1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
64 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
65 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
66 now-deprecated earlier definitions used only specified the range up to TLS
67 1.2 (in the older-version library docs).
69 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
71 JH/15 Rework TLS client-side context management. Stop using a global, and
72 explicitly pass a context around. This enables future use of TLS for
73 connections to service-daemons (eg. malware scanning) while a client smtp
74 connection is using TLS; with cutthrough connections this is quite likely.
76 JH/16 Fix ARC verification to do AS checks in reverse order.
78 JH/17 Support a "tls" option on the ${readsocket } expansion item.
80 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
81 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
82 Previously the "utf8" would be re-prepended for every additional message.
84 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
85 Previously thery were accepted, resulting in issues when attempting to
86 forward messages to a non-supporting MTA.
88 PP/02 Let -n work with printing macros too, not just options.
90 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
91 one parent address was copied, and bogus data was used at delivery-logging
92 time. Either a crash (after delivery) or bogus log data could result.
93 Discovery and analysis by Tim Stewart.
95 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
96 Previously if the string ended mid-character, we did not insert the
97 promised '?' replacement.
103 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
104 When redis_servers is set to a list of > 1 element, and the Redis servers
105 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
106 case of MOVED into a DEFER case instead, thus moving the query onto the
107 next server in the list. For a cluster of N elements, all N servers must
108 be defined in redis_servers.
110 GF/02 Catch and remove uninitialized value warning in exiqsumm
111 Check for existence of @ARGV before looking at $ARGV[0]
113 JH/01 Replace the store_release() internal interface with store_newblock(),
114 which internalises the check required to safely use the old one, plus
115 the allocate and data copy operations duplicated in both (!) of the
116 extant use locations.
118 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
119 modifier. This matches the restriction on the commandline.
121 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
122 Previously only the last row was returned.
124 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
125 we assumed that tags in the header were well-formed, and parsed the
126 element content after inspecting only the first char of the tag.
127 Assumptions at that stage could crash the receive process on malformed
130 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
131 While running the DKIM ACL we operate on the Permanent memory pool so that
132 variables created with "set" persist to the DATA ACL. Also (at any time)
133 DNS lookups that fail create cache records using the Permanent pool. But
134 expansions release any allocations made on the current pool - so a dnsdb
135 lookup expansion done in the DKIM ACL releases the memory used for the
136 DNS negative-cache, and bad things result. Solution is to switch to the
137 Main pool for expansions.
138 While we're in that code, add checks on the DNS cache during store_reset,
139 active in the testsuite.
140 Problem spotted, and debugging aided, by Wolfgang Breyha.
142 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
143 When none of the hosts presented to a transport match an already-open
144 connection, close it and proceed with the list. Previously we would
145 queue the message. Spotted by Lena with Yahoo, probably involving
148 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
149 Previously a spurious "250 OK id=" response was appended to the proper
152 JH/08 The "support for" informational output now, which built with Content
153 Scanning support, has a line for the malware scanner interfaces compiled
154 in. Interface can be individually included or not at build time.
156 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
157 by the template makefile "src/EDITME". The "STREAM" support for an older
158 ClamAV interface method is removed.
160 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
161 rows affected is given instead).
163 JH/11 The runtime Berkeley DB library version is now additionally output by
164 "exim -d -bV". Previously only the compile-time version was shown.
166 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
167 SMTP connection. Previously, when one had more receipients than the
168 first, an abortive onward connection was made. Move to full support for
169 multiple onward connections in sequence, handling cutthrough connection
170 for all multi-message initiating connections.
172 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
173 routers. Previously, a multi-recipient message would fail to match the
174 onward-connection opened for the first recipient, and cause its closure.
176 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
177 a timeout on read on a GnuTLS initiating connection, resulting in the
178 initiating connection being dropped. This mattered most when the callout
179 was marked defer_ok. Fix to keep the two timeout-detection methods
182 JH/15 Relax results from ACL control request to enable cutthrough, in
183 unsupported situations, from error to silently (except under debug)
184 ignoring. This covers use with PRDR, frozen messages, queue-only and
187 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
189 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
190 metadata, resulting in a crash in free().
192 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
193 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
194 Broken also in d185889f4, with init system revamp.
196 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
197 Previously we abruptly closed the connection after reading a malware-
198 found indication; now we go on to read the "scan ok" response line,
201 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
202 ACL. Previously, a crash would result.
204 JH/19 Speed up macro lookups during configuration file read, by skipping non-
205 macro text after a replacement (previously it was only once per line) and
206 by skipping builtin macros when searching for an uppercase lead character.
208 JH/20 DANE support moved from Experimental to mainline. The Makefile control
209 for the build is renamed.
211 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
212 was allocated for every new TLS startup, meaning one per message. Fix
213 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
215 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
216 reported the original. Fix to report (as far as possible) the ACL
217 result replacing the original.
219 JH/23 Fix memory leak during multi-message connections using STARTTLS under
220 OpenSSL. Certificate information is loaded for every new TLS startup,
221 and the resources needed to be freed.
223 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
225 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
228 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
229 DATA response info to the (existing) per-recipient response info for
230 the "C=" log element. It can have useful tracking info from the
231 destination system. Patch from Simon Arlott.
233 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
234 length value. Previously this would segfault.
236 HS/02 Support Avast multiline protoocol, this allows passing flags to
237 newer versions of the scanner.
239 JH/28 Ensure that variables possibly set during message acceptance are marked
240 dead before release of memory in the daemon loop. This stops complaints
241 about them when the debug_store option is enabled. Discovered specifically
242 for sender_rate_period, but applies to a whole set of variables.
243 Do the same for the queue-runner and queue-list loops, for variables set
244 from spool message files. Do the same for the SMTP per-message loop, for
245 certain variables indirectly set in ACL operations.
247 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
248 as a multi-recipient message from a mailinglist manager). The coding had
249 an arbitrary cutoff number of characters while checking for more input;
250 enforced by writing a NUL into the buffer. This corrupted long / fast
251 input. The problem was exposed more widely when more pipelineing of SMTP
252 responses was introduced, and one Exim system was feeding another.
253 The symptom is log complaints of SMTP syntax error (NUL chars) on the
254 receiving system, and refused recipients seen by the sending system
255 (propating to people being dropped from mailing lists).
256 Discovered and pinpointed by David Carter.
258 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
259 replaced by the ${authresults } expansion.
