit when Exim was running as
4912 exim, but did not say which command line option was involved. All I could
4913 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4914 running as the user.
4917 61. There was a buffer overflow vulnerability in the SPA authentication code
4918 (which came originally from the Samba project). I have added a test to the
4919 spa_base64_to_bits() function which I hope fixes it.
4922 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4923 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4925 63. The daemon start-up calls getloadavg() while still root for those OS that
4926 need the first call to be done as root, but it missed one case: when
4927 deliver_queue_load_max is set with deliver_drop_privilege. This is
4928 necessary for the benefit of the queue runner, because there is no re-exec
4929 when deliver_drop_privilege is set.
4931 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4932 This has been fixed.
4934 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4935 was in use, was not putting the data itself into the right store pool;
4936 consequently, it could be overwritten for a subsequent message in the same
4937 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4940 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4942 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4943 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4944 the function were treating the return as a boolean value, which happened to
4945 work because 0=false and not-0=true, but is not correct code.
4947 68. The host_aton() function was not handling scoped IPv6 addresses (those
4948 with, for example, "%eth0" on the end) correctly.
4950 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4951 negated items (that is, ~something) in unsigned ints. Some compilers
4952 apparently mutter when there is no cast.
4954 70. If an address verification called from an ACL failed, and did not produce a
4955 user-specific message (i.e. there was only a "system" message), nothing was
4956 put in $acl_verify_message. In this situation, it now puts the system
4959 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4960 session; change 4.31/43 added the unwanted input to the log line - except
4961 that it did not do this in the start of session case. It now does.
4963 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4964 This is wrong and can cause the other end to generate a synchronization
4965 error if it is another Exim or anything else that does the synchronization
4966 check. A QUIT command is no longer sent after a timeout.
4968 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4969 during host lookups.
4971 74. The defer_ok option of callout verification was not working if it was used
4972 when verifying addresses in header lines, that is, for this case:
4974 verify = header_sender/callout=defer_ok
4976 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4977 those file descriptors could be used for SMTP connections. If anything
4978 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4979 could be sent to the SMTP client, causing chaos. The daemon now opens
4980 stdin, stdout, and stderr to /dev/null when it puts itself into the
4983 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4984 log by default. The user can override this with suitable Perl magic.
4986 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4987 the log message when discard triggers, was not working for the DATA ACL or
4988 for the non-SMTP ACL.
4990 78. Error message wording change in sieve.c.
4992 79. If smtp_accept_max_per_host was set, the number of connections could be
4993 restricted to fewer than expected, because the daemon was trying to set up
4994 a new connection before checking whether the processes handling previous
4995 connections had finished. The check for completed processes is now done
4996 earlier. On busy systems, this bug wouldn't be noticed because something
4997 else would have woken the daemon, and it would have reaped the completed
5000 80. If a message was submitted locally by a user whose login name contained one
5001 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5002 It caused a spool format error. I have fixed the spool reading code. A
5003 related problem was that the "from" clause in the Received: line became
5004 illegal because of the space(s). It is now covered by ${quote_local_part.
5006 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5009 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5011 83. Patch from Sieve maintainer to fix unterminated string problem in
5012 "vacation" handling.
5014 84. Some minor changes to the Linux configuration files to help with other
5015 OS variants using glibc.
5017 85. One more patch for Sieve to update vacation handling to latest spec.
5020 ----------------------------------------------------
5021 See the note above about the 4.44 and 4.50 releases.
5022 ----------------------------------------------------
5028 1. Change 4.43/35 introduced a bug that caused file counts to be
5029 incorrectly computed when quota_filecount was set in an appendfile
5032 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5033 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5036 3. Give more explanation in the error message when the command for a transport
5037 filter fails to execute.
5039 4. There are several places where Exim runs a non-Exim command in a
5040 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5041 was being done only for the command run by the queryprogram router. It is
5042 now done for all such subprocesses. The other cases are: ${run, transport
5043 filters, and the commands run by the lmtp and pipe transports.
5045 5. Some older OS have a limit of 256 on the maximum number of file
5046 descriptors. Exim was using setrlimit() to set 1000 as a large value
5047 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5048 systems. I've change it so that if it can't get 1000, it tries for 256.
5050 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5051 was an oversight, and furthermore, ever since the addition of extra
5052 controls (e.g. 4.43/32), the checks on when to allow different forms of
5053 "control" were broken. There should now be diagnostics for all cases when a
5054 control that does not make sense is encountered.
5056 7. $recipients is now available in the predata ACL (oversight).
5058 8. Tidy the search cache before the fork to do a delivery from a message
5059 received from the command line. Otherwise the child will trigger a lookup
5060 failure and thereby defer the delivery if it tries to use (for example) a
5061 cached ldap connection that the parent has called unbind on.
5063 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5064 of $address_data from the recipient verification was clobbered by the
5065 sender verification.
5067 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5068 was its contents. (It was OK if the option was not defined at all.)
5070 11. A "Completed" log line is now written for messages that are removed from
5071 the spool by the -Mrm option.
5073 12. $host_address is now set to the target address during the checking of
5074 ignore_target_hosts.
5076 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5077 being passed; this would have caused $sender_host_name to have been used if
5078 matching the list had actually called for a host name (not very likely,
5079 since this list is usually IP addresses). A host name is now passed as
5082 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5083 code that specifies a non-restarting handler (typically sigaction() in
5084 modern systems) in an attempt to fix a rare and obscure crash bug.
5086 15. Narrowed the window for a race in the daemon that could cause it to ignore
5087 SIGCHLD signals. This is not a major problem, because they are used only to
5088 wake it up if nothing else does.
5090 16. A malformed maildirsize file could cause Exim to calculate negative values
5091 for the mailbox size or file count. Odd effects could occur as a result.
5092 The maildirsize information is now recalculated if the size or filecount
5095 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5096 support for a long time. Removed HAVE_SYS_VFS_H.
5098 18. Updated exipick to current release from John Jetmore.
5100 19. Allow an empty sender to be matched against a lookup in an address list.
5101 Previously the only cases considered were a regular expression, or an
5104 20. Exim went into a mad DNS lookup loop when doing a callout where the
5105 host was specified on the transport, if the DNS lookup yielded more than
5108 21. The RFC2047 encoding function was originally intended for short strings
5109 such as real names; it was not keeping to the 75-character limit for
5110 encoded words that the RFC imposes. It now respects the limit, and
5111 generates multiple encoded words if necessary. To be on the safe side, I
5112 have increased the buffer size for the ${rfc2047: expansion operator from
5115 22. Failure to deliver a bounce message always caused it to be frozen, even if
5116 there was an errors_to setting on the router. The errors_to setting is now
5119 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5120 canonical form (fully expanded) before being placed in
5121 $sender_host_address.
5123 24. Updated eximstats to version 1.33
5125 25. Include certificate and key file names in error message when GnuTLS fails
5126 to set them up, because the GnuTLS error message doesn't include the name
5127 of the failing file when there is a problem reading it.
5129 26. Expand error message when OpenSSL has problems setting up cert/key files.
5132 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5133 (this can affect the format of dates).
5135 28. exim_tidydb, when checking for the continued existence of a message for
5136 which it has found a message-specific retry record, was not finding
5137 messages that were in split spool directories. Consequently, it was
5138 deleting retry records that should have stayed in existence.
5140 29. eximstats updated to version 1.35
5141 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5142 1.35 - bugfix such that pie charts by volume are generated correctly
5144 30. The SPA authentication driver was not abandoning authentication and moving
5145 on to the next authenticator when an expansion was forced to fail,
5146 contradicting the general specification for all authenticators. Instead it
5147 was generating a temporary error. It now behaves as specified.
5149 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5150 (the order specifies the preference for clients). The order is now AES256,
5151 AES128, 3DES, ARCFOUR128.
5153 31. Small patch to Sieve code - explicitly set From: when generating an
5156 32. Exim crashed if a remote delivery caused a very long error message to be
5157 recorded - for instance if somebody sent an entire SpamAssassin report back
5158 as a large number of 550 error lines. This bug was coincidentally fixed by
5159 increasing the size of one of Exim's internal buffers (big_buffer) that
5160 happened as part of the Exiscan merge. However, to be on the safe side, I
5161 have made the code more robust (and fixed the comments that describe what
5164 33. Some experimental protocols are using DNS PTR records for new purposes. The
5165 keys for these records are domain names, not reversed IP addresses. The
5166 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5167 leaves it alone. Component reversal etc. now happens only for IP addresses.
5170 34. The host_aton() function is supposed to be passed a string that is known
5171 to be a valid IP address. However, in the case of IPv6 addresses, it was
5172 not checking this. This is a hostage to fortune. Exim now panics and dies
5173 if the condition is not met. A case was found where this could be provoked
5174 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5175 components; fortuitously, this particular loophole had already been fixed
5176 by change 4.50/55 or 4.44/33 above.
5178 If there are any other similar loopholes, the new check in host_aton()
5179 itself should stop them being exploited. The report I received stated that
5180 data on the command line could provoke the exploit when Exim was running as
5181 exim, but did not say which command line option was involved. All I could
5182 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5183 running as the user.
5186 35. There was a buffer overflow vulnerability in the SPA authentication code
5187 (which came originally from the Samba project). I have added a test to the
5188 spa_base64_to_bits() function which I hope fixes it.
5191 36. The daemon start-up calls getloadavg() while still root for those OS that
5192 need the first call to be done as root, but it missed one case: when
5193 deliver_queue_load_max is set with deliver_drop_privilege. This is
5194 necessary for the benefit of the queue runner, because there is no re-exec
5195 when deliver_drop_privilege is set.
5197 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5198 was in use, was not putting the data itself into the right store pool;
5199 consequently, it could be overwritten for a subsequent message in the same
5200 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5203 38. Sometimes the final signoff response after QUIT could fail to get
5204 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5205 < 0 before doing a fflush(). This bug looks as though it goes back to the
5206 introduction of TLS in release 3.20, but "sometimes" must have been rare
5207 because the tests only now provoked it.
5213 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5214 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5215 Now it returns an integer. A number of calls were still expecting a T/F
5216 return. Fortuitously, in all cases, the tests worked in OK situations,
5217 which is the norm. However, things would have gone wrong on any write
5218 failures on the smtp file descriptor. This function is used when sending
5219 messages over SMTP and also when doing verify callouts.
5221 2. When Exim is called to do synchronous delivery of a locally submitted
5222 message (the -odf or -odi options), it no longer closes stderr before doing
5225 3. Implemented the mua_wrapper option.
5227 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5229 5. Implemented the functions header_remove(), header_testname(),
5230 header_add_at_position(), and receive_remove_recipient(), and exported them
5233 6. If an ACL "warn" statement specified the addition of headers, Exim already
5234 inserted X-ACL-Warn: at the start if there was no header name. However, it
5235 was not making this test for the second and subsequent header lines if
5236 there were newlines in the string. This meant that an invalid header could
5237 be inserted if Exim was badly configured.
5239 7. Allow an ACL "warn" statement to add header lines at the start or after all
5240 the Received: headers, as well as at the end.
5242 8. Added the rcpt_4xx retry error code.
5244 9. Added postmaster_mailfrom=xxx to callout verification option.
5246 10. Added mailfrom=xxxx to the callout verification option, for verify=
5249 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5250 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5252 12. Inserted some casts to stop certain compilers warning when using pointer
5253 differences as field lengths or precisions in printf-type calls (mostly
5254 affecting debugging statements).
5256 13. Added optional readline() support for -be (dynamically loaded).
5258 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5259 same clock tick as a message's arrival, so that its received time was the
5260 same as the "first fail" time on the retry record, and that message
5261 remained on the queue past the ultimate address timeout, every queue runner
5262 would try a delivery (because it was past the ultimate address timeout) but
5263 after another failure, the ultimate address timeout, which should have then
5264 bounced the address, did not kick in. This was a "< instead of <=" error;
5265 in most cases the first failure would have been in the next clock tick
5266 after the received time, and all would be well.
5268 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5269 being recognized when the domain list was tested by the match_domain
5270 condition in an expansion string.
5272 16. Added the ${str2b64: operator.
5274 17. Exim was always calling setrlimit() to set a large limit for the number of
5275 processes, without checking whether the existing limit was already
5276 adequate. (It did check for the limit on file descriptors.) Furthermore,
5277 errors from getrlimit() and setrlimit() were being ignored. Now they are
5278 logged to the main and panic logs, but Exim does carry on, to try to do its
5279 job under whatever limits there are.
5281 18. Imported PCRE 5.0.
5283 19. Trivial typo in log message " temporarily refused connection" (the leading
5286 20. If the log selector return_path_on_delivery was set and an address was
5287 redirected to /dev/null, the delivery process crashed because it assumed
5288 that a return path would always be set for a "successful" delivery. In this
5289 case, the whole delivery is bypassed as an optimization, and therefore no
5292 21. Internal re-arrangement: the function for sending a challenge and reading
5293 a response while authentication was assuming a zero-terminated challenge
5294 string. It's now changed to take a pointer and a length, to allow for
5295 binary data in such strings.
5297 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5299 23. Exim was not respecting finduser_retries when seeking the login of the
5300 uid under which it was called; it was always trying 10 times. (The default
5301 setting of finduser_retries is zero.) Also, it was sleeping after the final
5302 failure, which is pointless.
5304 24. Implemented tls_on_connect_ports.
5306 25. Implemented acl_smtp_predata.
5308 26. If the domain in control=submission is set empty, Exim assumes that the
5309 authenticated id is a complete email address when it generates From: or
5310 Sender: header lines.
5312 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5313 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5314 chown and chgrp in /bin and hostname in /usr/bin.
5316 28. Exim was keeping the "process log" file open after each use, just as it
5317 does for the main log. This opens the possibility of it remaining open for
5318 long periods when the USR1 signal hits a daemon. Occasional processlog
5319 errors were reported, that could have been caused by this. Anyway, it seems
5320 much more sensible not to leave this file open at all, so that is what now
5323 29. The long-running daemon process does not normally write to the log once it
5324 has entered its main loop, and it closes the log before doing so. This is
5325 so that log files can straightforwardly be renamed and moved. However,
5326 there are a couple of unusual error situations where the daemon does write
5327 log entries, and I had neglected to close the log afterwards.
5329 30. The text of an SMTP error response that was received during a remote
5330 delivery was being truncated at 512 bytes. This is too short for some of
5331 the long messages that one sometimes sees. I've increased the limit to
5334 31. It is now possible to make retry rules that apply only when a message has a
5335 specific sender, in particular, an empty sender.
5337 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5338 it possible to be selective about when SMTP synchronization is enforced.
5340 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5342 32. Implemented hosts_connection_nolog.
5344 33. Added an ACL for QUIT.
5346 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5349 35. Added mailbox_size and mailbox_filecount to appendfile.
5351 36. Added control = no_multiline_responses to ACLs.
5353 37. There was a bug in the logic of the code that waits for the clock to tick
5354 in the case where the clock went backwards by a substantial amount such
5355 that the microsecond fraction of "now" was more than the microsecond
5356 fraction of "then" (but the whole seconds number was less).
5358 38. Added support for the libradius Radius client library this is found on
5359 FreeBSD (previously only the radiusclient library was supported).
5365 1. When certain lookups returned multiple values in the form name=value, the
5366 quoting of the values was not always being done properly. Specifically:
5367 (a) If the value started with a double quote, but contained no whitespace,
5369 (b) If the value contained whitespace other than a space character (i.e.
5370 tabs or newlines or carriage returns) it was not quoted.
5371 This fix has been applied to the mysql and pgsql lookups by writing a
5372 separate quoting function and calling it from the lookup code. The fix
5373 should probably also be applied to nisplus, ibase and oracle lookups, but
5374 since I cannot test any of those, I have not disturbed their existing code.
5376 2. A hit in the callout cache for a specific address caused a log line with no
5377 reason for rejecting RCPT. Now it says "Previous (cached) callout
5378 verification failure".
5380 3. There was an off-by-one bug in the queryprogram router. An over-long
5381 return line was truncated at 256 instead of 255 characters, thereby
5382 overflowing its buffer with the terminating zero. As well as fixing this, I
5383 have increased the buffer size to 1024 (and made a note to document this).
5385 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5386 when Exim is waiting for an SMTP response from a remote server, Exim
5387 restarts its select() call on the socket, thereby resetting its timeout.
5388 This is not a problem when such interrupts are rare. Somebody set up a cron
5389 job to run exiwhat every 2 minutes, which is less than the normal select()
5390 timeout (5 or 10 minutes). This meant that the select() timeout never
5391 kicked in because it was always reset. I have fixed this by comparing the
5392 time when an interrupt arrives with the time at the start of the first call
5393 to select(). If more time than the timeout has elapsed, the interrupt is
5394 treated as a timeout.
5396 5. Some internal re-factoring in preparation for the addition of Sieve
5397 extensions (by MH). In particular, the "personal" test is moved to a
5398 separate function, and given an option for scanning Cc: and Bcc: (which is
5399 not set for Exim filters).
5401 6. When Exim created an email address using the login of the caller as the
5402 local part (e.g. when creating a From: or Sender: header line), it was not
5403 quoting the local part when it contained special characters such as @.
5405 7. Installed new OpenBSD configuration files.
5407 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5408 try to make them clearer.
5410 9. Callout options, other than the timeout value, were being ignored when
5411 verifying sender addresses in header lines. For example, when using
5413 verify = header_sender/callout=no_cache
5415 the cache was (incorrectly) being used.
5417 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5418 only the Cygwin environment.
5420 11. When return_path_on_delivery was set as a log selector, if different remote
5421 addresses in the same message used different return paths and parallel
5422 remote delivery occurred, the wrong values would sometimes be logged.
5423 (Whenever a remote delivery process finished, the return path value from
5424 the most recently started remote delivery process was logged.)
5426 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5427 lines when AUTH and/or TLS are in use. This is the "received protocol"
5428 field. Exim used to use "asmtp" for authenticated SMTP, without any
5429 indication (in the protocol name) for TLS use. Now it follows the RFC and
5430 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5431 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5432 names appear in log lines as well as in Received: header lines.
5434 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5435 extensions, and comparison tests, and to fix some bugs.
5437 14. Changes to the "personal" filter test:
5439 (1) The test was buggy in that it was just doing the equivalent of
5440 "contains" tests on header lines. For example, if a user's address was
5441 anne@some.where, the "personal" test would incorrectly be true for
5443 To: susanne@some.where
5445 This test is now done by extracting each address from the header in turn,
5446 and checking the entire address. Other tests that are part of "personal"
5447 are now done using regular expressions (for example, to check local parts
5448 of addresses in From: header lines).
5450 (2) The list of non-personal local parts in From: addresses has been
5451 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5452 taken from the Sieve specification recommendations.
5454 (3) If the message contains any header line starting with "List-" it is
5455 treated as non-personal.
5457 (4) The test for "circular" in the Subject: header line has been removed
5458 because it now seems ill-conceived.
5460 15. Minor typos in src/EDITME comments corrected.
5462 16. Installed latest exipick from John Jetmore.
5464 17. If headers_add on a router specified a text string that was too long for
5465 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5466 of string_sprintf() is now avoided.
5468 18. $message_body_size was not set (it was always zero) when running the DATA
5469 ACL and the local_scan() function.
5471 19. For the "mail" command in an Exim filter, no default was being set for
5472 the once_repeat time, causing a random time value to be used if "once" was
5473 specified. (If the value happened to be <= 0, no repeat happened.) The
5474 default is now 0s, meaning "never repeat". The "vacation" command was OK
5475 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5476 (I found it when inspecting the code).
5478 20. There is now an overall timeout for performing a callout verification. It
5479 defaults to 4 times the callout timeout, which applies to individual SMTP
5480 commands during the callout. The overall timeout applies when there is more
5481 than one host that can be tried. The timeout is checked before trying the
5482 next host. This prevents very long delays if there are a large number of
5483 hosts and all are timing out (e.g. when the network connections are timing
5484 out). The value of the overall timeout can be changed by specifying an
5485 additional sub-option for "callout", called "maxwait". For example:
5487 verify = sender/callout=5s,maxwait=20s
5489 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5490 to the end before writing, but this should make it even safer).