261 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
263 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
264 allows proper process termination in container environments.
266 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
267 Previously the "final dot" had a newline after it; ensure it is CR,LF.
269 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
270 and "err_perm", deprecated since 4.83 when the RFC-defined words
271 "temperror" and "permerror" were introduced.
273 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
274 transport-filters or DKIM-signing. The restriction was lost in the
275 consolidation of verify-callout and delivery SMTP handling.
276 Extend the restriction to also cover ARC-signing.
278 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
279 in defer=pass mode supply a 450 to the initiator. Previously the message
282 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
283 tls_require_ciphers is used as before.
285 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
286 "pass_unscanned". Only tmpfails from the scanner are written to
287 the paniclog, as they may require admin intervention (permission
288 denied, license issues). Other scanner errors (like decompression
289 bombs) do not cause a paniclog entry.
291 JH/36 Fix reinitialisation of DKIM logging variable between messages.
292 Previously it was possible to log spurious information in receive log
295 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
296 triggered odd behaviour from Outlook Express clients.
298 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
301 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
302 since the IETF WG has not yet settled on that versus the original
303 "bare" representation.
305 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
306 Previously the millisecond value corrupted the output.
307 Fix also for syslog_pid=no and log_selector +pid, for which the pid
308 corrupted the output.
314 JH/01 Rework error string handling in TLS interface so that the caller in
315 more cases is responsible for logging. This permits library-sourced
316 string to be attached to addresses during delivery, and collapses
317 pairs of long lines into single ones.
319 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
320 during configuration. Wildcards are allowed and expanded.
322 JH/02 Rework error string handling in DKIM to pass more info back to callers.
323 This permits better logging.
325 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
326 do not close it down and have the child transport start it up again on
327 the passed-on TCP connection. Instead, proxy the child (and any
328 subsequent ones) for TLS via a unix-domain socket channel. Logging is
329 affected: the continued delivery log lines do not have any DNSSEC, TLS
330 Certificate or OCSP information. TLS cipher information is still logged.
332 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
333 identical IP addresses on different listening ports. Will also affect
336 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
337 add noisy ifdef guards to special-case this sillyness.
338 Patch from Bernd Kuhls.
340 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
341 than 255 are no longer allowed.
343 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
344 Disable the session-cache too, which might reduce our load. Since we
345 currrectly use a new context for every connection, both as server and
346 client, there is no benefit for these.
347 GnuTLS appears to not support tickets server-side by default (we don't
348 call gnutls_session_ticket_enable_server()) but client side is enabled
349 by default on recent versions (3.1.3 +) unless the PFS priority string
352 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
353 <https://reproducible-builds.org/specs/source-date-epoch/>.
355 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
356 the check for any unsuccessful recipients did not notice the limit, and
357 erroneously found still-pending ones.
359 JH/08 Pipeline CHUNKING command and data together, on kernels that support
360 MSG_MORE. Only in-clear (not on TLS connections).
362 JH/09 Avoid using a temporary file during transport using dkim. Unless a
363 transport-filter is involved we can buffer the headers in memory for
364 creating the signature, and read the spool data file once for the
365 signature and again for transmission.
367 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
368 4.77 as the kernel support then wasn't solid, having issues in 64bit
369 mode. Now, it's been long enough. Add support for FreeBSD also.
371 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
372 case where the routing stage had gathered several addresses to send to
373 a host before calling the transport for the first, we previously failed
374 to close down TLS in the old transport process before passing the TCP
375 connection to the new process. The new one sent a STARTTLS command
376 which naturally failed, giving a failed delivery and bloating the retry
377 database. Investigation and fix prototype from Wolfgang Breyha.
379 JH/12 Fix check on SMTP command input synchronisation. Previously there were
380 false-negatives in the check that the sender had not preempted a response
381 or prompt from Exim (running as a server), due to that code's lack of
382 awareness of the SMTP input buffering.
384 PP/04 Add commandline_checks_require_admin option.
385 Exim drops privileges sanely, various checks such as -be aren't a
386 security problem, as long as you trust local users with access to their
387 own account. When invoked by services which pass untrusted data to
388 Exim, this might be an issue. Set this option in main configuration
389 AND make fixes to the calling application, such as using `--` to stop
392 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
393 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
394 responses to those, into a single TLS record each way (this usually means
395 a single packet). As a side issue, smtp_enforce_sync now works on TLS
398 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
399 affects you only if you're dancing at the edge of the param size limits.
400 If you are, and this message makes sense to you, then: raise the
401 configured limit or use OpenSSL 1.1. Nothing we can do for older
404 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
405 cmdline element to get the documented default one. Previously it was
408 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
409 are unsuable for non-message-specific reasons - eg. connection timeout,
410 banner-time rejection.
412 JH/16 Fix logging of delivery remote port, when specified by router, under
415 PP/06 Repair manualroute's ability to take options in any order, even if one
416 is the name of a transport.
419 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
421 JH/17 Change the list-building routines interface to use the expanding-string
422 triplet model, for better allocation and copying behaviour.
424 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
425 Previously it was constructed the first time a possibly-matching string
426 was met in the configuration file input during startup; now it is done
429 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
430 compatible one, to avoid the (poorly documented) possibility of a config
431 file in the working directory redirecting the DB files, possibly correpting
432 some existing file. CVE-2017-10140 assigned for BDB.
434 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
435 cache-hot. Previously, although the result was properly cached, the
436 initial verify call returned a defer.
438 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
439 the main verify for receipient in uncached-mode.
441 JH/22 Retire historical build files to an "unsupported" subdir. These are
442 defined as "ones for which we have no current evidence of testing".
444 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
445 if present. Previously it was ignored.
447 JH/24 Start using specified-initialisers in C structure init coding. This is
448 a C99 feature (it's 2017, so now considered safe).
450 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
451 if was a fixed-sized field and bitmask ops via macros; it is now more
454 PP/07 GitHub PR 56: Apply MariaDB build fix.
455 Patch provided by Jaroslav Škarvada.
457 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
458 during Coverity cleanups [4.87 JH/47]
459 Diagnosis and fix provided by Michael Fischer v. Mollard.
461 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
462 the right size to place the terminating semicolon on its own folded
463 line, the header hash was calculated to an incorrect value thanks to
464 the (relaxed) space the fold became.
466 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
467 and confused the parent.