5492 22. Exim was forgetting that it had advertised PIPELINING for the second and
5493 subsequent messages on an SMTP connection. It was also not resetting its
5494 memory on STARTTLS and an internal HELO.
5496 23. When Exim logs an SMTP synchronization error within a session, it now
5497 records whether PIPELINING has been advertised or not.
5499 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5500 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5501 rather than long int.
5503 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5505 26. Added the never_mail option to autoreply.
5511 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5512 crash if the getsockname() call failed; this can happen if a connection is
5513 closed very soon after it is established. The problem was simply in the
5514 order in which certain operations were done, causing Exim to try to write
5515 to the SMTP stream before it had set up the file descriptor. The bug has
5516 been fixed by making things happen in the correct order.
5522 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5523 before the connection was closed, thus losing the rejection response.
5525 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5526 some early Solaris releases, but causes trouble in current releases where
5527 socklen_t is defined.
5529 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5532 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5535 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5536 that releasing the top of it at the end releases what was used for sub-
5537 expansions (unless the block got too big). However, discard this block if
5538 the first thing is a variable or header, so that we can use its block when
5539 it is dynamic (useful for very large $message_headers, for example).
5541 6. Lookups now cache *every* query, not just the most recent. A new, separate
5542 store pool is used for this. It can be recovered when all lookup caches are
5543 flushed. Lookups now release memory at the end of their result strings.
5544 This has involved some general refactoring of the lookup sources.
5546 7. Some code has been added to the store_xxx() functions to reduce the amount
5547 of flapping under certain conditions.
5549 8. log_incoming_interface used to affect only the <= reception log lines. Now
5550 it causes the local interface and port to be added to several more SMTP log
5551 lines, for example "SMTP connection from", and rejection lines.
5553 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5555 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5557 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5558 could be overwritten at the end of the current message (or the start of a
5559 new message if it was set in a HELO ACL). The value is now preserved for
5560 the duration of the SMTP connection.
5562 12. If a transport had a headers_rewrite setting, and a matching header line
5563 contained an unqualified address, that address was qualified, even if it
5564 did not match any rewriting rules. The underlying bug was that the values
5565 of the flags that permit the existence of unqualified sender and recipient
5566 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5567 non-local messages, and by -bnq for local messages) were not being
5568 preserved with the message after it was received.
5570 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5571 "next input=" as part of the text comprising the host identity instead of
5572 the correct text. The code was using the same buffer for two different
5573 strings. However, depending on which order the printing function evaluated
5574 its arguments, the bug did not always show up. Under Linux, for example, my
5575 test suite worked just fine.
5577 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5578 doesn't work with some older versions of Perl. It has been changed to "my",
5579 which in any case is probably the better facility to use.
5581 15. A really picky compiler found some instances of statements for creating
5582 error messages that either had too many or two few arguments for the format
5585 16. The size of the buffer for calls to the DNS resolver has been increased
5586 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5587 for addresses that have a lot of PTR records. This alleviates a problem; it
5588 does not fully solve it.
5590 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5591 buffer now truncates the list and logs the incident, which is the same
5592 action as happens when Exim is looking up a host name and its aliases.
5593 Previously in this situation something unpredictable would happen;
5594 sometimes it was "internal error: store_reset failed".
5596 18. If a server dropped the connection unexpectedly when an Exim client was
5597 using GnuTLS and trying to read a response, the client delivery process
5598 crashed while trying to generate an error log message.
5600 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5601 string, for example:
5603 warn message = H1: something\nH2: something
5605 the text was added as a single header line from Exim's point of view
5606 though it ended up OK in the delivered message. However, searching for the
5607 second and subsequent header lines using $h_h2: did not work. This has been
5608 fixed. Similarly, if a system filter added multiple headers in this way,
5609 the routers could not see them.
5611 20. Expanded the error message when iplsearch is called with an invalid key to
5612 suggest using net-iplsearch in a host list.
5614 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5615 ACLs are no longer actually imposed (and a message to that effect is
5618 22. If a "gecos" field in a passwd entry contained escaped characters, in
5619 particular, if it contained a \" sequence, Exim got it wrong when building
5620 a From: or a Sender: header from that name. A second bug also caused
5621 incorrect handling when an unquoted " was present following a character
5622 that needed quoting.
5624 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5625 was not being matched caselessly.
5627 24. Arranged for all hyphens in the exim.8 source to be escaped with
5630 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5631 properly. Recipient callout cache records were still being keyed to include
5632 the sender, even when use_sender was set false. This led to far more
5633 callouts that were necessary. The sender is no longer included in the key
5634 when use_sender is false.
5636 26. Added "control = submission" modifier to ACLs.
5638 27. Added the ${base62d: operator to decode base 62 numbers.
5640 28. dnsdb lookups can now access SRV records.
5642 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5643 the configuration file.
5645 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5646 (-v) mode. This makes the output for a verbose queue run more intelligible.
5648 31. Added a use_postmaster feature to recipient callouts.
5650 32. Added the $body_zerocount variable, containing the number of binary zero
5651 bytes in the message body.
5653 33. The time of last modification of the "new" subdirectory is now used as the
5654 "mailbox time last read" when there is a quota error for a maildir
5657 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5659 35. Added +ignore_unknown as a special item in host lists.
5661 36. Code for decoding IPv6 addresses in host lists is now included, even if
5662 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5663 address was recognized as an IP address, but was then not correctly decoded
5664 into binary, causing unexpected and incorrect effects when compared with
5671 1. Very minor rewording of debugging text in manualroute to say "list of
5672 hosts" instead of "hostlist".
5674 2. If verify=header_syntax was set, and a header line with an unqualified
5675 address (no domain) and a large number of spaces between the end of the
5676 name and the colon was received, the reception process suffered a buffer
5677 overflow, and (when I tested it) crashed. This was caused by some obsolete
5678 code that should have been removed. The fix is to remove it!
5680 3. When running in the test harness, delay a bit after writing a bounce
5681 message to get a bit more predictability in the log output.
5683 4. Added a call to search_tidyup() just before forking a reception process. In
5684 theory, someone could use a lookup in the expansion of smtp_accept_max_
5685 per_host which, without the tidyup, could leave open a database connection.
5687 5. Added the variables $recipient_data and $sender_data which get set from a
5688 lookup success in an ACL "recipients" or "senders" condition, or a router
5689 "senders" option, similar to $domain_data and $local_part_data.
5691 6. Moved the writing of debug_print from before to after the "senders" test
5694 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5695 problems for message scanning, either using a data ACL, or using
5696 local_scan() because the Received: header was not generated till after they
5697 were called (in order to set the time as the time of reception completion).
5698 I have revised the way this works. The header is now generated after the
5699 body is received, but before the ACL or local_scan() are called. After they
5700 are run, the timestamp in the header is updated.
5706 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5707 before starting a queue runner without re-exec. This happened only when
5708 deliver_drop_privilege was set or when the Exim user was set to root. The
5709 effect of the bug was that timeouts during subsequent deliveries caused
5710 crashes instead of being properly handled. The handler is now left at its
5711 default (and expected) setting.
5713 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5714 message, again when deliver_drop_privilege is set or Exim is run as root.
5715 The bug described in (1) was not present in this case, but the tidying up
5716 of the other signals was missing. I have made the two cases consistent.
5718 3. The ignore_target_hosts setting on a manualroute router was being ignored
5719 for hosts that were looked up using the /MX notation.
5721 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5724 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5725 operated on the sender address. After changing the $sender_address to <>
5726 for the sender address verify, Exim was re-instated it as the original
5727 (before rewriting) address, but remembering that it had rewritten it, so it
5728 wasn't rewriting it again. This bug also had the effect of breaking the
5729 sender address verification caching when the sender address was rewritten.
5731 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5732 This has been changed so that if the ip literal address matches
5733 ignore_target_hosts, the router declines.
5735 7. Added expansion conditions match_domain, match_address, and match_local_
5736 part (NOT match_host).
5738 8. The placeholder for the Received: header didn't have a length field set.
5740 9. Added code to Exim itself and to exim_lock to test for a specific race
5741 condition that could lead to file corruption when using MBX delivery. The
5742 issue is with the lockfile that is created in /tmp. If this file is removed
5743 after a process has opened it but before that process has acquired a lock,
5744 there is the potential for a second process to recreate the file and also
5745 acquire a lock. This could lead to two Exim processes writing to the file
5746 at the same time. The added code performs the same test as UW imapd; it
5747 checks after acquiring the lock that its file descriptor still refers to
5748 the same named file.
5750 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5751 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5754 11. Added the smtp_active_hostname option. If used, this will typically be made
5755 to depend on the incoming interface address. Because $interface_address is
5756 not set up until the daemon has forked a reception process, error responses
5757 that can happen earlier (such as "too many connections") no longer contain
5760 12. If an expansion in a condition on a "warn" statement fails because a lookup
5761 defers, the "warn" statement is abandoned, and the next ACL statement is
5762 processed. Previously this caused the whole ACL to be aborted.
5764 13. Added the iplsearch lookup type.
5766 14. Added ident_timeout as a log selector.
5768 15. Added tls_certificate_verified as a log selector.
5770 16. Added a global option tls_require_ciphers (compare the smtp transport
5771 option of the same name). This controls incoming TLS connections.
5773 17. I finally figured out how to make tls_require_ciphers do a similar thing
5774 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5775 before starting the TLS session.
5777 18. Tabs are now shown as \t in -bP output.
5779 19. If the log selector return_path_on_delivery was set, Exim crashed when
5780 bouncing a message because it had too many Received: header lines.
5782 20. If two routers both had headers_remove settings, and the first one included
5783 a superfluous trailing colon, the final name in the first list and the
5784 first name in the second list were incorrectly joined into one item (with a
5785 colon in the middle).
5791 1. Added -C and -D options to the exinext utility, mainly to make it easier
5792 to include in the automated testing, but these could be helpful when
5793 multiple configurations are in use.
5795 2. The exinext utility was not formatting the output nicely when there was
5796 an alternate port involved in the retry record key, nor when there was a
5797 message id as well (for retries that were specific to a specific message
5798 and a specific host). It was also confused by IPv6 addresses, because of
5799 the additional colons they contain. I have fixed the IPv4 problem, and
5800 patched it up to do a reasonable job for IPv6.
5802 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5803 delivery, the log line now contains "pipelined" if PIPELINING was used.
5805 4. An SMTP transport process used to panic and die if the bind() call to set
5806 an explicit outgoing interface failed. This has been changed; it is now
5807 treated in the same way as a connect() failure.
5809 5. A reference to $sender_host_name in the part of a conditional expansion
5810 that was being skipped was still causing a DNS lookup. This no longer
5813 6. The def: expansion condition was not recognizing references to header lines
5814 that used bh_ and bheader_.
5816 7. Added the _cache feature to named lists.
5818 8. The code for checking quota_filecount in the appendfile transport was
5819 allowing one more file than it should have been.
5821 9. For compatibility with Sendmail, the command line option
5829 and sets the incoming protocol and host name (for trusted callers). The
5830 host name and its colon can be omitted when only the protocol is to be set.
5831 Note the Exim already has two private options, -pd and -ps, that refer to
5832 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5833 "s", but I don't think that's a major issue.
5835 10. A number of refactoring changes to the code, none of which should affect
5838 (a) The number of logging options was getting close to filling up the
5839 32-bit word that was used as a bit map. I have split them into two classes:
5840 those that are passed in the argument to log_write(), and those that are
5841 only ever tested independently outside of that function. These are now in
5842 separate 32-bit words, so there is plenty of room for expansion again.
5843 There is no change in the user interface or the logging behaviour.
5845 (b) When building, for example, log lines, the code previously used a
5846 macro that called string_cat() twice, in order to add two strings. This is
5847 not really sufficiently general. Furthermore, there was one instance where
5848 it was actually wrong because one of the argument was used twice, and in
5849 one call a function was used. (As it happened, calling the function twice
5850 did not affect the overall behaviour.) The macro has been replaced by a
5851 function that can join an arbitrary number of extra strings onto a growing
5854 (c) The code for expansion conditions now uses a table and a binary chop
5855 instead of a serial search (which was left over from when there were very
5856 few conditions). Also, it now recognizes conditions like "pam" even when
5857 the relevant support is not compiled in: a suitably worded error message is
5858 given if an attempt is made to use such a condition.
5860 11. Added ${time_interval:xxxxx}.
5862 12. A bug was causing one of the ddress fields not to be passed back correctly
5863 from remote delivery subprocesses. The field in question was not being
5864 subsequently used, so this caused to problems in practice.
5866 13. Added new log selectors queue_time and deliver_time.
5868 14. Might have fixed a bug in maildirsizefile handling that threw up
5869 "unexpected character" debug warnings, and recalculated the data
5870 unnecessarily. In any case, I expanded the warning message to give more
5873 15. Added the message "Restricted characters in address" to the statements in
5874 the default ACL that block characters like @ and % in local parts.
5876 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5877 Three changes have been made:
5879 (a) There was a serious bug; a negative response to MAIL caused the whole
5880 recipient domain to be cached as invalid, thereby blocking all messages
5881 to all local parts at the same domain, from all senders. This bug has
5882 been fixed. The domain is no longer cached after a negative response to
5883 MAIL if the sender used is not empty.
5885 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5888 (c) A new callout option, "use_sender" has been added for people who want
5889 the modified behaviour.
5895 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5898 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5899 indeed breaks things for older releases.
5901 3. Added additional logging to the case where there is a problem reading data
5902 from a filter that is running in a subprocess using a pipe, in order to
5903 try to track down a specific problem.
5905 4. Testing facility fudge: when running in the test harness and attempting
5906 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5907 getting "No route to host". Convert this to a timeout.
5909 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5912 6. Some OS don't have socklen_t but use size_t instead. This affects the
5913 fifth argument of getsockopt() amongst other things. This is now
5914 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5915 can be set for individual OS. I have set it for SunOS5, OSF1, and
5916 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5917 some earlier ones do not.
5919 7. Change 4.30/15 was not doing the test caselessly.
5921 8. The standard form for an IPv6 address literal was being rejected by address
5922 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5923 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5924 this, as well as the form without the "IPv6" on the front (but only when
5925 address literals are enabled, of course).
5927 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5929 10. Exim crashed if a message with an empty sender address specified by -f
5930 encountered a router with an errors_to setting. This could be provoked only
5931 by a command such as
5935 where an empty string was supplied; "<>" did not hit this bug.
5937 11. Installed PCRE release 4.5.
5939 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5940 remained set. It is now erased.
5942 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5943 times from message ids (which are base 36 rather than the normal 62).
5945 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5946 were being counted as actual protocol errors, and logged if the log
5947 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5948 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5949 and DATA following a set of rejected RCPTs do not count as protocol errors.
5950 In other words, Exim assumes they were pipelined, though this may not
5951 actually be the case. Of course, in all cases the client gets an
5952 appropriate error code.
5954 15. If a lookup fails in an ACL condition, a message about the failure may
5955 be available; it is used if testing the ACL cannot continue, because most
5956 such messages specify what the cause of the deferral is. However, some
5957 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5958 that caused an old message to be retained and used if a later statement
5959 caused a defer, replacing the real cause of the deferral.
5961 16. If an IP address had so many PTR records that the DNS lookup buffer
5962 was not large enough to hold them, Exim could crash while trying to process
5963 the truncated data. It now detects and logs this case.
5965 17. Further to 4.21/58, another change has been made: if (and only if) the
5966 first line of a message (the first header line) ends with CRLF, a bare LF
5967 in a subsequent header line has a space inserted after it, so as not to
5968 terminate the header.
5970 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5971 unnecessarily, used atoi() instead of strtol(), and didn't check the
5972 termination when getting file sizes from file names by regex.
5974 19. Completely re-implemented the support for maildirsize files, in the light
5975 of a number of problems with the previous contributed implementation
5976 (4.30/29). In particular:
5978 . If the quota is zero, the maildirsize file is maintained, but no quota is
5981 . If the maildir directory does not exist, it is created before any attempt
5982 to write a maildirsize file.
5984 . The quota value in the file is just a cache; if the quota is changed in
5985 the transport, the new value overrides.
5987 . A regular expression is available for excluding directories from the
5990 20. The autoreply transport checks the characters in options that define the
5991 message's headers; it allows continued headers, but it was checking with
5992 isspace() after an embedded newline instead of explicitly looking for a
5995 21. If all the "regular" hosts to which an address was routed had passed their
5996 expiry times, and had not reached their retry times, the address was
5997 bounced, even if fallback hosts were defined. Now Exim should go on to try
6000 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6001 equivalent code in the SMTP transport. Some hosts send humungous responses
6002 to HELO/EHLO, more than 1024 it seems.
6004 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6005 gives compiler warnings in some environments. I've now done it "properly",
6008 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6009 (because of problems with the built-in one) was declared to return uschar *
6010 instead of char *, causing compiler failure.
6012 25. Fixed a file descriptor leak when processing alias/forward files.
6014 26. Fixed a minor format string issue in dbfn.c.
6016 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6018 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6019 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6022 29. When a sender address is verified, it is cached, to save repeating the test
6023 when there is more than one recipient in a message. However, when the
6024 verification involves a callout, it is possible for different callout
6025 options to be set for different recipients. It is too complicated to keep
6026 track of this in the cache, so now Exim always runs a verification when a
6027 callout is required, relying on the callout cache for the optimization.
6028 The overhead is duplication of the address routing, but this should not be
6031 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6032 to be verified with callout=postmaster, and the main callout worked but the
6033 postmaster check failed, the verification correctly failed. However, if a
6034 subsequent RCPT command asked for sender verification *without* the
6035 postmaster check, incorrect caching caused this verification also to fail,
6038 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6039 it was not caching the DNS options (qualify_single, search_parents) that
6040 were used when the lookup failed. A subsequent lookup with different
6041 options therefore always gave the same answer, though there were cases
6042 where it should not have. (Example: a "domains = !$mx_any" option on a
6043 dnslookup router: the "domains" option is always processed without any
6044 widening, but the router might have qualify_single set.) Now Exim uses the
6045 cached value only when the same options are set.
6047 32. Added John Jetmore's "exipick" utility to the distribution.
6049 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6050 than a timeout (e.g. a certificate is required, and is not provided), an
6051 Exim server now closes the connection immediately. Previously it waited for
6052 the client to close - but if the client is SSL, it seems that they each
6053 wait for each other, leading to a delay before one of them times out.
6055 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6056 maintained 0.8.x compatibility because I don't think many are using it, and
6057 it is clearly obsolete.
6059 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6062 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6063 client certificate was expired. A simple patch fixes this, though I don't
6064 understand the full logic of why the verify callback is called multiple
6067 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6068 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6069 which causes problems with some clients (such as the Certicom SSL Plus
6070 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6071 disables the coutermeasure allowing Eudora to connect."
6073 38. Exim was not checking that a write() to a log file succeeded. This could
6074 lead to Bad Things if a log got too big, in particular if it hit a file
6075 size limit. Exim now panics and dies if it cannot write to a log file, just
6076 as it does if it cannot open a log file.
6078 39. Modified OS/Makefile-Linux so that it now contains
6080 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6082 The two -D definitions ensure that Exim is compiled with large file
6083 support, which makes it possible to handle log files that are bigger than
6086 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6087 instance) a domain was checked against a named list that involved a lookup,
6088 causing $domain_data to be set, then another domain was checked against the
6089 same list, then the first domain was re-checked, the value of $domain_data
6090 after the final check could be wrong. In particular, if the second check
6091 failed, it could be set empty. This bug probably also applied to
6094 41. The strip_trailing_dot option was not being applied to the address given
6095 with the -f command-line option.
6097 42. The code for reading a message's header from the spool was incrementing
6098 $received_count, but never initializing it. This meant that the value was
6099 incorrect (doubled) while delivering a message in the same process in which
6100 it was received. In the most common configuration of Exim, this never
6101 happens - a fresh exec is done - but it can happen when
6102 deliver_drop_privilege is set.