469 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
470 which could crash as a result. This could lead to undeliverable messages.
472 JH/28 Logging: "next input sent too soon" now shows where input was truncated
475 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
476 matters on fast-turnover and PID-randomising systems, which were getting
477 out-of-order delivery.
479 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
480 a possibly-overlapping copy. The symptom was that "Remote host closed
481 connection in response to HELO" was logged instead of the actual 4xx
484 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
485 Previously only that bufferd was discarded, resulting in SYMTP command
488 JH/32 DKIM: when a message has multiple signatures matching an identity given
489 in dkim_verify_signers, run the dkim acl once for each. Previously only
490 one run was done. Bug 2189.
492 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
493 "panic the current process" to "deliberately defer". The panic log is
494 still written with the problem list name; the mail and reject logs now
495 get a temp-reject line for the message that was being handled, saying
496 something like "domains check lookup or other defer". The SMTP 451
497 message is still "Temporary local problem".
499 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
500 A crafted sequence of BDAT commands could result in in-use memory beeing
501 freed. CVE-2017-16943.
503 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
504 from SMTP input. Previously it was always done; now only done for DATA
505 and not BDAT commands. CVE-2017-16944.
507 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
508 to the message (such as an overlong header line). Previously this was
509 not done and we did not exit BDAT mode. Followon from the previous item
510 though a different problem.
516 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
517 than -2003 did; needs libidn2 in addition to libidn.
519 JH/02 The path option on a pipe transport is now expanded before use.
521 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
522 Patch provided by "Björn", documentation fix added too.
524 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
525 missing a wire-to-host endian conversion.
527 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
528 close after a BDAT command line could be taken as a following command,
529 giving a synch failure. Fix by only checking for synch immediately
530 before acknowledging the chunk.
532 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
533 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
534 macro. Patches provided by Josh Soref.
536 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
537 Previously we did not; the RFC seems ambiguous and VRFY is not listed
538 by IANA as a service extension. However, John Klensin suggests that we
541 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
542 the dkim code may be unix-mode line endings rather than smtp wire-format
543 CRLF, so prepend a CR to any bare LF.
545 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
546 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
548 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
549 the first were themselves being wrongly included in the feed into dkim
550 processing; with most chunk sizes in use this resulted in an incorrect
551 body hash calculated value.
553 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
554 DKIM signature block, for verification. Although advised against by
555 standards it is specifically not ruled illegal.
557 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
559 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
560 missing a body hash (the bh= tag).
562 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
563 It seems that HAProxy sends the Proxy Protocol information in clear and
564 only then does a TLS startup, so do the same.
566 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
567 TCP connections (such as for Spamd) unless the daemon successfully set
568 Fast Open mode on its listening sockets. This fixes breakage seen on
569 too-old kernels or those not configured for Fast Open, at the cost of
570 requiring both directions being enabled for TFO, and TFO never being used
571 by non-daemon-related Exim processes.
573 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
574 endings, at least on the first header line. Try to canonify any that get
575 past that check, despite the cost.
577 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
578 now limited to an arbitrary five deep, while parsing addresses with the
579 strip_excess_angle_brackets option enabled.
581 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
582 instead leave the unprompted TLS handshake in socket buffer for the
583 TLS library to consume.
585 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
587 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
589 JH/16 Drop variables when they go out of scope. Memory management drops a whole
590 region in one operation, for speed, and this leaves assigned pointers
591 dangling. Add checks run only under the testsuite which checks all
592 variables at a store-reset and panics on a dangling pointer; add code
593 explicitly nulling out all the variables discovered. Fixes one known
594 bug: a transport crash, where a dangling pointer for $sending_ip_address
595 originally assigned in a verify callout, is re-used.
597 PP/06 Drop '.' from @INC in various Perl scripts.
599 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
601 PP/08 Reduce a number of compilation warnings under clang; building with
602 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
603 should be warning-free.
605 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
607 HS/01 Fix portability problems introduced by PP/08 for platforms where
608 realloc(NULL) is not equivalent to malloc() [SunOS et al].
610 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
611 chunk. This allows us to accept broken chunked messages. We need a more
612 general solution here.
614 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
615 already-broken messages in the queue.
617 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
619 JH/19 Fix reference counting bug in routing-generated-address tracking.
625 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
626 supports it and a size is available (ie. the sending peer gave us one).
628 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
629 years of being deprecated). The replacements are the ACLs
630 acl_smtp_mime and acl_not_smtp_mime.
632 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
633 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
634 or even an in-clear connection were permitted. Now, if the host lookup
635 was dnssec and dane was requested then the host is only used if the
636 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
637 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
638 if one fails this test.
639 This means that a poorly-configured remote DNS will make it incommunicado;
640 but it protects against a DNS-interception attack on it.
642 JH/04 Bug 1810: make continued-use of an open smtp transport connection
643 non-noisy when a race steals the message being considered.
645 JH/05 If main configuration option tls_certificate is unset, generate a
646 self-signed certificate for inbound TLS connections.
648 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
649 in rewrites and routers.
651 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
652 and logged a warning sing 4.83; now they are a configuration file error.
654 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
655 (lacking @domain). Apply the same qualification processing as RCPT.
657 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
659 JH/10 Support ${sha256:} applied to a string (as well as the previous
662 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
663 a cutthrough deliver is pending, as we always want to make a connection.
664 This also avoids re-routing the message when later placing the cutthrough
665 connection after a verify cache hit.
666 Do not update it with the verify result either.
668 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
669 when routing results in more than one destination address.
671 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
672 signing (which inhibits the cutthrough capability). Previously only
673 the presence of an option was tested; now an expansion evaluating as
674 empty is permissible (obviously it should depend only on data available
675 when the cutthrough connection is made).
677 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
678 the relevant preceding SMTP command did not note the pipelining mode.
680 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
681 Previously they were not counted.
683 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
684 as one having no matching records. Previously we deferred the message
685 that needed the lookup.
687 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
688 distinguished as "(=".
690 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
691 for missing MX records. Previously it only worked for missing A records.
693 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
695 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
696 after the data-go-ahead and data-ack. Patch from Jason Betts.
698 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
699 even for a "none" policy. Patch from Tony Meyer.
701 JH/22 Fix continued use of a connection for further deliveries. If a port was
702 specified by a router, it must also match for the delivery to be
705 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
706 When one of the recipients of a message was unsuitable for the connection
707 (has no matching addresses), we lost track of needing to mark it
708 deferred. As a result mail would be lost.