6104 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6105 it now includes up to 150 characters of the unexpected data in the log
6108 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6109 and building data strings. The size of both of these buffers was 10 000
6110 bytes - far larger than anybody would *ever* want, thought I. Needless to
6111 say, somebody hit the limit. I have increased the maximum line length to
6112 20 000 and the maximum data length of concatenated lines to 100 000. I have
6113 also fixed two bugs, because there was no checking on these buffers. Tsk,
6114 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6115 buffer is too small.
6117 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6118 lsearch lookups. Now it does.
6120 46. When parsing a route_list item in a manualroute router, a fixed-length
6121 buffer was used for the list of hosts. I made this 1024 bytes long,
6122 thinking that nobody would ever have a list of hosts that long. Wrong.
6123 Somebody had a whole pile of complicated expansion conditions, and the
6124 string was silently truncated, leading to an expansion error. It turns out
6125 that it is easier to change to an unlimited length (owing to other changes
6126 that have happened since this code was originally written) than to build
6127 structure for giving a limitation error. The length of the item that
6128 expands into the list of hosts is now unlimited.
6130 47. The lsearch lookup could not handle data where the length of text line was
6131 more than 4095 characters. Such lines were truncated, leading to shortened
6132 data being returned. It should now handle lines of any length.
6134 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6135 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6138 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6139 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6140 no longer happens in comment lines. A list of replacements is now placed
6141 at the head of all of the source files, except those whose only change is
6142 to replace PERL_COMMAND in the very first #! line.
6144 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6145 messages on the queue, with a bottom-up merge sort, using code contributed
6146 by Michael Haardt. This should make operations like -bp somewhat faster on
6147 large queues. It won't affect queue runners, except when queue_run_in_order
6150 51. Installed eximstats 1.31 in the distribution.
6152 52. Added support for SRV lookups to the dnslookup router.
6154 53. If an ACL referred to $message_body or $message_body_end, the value was not
6155 reset for any messages that followed in the same SMTP session.
6157 54. The store-handling optimization for building very long strings was not
6158 differentiating between the different store pools. I don't think this
6159 actually made any difference in practice, but I've tidied it.
6161 55. While running the routers to verify a sender address, $sender_address
6162 was still set to the sender address. This is wrong, because when routing to
6163 send a bounce to the sender, it would be empty. Therefore, I have changed
6164 it so that, while verifying a sender address, $sender_address is set to <>.
6165 (There is no change to what happens when verifying a recipient address.)
6167 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6168 target A or AAAA records (if not already returned) without resetting the
6169 qualify_single or search_parents options of the DNS resolver. These are
6170 inappropriate in this case because the targets of MX and SRV records must
6171 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6172 target that, when qualified, matched something in the local domain. These
6173 two options are now turned off when doing these lookups.
6175 57. It seems that at least some releases of Reiserfs (which does not have the
6176 concept of a fixed number of inodes) returns zero and not -1 for the
6177 number of available inodes. This interacted badly with check_spool_inodes,
6178 which assumed that -1 was the "no such thing" setting. What I have done is
6179 to check that the total number of inodes is greater than zero before doing
6180 the test of how many are available.
6182 58. When a "warn" ACL statement has a log_message modifier, the message is
6183 remembered, and not repeated. This is to avoid a lot of repetition when a
6184 message has many recipients that cause the same warning to be written.
6185 However, Exim was preserving the list of already written lines for an
6186 entire SMTP session, which doesn't seem right. The memory is now reset if a
6187 new message is started.
6189 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6190 debugging output unless log_rewrite was also set.
6192 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6193 of (int)(handle) when we know that handle contains (void *)(-1).
6195 61. The Exim daemon panic-logs an error return when it closes the incoming
6196 connection. However "connection reset by peer" seems to be common, and
6197 isn't really an error worthy of noting specially, so that particular error
6200 62. When Exim is trying to find all the local interfaces, it used to panic and
6201 die if the ioctl to get the interface flags failed. However, it seems that
6202 on at least one OS (Solaris 9) it is possible to have an interface that is
6203 included in the list of interfaces, but for which you get a failure error
6204 for this call. This happens when the interface is not "plumbed" into a
6205 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6206 failure of the "get flags" call assumes that the interface is down.
6208 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6209 makes life easier for people who are doing arithmetic on fields extracted
6210 from dates, where you often get leading zeros that should not be
6211 interpreted as octal.
6213 64. Added qualify_domain to the redirect router, to override the global
6216 65. If a pathologically long header line contained very many addresses (the
6217 report of this problem mentioned 10 000) and each of them was rewritten,
6218 Exim could use up a very large amount of memory. (It kept on making new
6219 copies of the header line as it rewrote, and never released the old ones.)
6220 At the expense of a bit more processing, the header rewriting function has
6221 been changed so that it no longer eats memory in this way.
6223 66. The generation of the Received: header has been moved from the time that a
6224 message starts to be received, to the time that it finishes. The timestamp
6225 in the Received: header should now be very close to that of the <= log
6226 line. There are two side-effects of this change:
6228 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6229 logged header lines no longer include the local Received: line, because
6230 it has not yet been created. The same applies to a copy of the message
6231 that is returned to a non-SMTP sender when a message is rejected.
6233 (b) When a filter file is tested using -bf, no additional Received: header
6234 is added to the test message. After some thought, I decided that this
6237 This change does not affect the value of $received_for. It is still set
6238 after address rewriting, but before local_scan() is called.
6240 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6242 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6243 gave an unhelpful panic error message, and a defer error. I have managed to
6244 change this behaviour so that it now rejects any supplied certificate,
6245 which seems right, as the list of acceptable certificates is empty.
6247 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6248 gave an unhelpful defer error. I have not managed to make this reject any
6249 supplied certificates, but the error message it gives is "no certificate
6250 supplied", which is not helpful.
6252 70. exigrep's output now also includes lines that are not associated with any
6253 message, but which match the given pattern. Implemented by a patch from
6254 Martin Sluka, which also tidied up the Perl a bit.
6256 71. Recipient callout verification, like sender verification, was using <> in
6257 the MAIL FROM command. This isn't really the right thing, since the actual
6258 sender may affect whether the remote host accepts the recipient or not. I
6259 have changed it to use the actual sender in the callout; this means that
6260 the cache record is now keyed on a recipient/sender pair, not just the
6261 recipient address. There doesn't seem to be a real danger of callout loops,
6262 since a callout by the remote host to check the sender would use <>.
6263 [SEE ABOVE: changed after hitting problems.]
6265 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6266 temporary errors. However, in the case of such a code being given after
6267 the end of a data transmission (i.e. after ".") Exim was failing to write
6268 a retry record for the message. (Yes, there was some broken host that was
6269 actually sending 8xx at this point.)
6271 73. An unknown lookup type in a host list could cause Exim to panic-die when
6272 the list was checked. (An example that provoked this was putting <; in the
6273 middle of a list instead of at the start.) If this happened during a DATA
6274 ACL check, a -D file could be left lying around. This kind of configuration
6275 error no longer causes Exim to die; instead it causes a defer error. The
6276 incident is still logged to the main and panic logs.
6278 74. Buglet left over from Exim 3 conversion. The message "too many messages
6279 in one connection" was written to the rejectlog but not the mainlog, except
6280 when address rewriting (yes!) was being logged.
6282 75. Added write_rejectlog option.
6284 76. When a system filter was run not as root (that is, when system_filter_user
6285 was set), the values of the $n variables were not being returned to the
6286 main process; thus, they were not subsequently available in the $sn
6289 77. Added +return_path_on_delivery log selector.
6291 78. A connection timeout was being treated differently from recipients deferred
6292 when testing hosts_max_try with a message that was older than the host's
6293 retry timeout. (The host should not be counted, thus allowing all hosts to
6294 be tried at least once before bouncing.) This may have been the cause of an
6295 occasionally reported bug whereby a message would remain on the queue
6296 longer than the retry timeout, but would be bounced if a delivery was
6297 forced. I say "may" because I never totally pinned down the problem;
6298 setting up timeout/retry tests is difficult. See also the next item.
6300 79. The ultimate address timeout was not being applied to errors that involved
6301 a combination of host plus message (for example, a timeout on a MAIL
6302 command). When an address resolved to a number of possible hosts, and they
6303 were not all tried for each delivery (e.g. because of hosts_max_try), a
6304 message could remain on the queue longer than the retry timeout.
6306 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6309 81. Fixed an obscure SMTP outgoing bug which required at least the following
6310 conditions: (a) there was another message waiting for the same server;
6311 (b) the server returned 5xx to all RCPT commands in the first message so
6312 that the message was not completed; (c) the server dropped the connection
6313 or gave a negative response to the RSET that Exim sends to abort the
6314 transaction. The observed case was a dropped connection after DATA that had
6315 been sent in pipelining mode. That is, the server had advertised PIPELINING
6316 but was not implementing it correctly. The effect of the bug was incorrect
6317 behaviour, such as trying another host, and this could lead to a crash.
6323 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6324 and daemon.c were passed as pointers to ints; they should have been
6325 pointers to socklen_t variables (which are typically unsigned ints).
6327 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6330 3. Fixed a really odd bug that affected only the testing scheme; patching a
6331 certain fixed string in the binary changed the value of another string that
6332 happened to be identical to the end of the original first string.
6334 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6335 name", it returns that address as the IP address. On some operating
6336 systems (e.g. Solaris), it also passes back the IP address string as the
6337 "host name". However, on others (e.g. Linux), it passes back an empty
6338 string. Exim wasn't checking for this, and was changing the host name to an
6339 empty string, assuming it had been canonicalized.
6341 5. Although rare, it is permitted to have more than one PTR record for a given
6342 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6343 all the names associated with an address, because they do in Solaris.
6344 However, it seems that they do not in Linux for data that comes from the
6345 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6346 I found this out when I moved to a new Linux workstation and tried to run
6347 the Exim test suite.
6349 To get round this problem I have changed the code so that it now does its
6350 own call to the DNS to look up PTR records when searching for a host name.
6351 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6352 addresses that are only in /etc/hosts are still found.
6354 This behaviour is, however, controlled by an option called host_lookup_
6355 order, which defaults to "bydns:byaddr". If people want to use the other
6356 order, or indeed, just use one or the other means of lookup, they can
6357 specify it in this variable.
6359 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6360 some operating systems, this comes back from gethostbyaddr() as an empty
6361 string, and this is what Exim used to test for. However, it seems that in
6362 other systems, "." is yielded. Exim now tests for this case too.
6364 7. The values of check_spool_space and check_log_space are now held internally
6365 as a number of kilobytes instead of an absolute number of bytes. If a
6366 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6367 kilobyte. This means that much larger values can be stored.
6369 8. Exim monitor: an attempt to get the action menu when not actually pointing
6370 at a message produces an empty menu entitled "No message selected". This
6371 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6372 no entries in it ("Shell widget menu has zero width and/or height"). So I
6373 have added a single, blank menu entry in this case.
6375 9. Added ${quote_local_part.
6377 10. MIME decoding is now applied to the contents of Subject: header lines when
6380 11. Now that a reference to $sender_host_address automatically causes a reverse
6381 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6382 host lookup before query-style lookups in lists that might use this
6383 variable. This has therefore been abolished, and the "net-" prefix is no
6384 longer necessary for query-style lookups.
6386 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6387 have been a typo for LFLAGS, so it has been changed.
6389 13. The install script calls Exim with "-C /dev/null" in order to find the
6390 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6391 to be output. However, since Exim outputs its version number before the
6392 error, it didn't break the script. It just looked ugly. I fixed this by
6393 always allowing "-C /dev/null" if the caller is root.
6395 14. Ignore overlarge ACL variable number when reading spool file - insurance
6396 against a later release with more variables having written the file.
6398 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6399 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6400 this, as well as the form without the "IPv6" on the front.
6402 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6403 OS/Makefile-Darwin file.
6405 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6406 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6408 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6409 variable. However, some OS use ints for inodes. I've added cast to long int
6410 to get rid of the compiler warning.
6412 19. I had forgotten to lock out "/../" in configuration file names when
6413 ALT_CONFIG_PREFIX was set.
6415 20. Routers used for verification do not need to specify transports. However,
6416 if such a router generated a host list, and callout was configured, Exim
6417 crashed, because it could not find a port number from the (non-existent)
6418 transport. It now assumes port 25 in this circumstance.
6420 21. Added the -t option to exigrep.
6422 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6423 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6424 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6425 them is set, LOOKUP_LSEARCH is forced.
6427 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6429 24. Added sender and host information to the "rejected by local_scan()" log
6430 line; previously there was no indication of these.
6432 25. Added .include_if_exists.
6434 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6435 Linux. It turns out that not all file systems support this. Apparently some
6436 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6437 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6438 not supported on the file descriptor, is now ignored when Exim is trying to
6439 sync a directory. This applies only to Linux.
6441 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6443 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6444 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6447 29. Added support for maildirsize files from supplied patch (modified a bit).
6449 30. The use of :fail: followed by an empty string could lead Exim to respond to
6450 sender verification failures with (e.g.):
6452 550 Verification failed for <xxx>
6453 550 Sender verify failed
6455 where the first response line was missing the '-' that indicates it is not
6456 the final line of the response.
6458 31. The loop for finding the name of the user that called Exim had a hardwired
6459 limit of 10; it now uses the value of finduser_retries, which is used for
6460 all other user lookups.
6462 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6465 33. Exim was neglecting to zero errno before one call of strtol() when
6466 expanding a string and expecting an integer value. On some systems this
6467 resulted in spurious "integer overflow" errors. Also, it was casting the
6468 result into an int without checking.
6470 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6471 did not work. The code looks as if it has *never* worked, though it appears
6472 to have been documented since at least release 1.62. I have made it work.
6474 35. The "timeout_DNS" error in retry rules, also documented since at least
6475 1.62, also never worked. As it isn't clear exactly what this means, and
6476 clearly it isn't a major issue, I have abolished the feature by treating it
6477 as "timeout", and writing a warning to the main and panic logs.
6479 36. The display of retry rules for -brt wasn't always showing the error code
6482 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6483 timeout_connect_A, timeout_connect_MX.
6485 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6486 to the empty sender.
6488 39. The daemon was not analysing the content of -oX till after it had closed
6489 stderr and disconnected from the controlling terminal. This meant that any
6490 syntax errors were only noted on the panic log, and the return code from
6491 the command was 0. By re-arranging the code a little, I've made the
6492 decoding happen first, so such errors now appear on stderr, and the return
6493 code is 1. However, the actual setting up of the sockets still happens in
6494 the disconnected process, so errors there are still only recorded on the
6497 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6498 connections (as happens on some IP stacks) was logged at start up time as
6499 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6500 it from "IPv6 and IPv4", which means that two separate sockets are being
6503 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6504 says whether AF_INET or AF_INET6 was passed as an argument.
6506 42. Exiwhat output was messed up when time zones were included in log
6509 43. Exiwhat now gives more information about the daemon's listening ports,
6510 and whether -tls-on-connect was used.
6512 44. The "port" option of the smtp transport is now expanded.
6514 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6515 silently ignored. Now an error message is written to the main and panic
6518 46. There's a new ACL modifier called "logwrite" which writes to a log file
6519 as soon as it is encountered.
6521 47. Added $local_user_uid and $local_user_gid at routing time.
6523 48. Exim crashed when trying to verify a sender address that was being
6526 49. Exim was recognizing only a space character after ".include". It now also
6527 recognizes a tab character.
6529 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6530 extracting the relevant information from the specification. The man page no
6531 longer contains scrambled data for the -d option, and I've added a section
6532 at the front about calling Exim under different names.
6534 51. Added "extra_headers" argument to the "mail" command in filter files.
6536 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6539 53. Installed eximstats 1.29.
6541 54. Added transport_filter_timeout as a generic transport option.
6543 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6544 Cc: header lines. This was required by RFC 822, but it not required by RFC
6547 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6548 incoming messages that did not have them. Now it does so only if the
6549 message originates locally, that is, if there is no associated remote host
6550 address. When Resent- header lines are present, this applies to the Resent-
6551 lines rather than the non-Resent- lines.
6553 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6554 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6556 58. Messages for configuration errors now include the name of the main
6557 configuration file - useful now that there may be more than one file in a
6558 list (.included file names were always shown).
6560 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6561 for those rare installations that do not start the daemon as root or run it
6562 setuid root. I've cut out the call to initgroups() if the daemon is not
6565 60. The Exim user and group can now be bound into the binary as text strings
6566 that are looked up at the start of Exim's processing.
6568 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6570 62. Added $mailstore_basename variable.
6572 63. Installed patch to sieve.c from Michael Haardt.
6574 64. When Exim failed to open the panic log after failing to open the main log,
6575 the original message it was trying to log was written to stderr and debug
6576 output, but if they were not available (the usual case in production), it
6577 was lost. Now it is written to syslog before the two lines that record the
6578 failures to open the logs.
6580 65. Users' Exim filters run in subprocesses under the user's uid. It is
6581 possible for a "deliver" command or an alias in a "personal" command to
6582 provoke an address rewrite. If logging of address rewriting is configured,
6583 this fails because the process is not running as root or exim. There may be
6584 a better way of dealing with this, but for the moment (because 4.30 needs
6585 to be released), I have disabled address rewrite logging when running a
6586 filter in a non-root, non-exim process.
6592 1. The buildconfig auxiliary program wasn't quoting the value set for
6593 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6594 not defined. This bug was masked in 4.22 by the effect that was fixed in
6597 2. Some messages that were rejected after a message id was allocated were
6598 shown as "incomplete" by exigrep. It no longer does this for messages that
6599 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6601 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6602 have allow_domain_literals set, the ID did not get logged in the <= line.
6603 Domain literals are now always recognized in Message-ID: header lines.
6605 4. The first argument for a ${extract expansion item is the key name or field
6606 number. Leading and trailing spaces in this item were not being ignored,
6607 causing some misleading effects.
6609 5. When deliver_drop_privilege was set, single queue runner processes started
6610 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6611 same command in the process it spins off) were not dropping privilege.
6613 6. When the daemon running as "exim" started a queue runner, it always
6614 re-executed Exim in the spun-off process. This is a waste of effort when
6615 deliver_drop_privilege is set. The new process now just calls the
6616 queue-runner function directly.
6622 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6625 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6626 ignored. Though the use of -oP was forcing the writing of a pid file, it
6627 was always written to the default place.
6629 3. If the message "no IP address found for host xxxx" is generated during
6630 incoming verification, it is now followed by identification of the incoming
6631 connection (so you can more easily find what provoked it).
6633 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6635 5. Added some features to "harden" Exim a bit more against certain attacks:
6637 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6638 be put in Local/Makefile. This is like the never_users runtime option,
6639 but it cannot be overridden. The default setting is "root".
6641 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6642 prefix string with which any file named in a -C command line option
6645 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6646 is retained for -C and -D only if the caller of Exim is root. Without
6647 it, the exim user may also use -C and -D and retain privilege.
6649 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6650 command line option is disabled.
6652 6. Macro names set by the -D option must start with an upper case letter, just
6653 like macro names defined in the configuration file.
6655 7. Added "dereference=" facility to LDAP.
6657 8. Two instances of the typo "uknown" in the source files are fixed.
6659 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6660 the Configure-Makefile script screwed up while processing it.
6662 10. Incorporated PCRE 4.4.
6664 11. The SMTP synchronization check was not operating right at the start of an
6665 SMTP session. For example, it could not catch a HELO sent before the client
6666 waited for the greeting. There is now a check for outstanding input at the
6667 point when the greeting is written. Because of the duplex, asynchronous
6668 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6669 way, but not yet received, when the check is performed.
6671 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6672 on TCP/IP sockets, because this apparently causes some broken clients to
6675 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6676 unchanged) from the Cygwin maintainer.
6678 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6679 was supported instead of testing for mbx. Effectively a typo.
6681 15. The spa authenticator server code was not checking that the input it
6682 received was valid base64.
6684 16. The debug output line for the "set" modifier in ACLs was not showing the
6685 name of the variable that was being set.
6687 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6689 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6690 lookup on its own. Something else was needed to trigger the lookup. For
6691 example, a match in host_lookup or the need for a host name in a host list.
6692 Now, if $sender_host_name is referenced and the host name has not yet been
6693 looked up, a lookup is performed. If the lookup fails, the variable remains
6694 empty, and $host_lookup_failed is set to "1".