710 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
712 JH/25 Decoding ACL controls is now done using a binary search; the source code
713 takes up less space and should be simpler to maintain. Merge the ACL
714 condition decode tables also, with similar effect.
716 JH/26 Fix problem with one_time used on a redirect router which returned the
717 parent address unchanged. A retry would see the parent address marked as
718 delivered, so not attempt the (identical) child. As a result mail would
721 JH/27 Fix a possible security hole, wherein a process operating with the Exim
722 UID can gain a root shell. Credit to http://www.halfdog.net/ for
723 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
726 JH/28 Enable {spool,log} filesystem space and inode checks as default.
727 Main config options check_{log,spool}_{inodes,space} are now
728 100 inodes, 10MB unless set otherwise in the configuration.
730 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
731 Previously it only applied to the main-section connection policy
734 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
736 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
737 by me. Added RFC7919 DH primes as an alternative.
739 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
740 are not in the system include path.
742 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
743 GnuTLS, when a session startup failed (eg because the client disconnected)
744 Exim did stdio operations after fclose. This was exposed by a recent
745 change which nulled out the file handle after the fclose.
747 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
748 signed directly by the cert-signing cert, rather than an intermediate
749 OCSP-signing cert. This is the model used by LetsEncrypt.
751 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
753 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
754 an incoming connection.
756 HS/02 Bug 1802: Do not half-close the connection after sending a request
759 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
760 fallback to "prime256v1".
762 JH/34 SECURITY: Use proper copy of DATA command in error message.
763 Could leak key material. Remotely exploitable. CVE-2016-9963.
769 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
770 and 3.4.4 - once the server is enabled to respond to an OCSP request
771 it does even when not requested, resulting in a stapling non-aware
772 client dropping the TLS connection.
774 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
775 support variable-length bit vectors. No functional change.
777 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
778 The I= interface field on outgoing lines is now after the H= remote
779 host field, same as incoming lines. There is a separate
780 outgoing_interface log selector which allows you to disable the
783 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
784 If not running log_selector +smtp_connection the mainlog would be held
785 open indefinitely after a "too many connections" event, including to a
786 deleted file after a log rotate. Leave the per net connection logging
787 leaving it open for efficiency as that will be quickly detected by the
788 check on the next write.
790 HS/01 Bug 1671: Fix post transport crash.
791 Processing the wait-<transport> messages could crash the delivery
792 process if the message IDs didn't exist for some reason. When
793 using 'split_spool_directory=yes' the construction of the spool
794 file name failed already, exposing the same netto behaviour.
796 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
797 mime_regex ACL conditions.
799 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
800 to DSN fail messages (bounces): remote IP, remote greeting, remote response
801 to HELO, local diagnostic string.
803 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
804 log line to debug. Even when configured with a tls authenticator many
805 client connections are expected to not authenticate in this way, so
806 an authenticate fail is not an error.
808 HS/02 Add the Exim version string to the process info. This way exiwhat
809 gives some more detail about the running daemon.
811 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
812 matter for fast-change records such as DNSBLs.
814 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
815 retry record, even if constant. There may be multiple transports with
816 different interface settings and the retry behaviour needs to be kept
819 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
821 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
823 JH/10 Bug 840: fix log_defer_output option of pipe transport
825 JH/11 Bug 830: use same host for all RCPTS of a message, even under
826 hosts_randomize. This matters a lot when combined with mua_wrapper.
828 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
829 ${quote_pgsql:<string>} operator.
831 JH/13 Bug 1708: avoid misaligned access in cached lookup.
833 JH/14 Change header file name for freeradius-client. Relevant if compiling
834 with Radius support; from the Gentoo tree and checked under Fedora.
836 JH/15 Bug 1712: Introduce $prdr_requested flag variable
838 JH/16 Bug 1714: Permit an empty string as expansion result for transport
839 option transport_filter, meaning no filtering.
841 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
843 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
844 defaults to "*" (all hosts). The variable is now available when not built
845 with TLS, default unset, mainly to enable keeping the testsuite sane.
846 If a server certificate is not supplied (via tls_certificate) an error is
847 logged, and clients will find TLS connections fail on startup. Presumably
848 they will retry in-clear.
849 Packagers of Exim are strongly encouraged to create a server certificate
850 at installation time.
852 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
853 with the $config_file variable.
855 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
856 in transport context, after the attempt, and per-recipient. The latter type
857 is per host attempted. The event data is the error message, and the errno
858 information encodes the lookup type (A vs. MX) used for the (first) host,
859 and the trailing two digits of the smtp 4xx response.
861 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
862 to write to mainlog (or rejectlog, paniclog) in the window between file
863 creation and permissions/ownership being changed. Particularly affects
864 installations where exicyclog is run as root, rather than exim user;
865 result is that the running daemon panics and dies.
867 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
869 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
870 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
871 "pri" and "weight". Note that the previous implicit priority given by the
872 list order is no longer honoured.
874 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
877 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
878 by defining SUPPORT_SOCKS.
880 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
881 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
882 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
883 variables are renamed to proxy_{local,external}_{address,port}.
885 JH/27 Move Internationalisation support from Experimental to mainline, enabled
886 for a build by defining SUPPORT_I18N
888 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
889 of the query string, and make ${quote_redis:} do that quoting.
891 JH/29 Move Events support from Experimental to mainline, enabled by default
892 and removable for a build by defining DISABLE_EVENT.
894 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
896 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
897 cached by the daemon.
899 JH/32 Move Redis support from Experimental to mainline, enabled for a build
900 by defining LOOKUP_REDIS. The libhiredis library is required.
902 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
903 keys are given for lookup.
905 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
906 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
907 only supported when built with TLS support. The PolarSSL SHA routines
908 are still used when the TLS library is too old for convenient support.
910 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
911 openssl_options), for security. OpenSSL forces this from version 1.1.0
912 server-side so match that on older versions.
914 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
915 allocation for $value could be released as the expansion processing
916 concluded, but leaving the global pointer active for it.
918 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
919 and to use the domains and local_parts ACL conditions.
921 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
922 incorrectly not doubled on cutthrough transmission, hence seen as a
923 body-termination at the receiving system - resulting in truncated mails.