6696 19. Added "eqi" as a case-independent comparison operator.
6698 20. The saslauthd authentication condition could segfault if neither service
6699 nor realm was specified.
6701 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6702 error message that was logged was misleading, and incoming SMTP
6703 connections were dropped. The message is now more accurate, and temporary
6704 errors are given to SMTP connections.
6706 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6707 (except RSET) with a 421 error, until QUIT is received. However, it was
6708 failing to send a response to QUIT.
6710 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6711 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6712 ACL is now run after the helo_try_verify_hosts code.
6714 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6715 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6716 case-independent, so other case variants are also recognized). Apparently
6717 some systems use these upper case variants.
6719 25. If more than two messages were waiting for the same host, and a transport
6720 filter was specified for the transport, Exim sent two messages over the
6721 same TCP/IP connection, and then failed with "socket operation on non-
6722 socket" when it tried to send the third.
6724 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6726 27. The extern definition of crypt16() in expand.c was not being excluded when
6727 the OS had its own crypt16() function.
6729 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6730 as a preferred synonym for return_size_limit, both as an option and as an
6733 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6735 30. Changed the default configuration ACL to relax the local part checking rule
6736 for addresses that are not in any local domains. For these addresses,
6737 slashes and pipe symbols are allowed within local parts, but the sequence
6738 /../ is explicitly forbidden.
6740 31. SPA server authentication was not clearing the challenge buffer before
6743 32. log_message in a "warn" ACL statement was writing to the reject log as
6744 well as to the main log, which contradicts the documentation and doesn't
6745 seem right (because no rejection is happening). So I have stopped it.
6747 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6748 However, I am unable to do any testing of this.
6750 34. Fixed an infelicity in the appendfile transport. When checking directories
6751 for a mailbox, to see if any needed to be created, it was accidentally
6752 using path names with one or more superfluous leading slashes; tracing
6753 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6755 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6756 contents are added to the log line that is written for every discarded
6757 recipient. (Previously a log_message setting was ignored.)
6759 36. The ${quote: operator now quotes the string if it is empty.
6761 37. The install script runs exim in order to find its version number. If for
6762 some reason other than non-existence or emptiness, which it checks, it
6763 could not run './exim', it was installing it with an empty version number,
6764 i.e. as "exim-". This error state is now caught, and the installation is
6767 38. An argument was missing from the function that creates an error message
6768 when Exim fails to connect to the socket for saslauthd authentication.
6769 This could cause Exim to crash, or give a corrupted message.
6771 39. Added isip, isip4, and isip6 to ${if conditions.
6773 40. The ACL variables $acl_xx are now saved with the message, and can be
6774 accessed later in routers, transports, and filters.
6776 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6777 strings in the file are not string-expanded.
6779 42. If a MAIL command specified a SIZE value that was too large to fit into an
6780 int variable, the check against message_size_limit failed. Such values are
6781 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6782 day this will have to be increased, but I don't think I want to be around
6783 when emails are that large.
6790 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6791 iconv() is not standard in FreeBSD.
6793 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6794 IPv6 enabled. The observed symptom was a segmentation fault on return from
6795 the function os_common_find_running_interfaces() in src/os.c.
6797 3. In the check_special_case() function in daemon.c I had used "errno" as an
6798 argument name, which causes warnings on some systems. This was basically a
6799 typo, since it was named "eno" in the comments!
6801 4. The code that waits for the clock to tick (at a resolution of some fraction
6802 of a second) so as to ensure message-id uniqueness was always waiting for
6803 at least one whole tick, when it could have waited for less. [This is
6804 almost certainly not relevant at current processor speeds, where it is
6805 unlikely to ever wait at all. But we try to future-proof.]
6807 5. The function that sleeps for a time interval that includes fractions of a
6808 second contained a race. It did not block SIGALRM between setting the
6809 timer, and suspending (a couple of lines later). If the interval was short
6810 and the sigsuspend() was delayed until after it had expired, the suspension
6811 never ended. On busy systems this could lead to processes getting stuck for
6814 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6815 process, before it forks any delivery processes. The open lookup caching
6816 mechanism meant that the open file or database connection was passed into
6817 the delivery process. The problem was that delivery processes always tidy
6818 up cached lookup data. This could cause a problem for the next delivery
6819 process started by the queue runner, because the external queue runner
6820 process does not know about the closure. So the next delivery process
6821 still has data in the lookup cache. In the case of a file lookup, there was
6822 no problem because closing a file descriptor in a subprocess doesn't affect
6823 the parent. However, if the lookup was caching a connection to a database,
6824 the connection was closed, and the second delivery process was likely to
6825 see errors such as "PGSQL: query failed: server closed the connection
6826 unexpectedly". The problem has been fixed by closing all cached lookups
6827 in a queue runner before running a delivery process.
6829 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6830 seem to have the "const" qualifier which it has on other OS. I've
6833 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6834 *of the same type* (client or server) with the same public name that an
6835 error should be diagnosed.
6837 9. When Exim looked up a host name for an IP address, but failed to find the
6838 original IP address when looking up the host name (a safety check), it
6839 output the message "<ip address> does not match any IP for NULL", which was
6840 confusing, to say the least. The bug was that the host name should have
6841 appeared instead of "NULL".
6843 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6844 user that is built into the binary, and the -C or -D options is used, root
6845 privilege is dropped before the configuration file is read. In addition,
6846 logging is switched to stderr instead of the normal log files. If the
6847 configuration then re-defines the Exim user, the unprivileged environment
6848 is probably not what is expected, so Exim logs a panic warning message (but
6851 However, if deliver_drop_privilege is set, the unprivileged state may well
6852 be exactly what is intended, so the warning has been cut out in that case,
6853 and Exim is allowed to try to write to its normal log files.
6859 1. smtp_return_error_details was not giving details for temporary sender
6860 or receiver verification errors.
6862 2. Diagnose a configuration error if two authenticators have the same public
6865 3. Exim used not to create the message log file for a message until the first
6866 delivery attempt. This could be confusing when incoming messages were held
6867 for policy or load reasons. The message log file is now created at the time
6868 the message is received, and an initial "Received" line is written to it.
6870 4. The automatically generated man page for command line options had a minor
6871 bug that caused no ill effects; however, a more serious problem was that
6872 the procedure for building the man page automatically didn't always
6873 operate. Consequently, release 4.20 contains an out-of-date version. This
6874 shouldn't happen again.
6876 5. When building Exim with embedded Perl support, the script that builds the
6877 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6878 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6880 6. The freeze_tell option was not being used for messages that were frozen on
6881 arrival, either by an ACL or by local_scan().
6883 7. Added the smtp_incomplete_transaction log selector.
6885 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6886 was accepting AUTH without a new EHLO.
6888 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6889 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6892 10. Logging of TCP/IP connections (when configured) now happens in the main
6893 daemon process instead of the child process, so that the TCP/IP connection
6894 count is more accurate (but it can never be perfect).
6896 11. The use of "drop" in a nested ACL was not being handled correctly in the
6897 outer ACL. Now, if condition failure induced by the nested "drop" causes
6898 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6899 or "require"), the connection is dropped.
6901 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6902 that yield "discard" can now be used with an "accept" or a "discard" verb,
6903 but an error is generated for any others (because I can't see a useful way
6904 to define what should happen).
6906 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6907 are now processed in the same way as lines in the Exim configuration file.
6908 In particular, continuation lines are supported.
6910 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6912 15. Added -ti meaning -t -i.
6914 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6915 domains, and warn by logging if others are found.
6917 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6918 returned from GIFCONF when Exim is trying to find the list of interfaces on
6919 a host. The code in os.c has been modified to copy each ifreq to an aligned
6920 structure in all cases.
6922 Also, in some cases, the returned ifreq's were being copied to a 'struct
6923 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6924 means the last couple of bytes of an IPv6 address could be chopped if the
6925 ifreq contained only a normal sockaddr (14 bytes storage).
6927 18. Named domain lists were not supported in the hosts_treat_as_local option.
6928 An entry such as +xxxx was not recognized, and was treated as a literal
6931 19. Ensure that header lines added by a DATA ACL are included in the reject log
6932 if the ACL subsequently rejects the message.
6934 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6935 MD5 (which is deprecated).
6937 21. When testing a filter file using -bf, Exim was writing a message when it
6938 took the sender from a "From " line in the message, but it was not doing so
6939 when it took $return_path from a Return-Path: header line. It now does.
6941 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6942 with a valid header line field name (a series of printing characters
6943 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6945 23. Changed "disc" in the source to "disk" to conform to the documentation and
6946 the book and for uniformity.
6948 24. Ignore Sendmail's -Ooption=value command line item.
6950 25. When execve() failed while trying to run a command in a pipe transport,
6951 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6952 could be confused with a return value of 69 from the command itself. This
6953 has been changed to 127, the value the shell returns if it is asked to run
6954 a non-existent command. The wording for the related log line suggests a
6955 non-existent command as the problem.
6957 26. If received_header_text expands to an empty string, do not add a Received:
6958 header line to the message. (Well, it adds a token one on the spool, but
6959 marks it "old" so that it doesn't get used or transmitted.)
6961 27. Installed eximstats 1.28 (addition of -nt option).
6963 28. There was no check for failure on the call to getsockname() in the daemon
6964 code. This can fail if there is a shortage of resources on the system, with
6965 ENOMEM, for example. A temporary error is now given on failure.
6967 29. Contrary to the C standard, it seems that in some environments, the
6968 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6969 program. Exim now does this explicitly; it affects the formatting of
6970 timestamps using strftime().
6972 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6973 complaints. I've now initialized all the variables, to avoid this.
6975 32. Header lines added by a system filter were not being "seen" during
6976 transport-time rewrites.
6978 33. The info_callback() function passed to OpenSSL is set up with type void
6979 (*)(SSL *, int, int), as described somewhere. However, when calling the
6980 function (actually a macro) that sets it up, the type void(*)() is
6981 expected. I've put in a cast to prevent warnings from picky compilers.
6983 34. If a DNS black list lookup found a CNAME record, but there were no A
6984 records associated with the domain it pointed at, Exim crashed.