924 Commonly the sender saw a TCP-level error, and retransmitted the message
925 via the normal store-and-forward channel. This could result in duplicates
926 received - but deduplicating mailstores were liable to retain only the
927 initial truncated version.
929 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
931 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
933 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
934 we're in there, support oversigning also; bug 1309.
936 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
938 HS/04 Add support for keep_environment and add_environment options.
940 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
941 either intentional arithmetic overflow during PRNG, or testing config-
944 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
945 delivery resulted in actual delivery. Cancel cutthrough before DATA
948 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
949 rejects a recipient: pass the reject to the originator.
951 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
952 Many were false-positives and ignorable, but it's worth fixing the
955 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
956 for the new environment-manipulation done at startup. Move the routines
957 from being local to tls.c to being global via the os.c file.
959 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
960 an extract embedded as result-arg for a map, the first arg for extract
961 is unavailable so we cannot tell if this is a numbered or keyed
962 extraction. Accept either.
968 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
971 JH/02 The smtp transport option "multi_domain" is now expanded.
973 JH/03 The smtp transport now requests PRDR by default, if the server offers
976 JH/04 Certificate name checking on server certificates, when exim is a client,
977 is now done by default. The transport option tls_verify_cert_hostnames
978 can be used to disable this per-host. The build option
979 EXPERIMENTAL_CERTNAMES is withdrawn.
981 JH/05 The value of the tls_verify_certificates smtp transport and main options
982 default to the word "system" to access the system default CA bundle.
983 For GnuTLS, only version 3.0.20 or later.
985 JH/06 Verification of the server certificate for a TLS connection is now tried
986 (but not required) by default. The verification status is now logged by
987 default, for both outbound TLS and client-certificate supplying inbound
990 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
993 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
994 Status Notification (bounce) messages are now MIME format per RFC 3464.
995 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
996 under the control of the dsn_advertise_hosts option, and routers may
997 have a dsn_lasthop option.
999 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1000 default, modifiable by a malware= option. The list separator for
1001 the options can now be changed in the usual way. Bug 68.
1003 JH/10 The smtp_receive_timeout main option is now expanded before use.
1005 JH/11 The incoming_interface log option now also enables logging of the
1006 local interface on delivery outgoing connections.
1008 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1009 if the interface and destination host and port all match.
1011 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1014 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1015 Patch from Andrew Lewis.
1017 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1018 now supports optional time-restrictions, weighting, and priority
1019 modifiers per server. Patch originally by <rommer@active.by>.
1021 JH/16 The spamd_address main option now supports a mixed list of local
1022 and remote servers. Remote servers can be IPv6 addresses, and
1023 specify a port-range.
1025 JH/17 Bug 68: The spamd_address main option now supports an optional
1026 timeout value per server.
1028 JH/18 Bug 1581: Router and transport options headers_add/remove can
1029 now have the list separator specified.
1031 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1034 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1037 JH/21 Support for the A6 type of dns record is withdrawn.
1039 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1040 rather than the verbs used.
1042 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1043 from 255 to 1024 chars.
1045 JH/24 Verification callouts now attempt to use TLS by default.
1047 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1048 are generic router options now. The defaults didn't change.
1050 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1051 Original patch from Alexander Shikoff, worked over by JH.
1053 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1054 files if ZCAT_COMMAND is not executable.
1056 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1058 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1060 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1061 Normally benign, it bites when the pair was led to by a CNAME;
1062 modern usage is to not canonicalize the domain to a CNAME target
1063 (and we were inconsistent anyway for A-only vs AAAA+A).
1065 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1067 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1068 when evaluating $sender_host_dnssec.
1070 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1071 $sender_helo_dnssec variable.
1073 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1075 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1077 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1079 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1080 documented as working, but never had. Support all but $spam_report.
1082 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1083 added for tls authenticator.
1085 HS/03 Add perl_taintmode main config option
1091 TL/01 When running the test suite, the README says that variables such as
1092 no_msglog_check are global and can be placed anywhere in a specific
1093 test's script, however it was observed that placement needed to be near
1094 the beginning for it to behave that way. Changed the runtest perl
1095 script to read through the entire script once to detect and set these
1096 variables, reset to the beginning of the script, and then run through
1097 the script parsing/test process like normal.
1099 TL/02 The BSD's have an arc4random API. One of the functions to induce
1100 adding randomness was arc4random_stir(), but it has been removed in
1101 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1102 function when detected.
1104 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1105 cause callback expansion.
1107 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1108 syntax errors in an expansion can be treated as a string instead of
1109 logging or causing an error, due to the internal use of bool_lax
1110 instead of bool when processing it.
1112 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1113 server certificates when making smtp deliveries.
1115 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1117 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1119 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1121 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1122 Merged patch from Sebastian Wiedenroth.
1124 JH/05 Fix results-pipe from transport process. Several recipients, combined
1125 with certificate use, exposed issues where response data items split
1126 over buffer boundaries were not parsed properly. This eventually
1127 resulted in duplicates being sent. This issue only became common enough
1128 to notice due to the introduction of connection certificate information,
1129 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1131 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1132 size buffer was used, resulting in syntax errors when an expansion
1135 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1136 version 3.3.6 or later.
1138 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1139 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1140 both become "event_action", the variables become $event_name, $event_data
1141 and $event_defer_errno. There is a new variable $verify_mode, usable in
1142 routers, transports and related events. The tls:cert event is now also
1143 raised for inbound connections, if the main configuration event_action
1146 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1147 early OCSP support, but no stapling (appears to be less than 1.0.0).
1149 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1150 server certificate names available under the smtp transport option
1151 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1154 JH/10 Time-related extraction expansions from certificates now use the main
1155 option "timezone" setting for output formatting, and are consistent
1156 between OpenSSL and GnuTLS compilations. Bug 1541.
1158 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1159 encoded parameter in the incoming message. Bug 1558.
1161 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1162 include certificate info, eximon was claiming there were spoolfile
1165 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1167 JH/14 Log delivery-related information more consistently, using the sequence
1168 "H=<name> [<ip>]" wherever possible.
1170 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1171 are problematic for Debian distribution, omit them from the release
1174 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1176 JH/16 Fix string representation of time values on 64bit time_t architectures.
1179 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1180 output list separator was used.
1185 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1186 checkers that were complaining about end of non-void function with no
1189 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1190 This was a regression introduced in 4.83 by another bugfix.
1192 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1194 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1195 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1201 TF/01 Correctly close the server side of TLS when forking for delivery.
1203 When a message was received over SMTP with TLS, Exim failed to clear up
1204 the incoming connection properly after forking off the child process to
1205 deliver the message. In some situations the subsequent outgoing
1206 delivery connection happened to have the same fd number as the incoming
1207 connection previously had. Exim would try to use TLS and fail, logging
1208 a "Bad file descriptor" error.
1210 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1211 utilities have not been installed.
1213 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1214 temporary space as the ACL may create new global variables.
1216 TL/01 LDAP support uses per connection or global context settings, depending
1217 upon the detected version of the libraries at build time.
1219 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1220 to extract and use the src ip:port in logging and expansions as if it
1221 were a direct connection from the outside internet. PPv2 support was
1222 updated based on HAProxy spec change in May 2014.
1224 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1226 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1227 Properly escape header and check for NULL return.
1229 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1232 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1234 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1235 characters in header names, implemented as a verify condition.
1236 Contributed by Michael Fischer v. Mollard.
1238 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1239 results permerror and temperror. Previous values are deprecated but
1240 still accepted. In a future release, err_perm and err_temp will be
1241 completely removed, which will be a backward incompatibility if the
1242 ACL tests for either of these two old results. Patch contributed by
1243 user bes-internal on the mailing list.
1245 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1247 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1248 selectors, in both main and reject logs.
1250 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1253 JH/07 Add malware type "sock" for talking to simple daemon.
1255 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1257 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1258 routers/transports under cutthrough routing.
1260 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1261 numbers. Touch up "bool" conditional to keep the same definition.
1263 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1265 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1267 JH/12 Expand items in router/transport headers_add or headers_remove lists
1268 individually rather than the list as a whole. Bug 1452.
1270 Required for reasonable handling of multiple headers_ options when
1271 they may be empty; requires that headers_remove items with embedded
1272 colons must have them doubled (or the list-separator changed).
1274 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1275 view the policy declared in the DMARC record. Currently, $dmarc_status
1276 is a combined value of both the record presence and the result of the
1279 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1281 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1282 dnslookup router and the smtp transport (applying to the forward
1285 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1286 of ldap servers used for a specific lookup. Patch provided by Heiko
1289 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1290 New variable $lookup_dnssec_authenticated for observability.
1292 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1293 Patch submitted by Lars Timman.
1295 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1297 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1298 Requires trusted mode and valid format message id, aborts otherwise.
1299 Patch contributed by Heiko Schlichting.
1301 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1302 certextract with support for various fields. Bug 1358.
1304 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1305 is requested by default, modifiable by smtp transport option
1308 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1309 operate on certificate variables to give certificate fingerprints
1310 Also new ${sha256:cert_variable}.
1312 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1314 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1315 Christian Aistleitner.
1317 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1319 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1320 file. Patch from Wolfgang Breyha.
1322 JH/25 Expand the coverage of the delivery $host and $host_address to
1323 client authenticators run in verify callout. Bug 1476.
1325 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1326 align with daemon_smtp_ports. Bug 72.
1328 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1329 support and error reporting did not work properly.
1331 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1332 and is readable. Patch from Andrew Colin Kissa.
1334 TL/14 Enhance documentation of ${run expansion and how it parses the
1335 commandline after expansion, particularly in the case when an
1336 unquoted variable expansion results in an empty value.
1338 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1340 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1341 by renaming away from it. Addresses GH issue 15, reported by
1344 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1346 TL/15 SECURITY: prevent double expansion in math comparison functions
1347 (can expand unsanitized data). Not remotely exploitable.
1354 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1356 PP/02 Make -n do something, by making it not do something.
1357 When combined with -bP, the name of an option is not output.
1359 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1362 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1363 $sender_host_name and config options to manage this, and basic check
1366 PP/05 DSCP support for outbound connections and control modifier for inbound.
1368 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1369 (Only plugin which currently uses this is kerberos4, which nobody should
1370 be using, but we should make it available and other future plugins might
1371 conceivably use it, even though it would break NAT; stuff *should* be
1372 using channel bindings instead).
1374 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1375 name; added for Sendmail compatibility; requires admin caller.
1376 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1377 just ignore it); requires trusted caller.
1378 Also parse but ignore: -Ac -Am -X<logfile>
1381 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1383 TL/02 Add +smtp_confirmation as a default logging option.
1385 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1386 Patch by Magnus Holmgren from 2007-02-20.
1388 TL/04 Bugzilla 1281 - Spec typo.
1389 Bugzilla 1283 - Spec typo.
1390 Bugzilla 1290 - Spec grammar fixes.
1392 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1394 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1396 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1397 mailing list by by Dmitry Isaikin.
1399 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1401 JH/02 Support "G" suffix to numbers in ${if comparisons.
1403 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1405 NM/01 Bugzilla 1197 - Spec typo
1406 Bugzilla 1196 - Spec examples corrections
1408 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1410 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1411 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1414 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1415 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1417 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1418 "acl {{name}{arg}...}", and optional args on acl condition
1421 JH/05 Permit multiple router/transport headers_add/remove lines.
1423 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1425 JH/07 Avoid using a waiting database for a single-message-only transport.
1426 Performance patch from Paul Fisher. Bugzilla 1262.
1428 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1431 JH/09 Add $headers_added variable, with content from use of ACL modifier
1432 add_header (but not yet added to the message). Bugzilla 199.
1434 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1435 Pulled from Bugzilla 817 by Wolfgang Breyha.
1437 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1439 (nb: this is the same fix as in Exim 4.80.1)
1441 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1444 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1447 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1449 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1450 advertise SMTP AUTH mechanism to us, instead of a generic
1451 protocol violation error. Also, make Exim more robust to bad
1452 data from the Dovecot auth socket.
1454 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1456 When a queue runner is handling a message, Exim first routes the
1457 recipient addresses, during which it prunes them based on the retry
1458 hints database. After that it attempts to deliver the message to
1459 any remaining recipients. It then updates the hints database using
1462 So if a recipient address works intermittently, it can get repeatedly
1463 deferred at routing time. The retry hints record remains fresh so the
1464 address never reaches the final cutoff time.
1466 This is a fairly common occurrence when a user is bumping up against
1467 their storage quota. Exim had some logic in its local delivery code
1468 to deal with this. However it did not apply to per-recipient defers
1469 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1471 This change adds a proper retry rule check during routing so that the
1472 final cutoff time is checked against the message's age. We only do
1473 this check if there is an address retry record and there is not a
1474 domain retry record; this implies that previous attempts to handle
1475 the address had the retry_use_local_parts option turned on. We use
1476 this as an approximation for the destination being like a local
1477 delivery, as in LMTP.
1479 I suspect this new check makes the old local delivery cutoff check
1480 redundant, but I have not verified this so I left the code in place.
1482 TF/02 Correct gecos expansion when From: is a prefix of the username.
1484 Test 0254 submits a message to Exim with the header
1488 When I ran the test suite under the user fanf2, Exim expanded
1489 the header to contain my full name, whereas it should have added
1490 a Resent-Sender: header. It erroneously treats any prefix of the
1491 username as equal to the username.
1493 This change corrects that bug.
1495 GF/01 DCC debug and logging tidyup
1496 Error conditions log to paniclog rather than rejectlog.
1497 Debug lines prefixed by "DCC: " to remove any ambiguity.
1499 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1501 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1502 Bug spotted by Jeremy Harris; was flawed since initial commit.
1503 Would have resulted in OCSP responses post-SNI triggering an Exim
1504 NULL dereference and crash.
1506 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1508 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1509 Bug detection, analysis and fix by Samuel Thibault.
1510 Bugzilla 1331, Debian bug #698092.
1512 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1514 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1515 Server implementation by Todd Lyons, client by JH.
1516 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1517 config variable "prdr_enable" controls whether the server
1518 advertises the facility. If the client requests PRDR a new
1519 acl_data_smtp_prdr ACL is called once for each recipient, after
1520 the body content is received and before the acl_smtp_data ACL.
1521 The client is controlled by both of: a hosts_try_prdr option
1522 on the smtp transport, and the server advertisement.
1523 Default client logging of deliveries and rejections involving
1524 PRDR are flagged with the string "PRDR".
1526 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1527 fclose(). Diagnosis by Todd Lyons.
1529 PP/17 Update configure.default to handle IPv6 localhost better.
1530 Patch by Alain Williams (plus minor tweaks).
1533 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1534 This is now consistent with GnuTLS, and is now documented: the
1535 previous undocumented portable approach to treating the option as
1536 unset was to force an expansion failure. That still works, and
1537 an empty string is now equivalent.
1539 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1540 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1541 not performing validation itself.
1543 PP/20 Added force_command boolean option to pipe transport.
1544 Patch from Nick Koston, of cPanel Inc.
1546 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1549 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1551 PP/21 Fix eximon continuous updating with timestamped log-files.
1552 Broken in a format-string cleanup in 4.80, missed when I repaired the
1553 other false fix of the same issue.
1554 Report and fix from Heiko Schlichting.
1557 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1558 Report from Prashanth Katuri.
1560 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1561 It's SecureTransport, so affects any MacOS clients which use the
1562 system-integrated TLS libraries, including email clients.
1564 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1565 using a MIME ACL for non-SMTP local injection.
1566 Report and assistance in diagnosis by Warren Baker.
1568 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1570 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1572 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1573 last id that failed so it may be referenced in subsequent ACL's.
1575 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1578 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1579 ldap library initialization, allowing self-signed CA's to be
1580 used. Also properly sets require_cert option later in code by
1581 using NULL (global ldap config) instead of ldap handle (per
1582 session). Bug diagnosis and testing by alxgomz.
1584 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1585 the src/util/ subdirectory.
1587 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1588 renamed to Transport Post Delivery Action by Jeremy Harris, as
1591 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1592 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1593 redis_servers = needs to be configured which will be used by the redis
1594 lookup. Patch from Warren Baker, of The Packet Hub.
1596 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1598 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1599 hostname or reverse DNS when processing a host list. Used suggestions
1600 from multiple comments on this bug.
1602 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1604 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1605 Missed a few lines, added it to make the runtest require no keyboard
1608 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1609 contains upper case chars. Make router use caseful_local_part.
1611 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1612 support when GnuTLS has been built with p11-kit.
1618 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1620 This, or similar/improved, will also be change PP/11 of 4.82.
1626 PP/01 Handle short writes when writing local log-files.
1627 In practice, only affects FreeBSD (8 onwards).
1628 Bugzilla 1053, with thanks to Dmitry Isaikin.
1630 NM/01 Bugzilla 949 - Documentation tweak
1632 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1635 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1637 PP/02 Implemented gsasl authenticator.
1639 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1641 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1642 `pkg-config foo` for cflags/libs.
1644 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1645 with rest of GSASL and with heimdal_gssapi.
1647 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1648 `pkg-config foo` for cflags/libs for the TLS implementation.
1650 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1651 properties get this fed in as external SSF. A number of robustness
1652 and debugging improvements to the cyrus_sasl authenticator.
1654 PP/08 cyrus_sasl server now expands the server_realm option.
1656 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1657 Patch by Jeremy Harris.
1659 PP/10 Added dbmjz lookup type.
1661 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1663 PP/12 MAIL args handles TAB as well as SP, for better interop with
1664 non-compliant senders.
1665 Analysis and variant patch by Todd Lyons.
1667 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1668 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1669 Patch from Dirk Mueller <dmueller@suse.com>
1671 PP/13 tls_peerdn now print-escaped for spool files.
1672 Observed some $tls_peerdn in wild which contained \n, which resulted
1673 in spool file corruption.
1675 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1676 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1677 or write after TLS renegotiation, which otherwise led to messages
1680 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1681 as a tracking header (ie: a signed header comes before the signature).
1682 Patch from Wolfgang Breyha.
1684 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1685 comma-sep list; embedded commas doubled.
1687 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1689 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1691 Report and patch from Dmitry Banschikov.
1693 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1694 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1695 failures. We appear to now support TLS1.1+ with Exim.
1697 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1698 lets Exim select keys and certificates based upon TLS SNI from client.
1699 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1700 before an outbound SMTP session. New log_selector, +tls_sni.
1702 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1703 NULL dereference. Report and patch from Alun Jones.
1705 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1706 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1707 Not seeing resolver debug output on NetBSD, but suspect this is a
1708 resolver implementation change.
1710 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1711 Left warnings. Added "eximon gdb" invocation mode.
1713 PP/21 Defaulting "accept_8bitmime" to true, not false.
1715 PP/22 Added -bw for inetd wait mode support.
1717 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1718 locate the relevant includes and libraries. Made this the default.
1720 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1721 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1723 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1724 This may cause build issues on older platforms.
1726 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1727 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1728 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1729 Added SNI support via GnuTLS too.
1730 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1732 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1734 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1735 Applied second patch from Janne, implementing suggestion to default
1736 multiple-strings-in-record handling to match SPF spec.
1738 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1740 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1741 read-only, out of scope).
1742 Patch from Wolfgang Breyha, report from Stuart Northfield.
1744 PP/29 Fix three issues highlighted by clang analyser static analysis.
1745 Only crash-plausible issue would require the Cambridge-specific
1746 iplookup router and a misconfiguration.
1747 Report from Marcin Mirosław.
1749 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1751 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1752 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1753 As part of this, removing so much warning spew let me fix some minor
1754 real issues in debug logging.
1756 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1757 assignment on my part. Fixed.
1759 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1760 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1761 Janne Snabb (who went above and beyond: thank you).
1763 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1764 string otherwise requires a connection and a bunch more work and it's
1765 relatively easy to get wrong. Should also expose TLS library linkage
1768 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1769 64-bit ${eval} (JH/03).
1771 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1772 GNU libc to support some of the 64-bit stuff, should not lead to
1773 conflicts. Defined before os.h is pulled in, so if a given platform
1774 needs to override this, it can.
1776 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1777 protection layer was required, which is not implemented.
1778 Bugzilla 1254, patch from Wolfgang Breyha.
1780 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1781 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1782 tls_dhparam take prime identifiers. Also unbreak combination of
1783 OpenSSL+DH_params+TLSSNI.
1785 PP/39 Disable SSLv2 by default in OpenSSL support.
1791 PP/01 Solaris build fix for Oracle's LDAP libraries.
1792 Bugzilla 1109, patch from Stephen Usher.
1794 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1796 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1799 TF/02 Fix a couple more cases where we did not log the error message
1800 when unlink() failed. See also change 4.74-TF/03.
1802 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1803 lock up or crash if it happened to be inside a call to libc when it
1804 got a SIGUSR1 from exiwhat.
1806 The SIGUSR1 handler appends the current process status to the process
1807 log which is later printed by exiwhat. It used to use the general
1808 purpose logging code to do this, but several functions it calls are
1809 not safe for signals.
1811 The new output code in the SIGUSR1 handler is specific to the process
1812 log, and simple enough that it's easy to inspect for signal safety.
1813 Removing some special cases also simplifies the general logging code.
1814 Removing the spurious timestamps from the process log simplifies
1817 TF/04 Improved ratelimit ACL condition.
1819 The /noupdate option has been deprecated in favour of /readonly which
1820 has clearer semantics. The /leaky, /strict, and /readonly update modes
1821 are mutually exclusive. The update mode is no longer included in the
1822 database key; it just determines when the database is updated. (This
1823 means that when you upgrade Exim will forget old rate measurements.)
1825 Exim now checks that the per_* options are used with an update mode that
1826 makes sense for the current ACL. For example, when Exim is processing a
1827 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1828 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1829 must specify per_mail/readonly. If you omit the update mode it defaults to
1830 /leaky where that makes sense (as before) or /readonly where required.
1832 The /noupdate option is now undocumented but still supported for
1833 backwards compatibility. It is equivalent to /readonly except that in
1834 ACLs where /readonly is required you may specify /leaky/noupdate or
1835 /strict/noupdate which are treated the same as /readonly.
1837 A useful new feature is the /count= option. This is a generalization
1838 of the per_byte option, so that you can measure the throughput of other
1839 aggregate values. For example, the per_byte option is now equivalent
1840 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1842 The per_rcpt option has been generalized using the /count= mechanism
1843 (though it's more complicated than the per_byte equivalence). When it is
1844 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1845 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1846 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1847 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1848 Note that using per_rcpt with a non-readonly update mode in more than
1849 one ACL will cause the recipients to be double-counted. (The per_mail
1850 and per_byte options don't have this problem.)
1852 The handling of very low rates has changed slightly. If the computed rate
1853 is less than the event's count (usually one) then this event is the first
1854 after a long gap. In this case the rate is set to the same as this event's
1855 count, so that the first message of a spam run is counted properly.
1857 The major new feature is a mechanism for counting the rate of unique
1858 events. The new per_addr option counts the number of different
1859 recipients that someone has sent messages to in the last time period. It
1860 behaves like per_rcpt if all the recipient addresses are different, but
1861 duplicate recipient addresses do not increase the measured rate. Like
1862 the /count= option this is a general mechanism, so the per_addr option
1863 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1864 example, measure the rate that a client uses different sender addresses
1865 with the options per_mail/unique=$sender_address. There are further
1866 details in the main documentation.
1868 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1870 TF/06 Removed a few PCRE remnants.
1872 TF/07 Automatically extract Exim's version number from tags in the git
1873 repository when doing development or release builds.
1875 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1876 Bugzilla 879. Patch from Paul Fisher.
1878 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1879 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1882 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1884 PP/05 Variable $av_failed, true if the AV scanner deferred.
1885 Bugzilla 1078. Patch from John Horne.
1887 PP/06 Stop make process more reliably on build failure.
1888 Bugzilla 1087. Patch from Heiko Schlittermann.
1890 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1891 Bugzilla 1089. Patch from Heiko Schlittermann.
1893 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1894 Bugzilla 1131. Patch from Holger Weiß.
1896 PP/09 Handle IPv6 addresses with SPF.
1897 Bugzilla 860. Patch from Wolfgang Breyha.
1899 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1901 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1904 PP/11 match_* no longer expand right-hand-side by default.
1905 New compile-time build option, EXPAND_LISTMATCH_RHS.
1906 New expansion conditions, "inlist", "inlisti".
1908 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1910 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1912 PP/14 fix log_write() format string regression from TF/03.
1913 Bugzilla 1152. Patch from Dmitry Isaikin.
1919 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1921 PP/02 Harmonised TLS library version reporting; only show if debugging.
1922 Layout now matches that introduced for other libraries in 4.74 PP/03.
1924 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1926 PP/04 New "dns_use_edns0" global option.
1928 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1931 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1932 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1934 TK/01 Updated PolarSSL code to 0.14.2.
1935 Bugzilla 1097. Patch from Andreas