6986 35. If a DNS black list lookup returned more than one A record, Exim ignored
6987 all but the first. It now scans all returned addresses if a particular IP
6988 value is being sought. In this situation, the contents of the
6989 $dnslist_value variable are a list of all the addresses, separated by a
6992 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6993 to accept a host name and all its aliases if the forward lookup for any of
6994 them yielded the IP address of the incoming connection. Now it accepts only
6995 those names whose forward lookup yields the correct IP address. Any other
6996 names are discarded. This closes a loophole whereby a rogue DNS
6997 administrator could create reverse DNS records to break through a
6998 wildcarded host restriction in an ACL.
7000 37. If a user filter or a system filter that ran in a subprocess used any of
7001 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7002 the wrong values were passed to the pipe command ($thisaddress had the
7003 value of $0, $0 had the value of $1, etc). This bug was introduced by
7004 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7006 38. Improved the line breaking for long SMTP error messages from ACLs.
7007 Previously, if there was no break point between 40 and 75 characters, Exim
7008 left the rest of the message alone. Two changes have been made: (a) I've
7009 reduced the minimum length to 35 characters; (b) if it can't find a break
7010 point between 35 and 75 characters, it looks ahead and uses the first one
7011 that it finds. This may give the occasional overlong line, but at least the
7012 remaining text gets split now.
7014 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7015 file descriptors might be low, and that setting 1000 would always raise it.
7016 It turns out that in some environments, the limit is already over 1000 and
7017 that lowering it causes trouble. So now Exim takes care not to decrease it.
7019 40. When delivering a message, the value of $return_path is set to $sender_
7020 address at the start of routing (routers may change the value). By an
7021 oversight, this default was not being set up when an address was tested by
7022 -bt or -bv, which affected the outcome if any router or filter referred to
7025 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7026 log only when "message" or "log_message" are set. However, if one of the
7027 conditions was an address verification, or a call to a nested ACL, the
7028 messages generated by the underlying test were being passed through. This
7029 no longer happens. The underlying message is available in $acl_verify_
7030 message for both "message" and "log_message" expansions, so it can be
7031 passed through if needed.
7033 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7034 new expansion $bh_ to give the encoded byte string without charset
7035 translation. Translation happens only if iconv() is available; HAVE_ICONV
7036 indicates this at build time. HEADERS_CHARSET gives the charset to
7037 translate to; headers_charset can change it in the configuration, and
7038 "headers charset" can change it in an individual filter file.
7040 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7041 that creates RFC 2047 encoded "words" labels them as that charset instead
7042 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7043 expansion operator; (ii) when Exim creates a From: line for a local
7044 message; (iii) when a header line is rewritten to include a "phrase" part.
7046 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7047 buggy, causing it to skip the first lines of messages whose message ID
7048 ended in 'D'. This would not have bitten before Exim release 4.14, because
7049 message IDs were unlikely to end in 'D' before then. The effect was to have
7050 incorrect size information for certain domains.
7052 45. #include "config.h" was missing at the start of the crypt16.c module. This
7053 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7056 46. If there was a timeout during a "random" callout check, Exim treated it as
7057 a failure of the random address, and carried on sending RSET and the real
7058 address. If the delay was just some slowness somewhere, the response to the
7059 original RCPT would be taken as a response to RSET and so on, causing
7060 mayhem of various kinds.
7062 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7063 when I implemented it. It didn't allow for the fact that some option values
7064 may legitimately be negative (e.g. size_addition), and it didn't even do
7065 the right test for positive values.
7067 48. Domain names in DNS records are case-independent. Exim always looks them up
7068 in lower case. Some resolvers return domain names in exactly the case they
7069 appear in the zone file, that is, they may contain uppercase letters. Not
7070 all resolvers do this - some return always lower case. Exim was treating a
7071 change of case by a resolver as a change of domain, similar to a widening
7072 of a domain abbreviation. This triggered its re-routing code and so it was
7073 trying to route what was effectively the same domain again. This normally
7074 caused routing to fail (because the router wouldn't handle the domain
7075 twice). Now Exim checks for this case specially, and just changes the
7076 casing of the domain that it ultimately uses when it transmits the message
7079 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7082 50. If a filter generated a file delivery with a non-absolute name (possible if
7083 no home directory exists for the router), the forbid_file option was not
7086 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7087 the existing equality matching.
7089 52. Exim was using ints instead of ino_t variables in some places where it was
7090 dealing with inode numbers.
7092 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7093 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7094 called TMPDIR, and if it finds it is different, it changes its value.
7096 54. The smtp_printf() function is now made available to local_scan() so
7097 additional output lines can be written before returning. There is also an
7098 smtp_fflush() function to enable the detection of a dropped connection.
7099 The variables smtp_input and smtp_batched_input are exported to
7102 55. Changed the default runtime configuration: the message "Unknown user"
7103 has been removed from the ACL, and instead placed on the localuser router,
7104 using the cannot_route_message feature. This means that any verification
7105 failures that generate their own messages won't get overridden. Similarly,
7106 the "Unrouteable address" message that was in the ACL for unverifiable
7107 relay addresses has also been removed.
7109 56. Added hosts_avoid_esmtp to the smtp transport.
7111 57. The exicyclog script was not checking for the esoteric option
7112 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7113 will work only if exicyclog is run under the appropriate euid.
7115 58. Following a discussion on the list, the rules by which Exim recognises line
7116 endings on incoming messages have been changed. The -dropcr and drop_cr
7117 options are now no-ops, retained only for backwards compatibility. The
7118 following line terminators are recognized: LF CRLF CR. However, special
7119 processing applies to CR:
7121 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7122 nor a local message in the state where . is a terminator.
7124 (ii) If a bare CR is encountered in a header line, an extra space is added
7125 after the line terminator so as not to end the header. The reasoning
7126 behind this is that bare CRs in header lines are most likely either
7127 to be mistakes, or people trying to play silly games.
7129 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7130 was being incorrectly given as 18 bytes larger than it should have been.
7131 This is a VOB (very old bug).
7133 60. This may never have affected anything current, but just in case it has:
7134 When the local host is found other than at the start of a list of hosts,
7135 the local host, those with the same MX, and any that follow, are discarded.
7136 When the list in question was part of a longer list of hosts, the following
7137 hosts (not currently being processed) were also being discarded. This no
7138 longer happens. I'm not sure if this situation could ever has previously
7141 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7144 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7145 header. This is something that is recommended in a new Internet Draft, and
7146 is something that is documented as being done by Sendmail. There are two
7147 possible values. For messages generated by the autoreply transport, Exim
7150 Auto-Submitted: auto-replied
7152 whereas for all other generated messages (e.g. bounces) it adds
7154 Auto-Submitted: auto-generated
7156 63. The "personal" condition in filters now includes a test for the
7157 Auto-Submitted: header. If it contains the string "auto-" the message it
7158 not considered personal.
7160 64. Added rcpt_include_affixes as a generic transport option.
7162 65. Added queue_only_override (default true).
7164 66. Added the syslog_duplication option.
7166 67. If what should have been the first header line of a message consisted of
7167 a space followed by a colon, Exim was mis-interpreting it as a header line.
7168 It isn't of course - it is syntactically invalid and should therefore be
7169 treated as the start of the message body. The misbehaviour could have
7170 caused a number of strange effects, including loss of data in subsequent
7171 header lines, and spool format errors.
7173 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7174 client host had authenticated. This control can now be exercised by an ACL
7175 for more flexibility.
7177 69. By default, callouts do not happen when testing with -bh. There is now a
7178 variant, -bhc, which does actually run the callout code, including
7179 consulting and updating the callout cache.
7181 70. Added support for saslauthd authentication, courtesy of Alexander
7184 71. If statvfs() failed on the spool or log directories while checking their
7185 size for availability, Exim confusingly gave the error "space shortage".
7186 Furthermore, in debugging mode it crashed with a floating point exception.
7187 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7188 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7189 serious problem, Exim now writes to the main and panic logs when this
7190 happens, with details of the failure. It then refuses to accept the
7191 incoming message, giving the message "spool directory problem" or "log
7192 directory problem" with a 421 code for SMTP messages.
7194 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7195 0, 1, and 2 exist, because some OS complain for execs without them (see
7196 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7197 descriptors. However, the code omitted to check that the open succeeded,
7198 causing mysterious errors if for some reason the permissions on /dev/null
7199 got screwed. Now Exim writes a message to the main and panic logs, and
7200 bombs out if it can't open /dev/null.
7202 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7203 interact so that it is all more flexible. It is supposed to remain
7204 backwards compatible. Also added extra_local_interfaces.
7206 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7207 to bomb out with an assertion failure - to the client this appears as a
7208 connection drop. This problem occurs in the part of the code that was taken
7209 from the Samba project. Fortunately, the assertion is in a very simple
7210 function, so I have fixed this by reproducing the function inline in the
7211 one place where it is called, and arranging for authentication to fail
7212 instead of killing the process with assert().
7214 75. The SPA client code was not working when the server requested OEM rather
7215 than Unicode encoding.
7217 76. Added code to make require_files with a specific uid setting more usable in
7218 the case where statting the file as root fails - usually a non-root-mounted
7219 NFS file system. When this happens and the failure is EACCES, Exim now
7220 forks a subprocess and does the per-uid checking as the relevant uid.
7222 77. Added process_log_path.
7224 78. If log_file_path was not explicitly set, a setting of check_log_space or
7225 check_log_inodes was ignored.
7227 79. If a space check for the spool or log partitions fails, the incident is now
7228 logged. Of course, in the latter case the data may get lost...
7230 80. Added the %p formatting code to string_format() so that it can be used to
7231 print addresses in debug_print(). Adjusted all the address printing in the
7232 debugging in store.c to use %p rather than %d.
7234 81. There was a concern that a line of code in smtp_in.c could overflow a
7235 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7236 initially expressed, the concern was not well-founded, because trailing
7237 spaces are removed early. However, if the trailing spaces were followed by
7238 a NULL, they did not get removed, so the overflow was possible. Two fixes
7241 (a) I re-wrote the offending code in a cleaner fashion.
7242 (b) If an incoming SMTP command contains a NULL character, it is rejected
7245 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7246 runs initgroups(), so that if the Exim user is in any additional groups,
7247 they will be used during message reception.
7253 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob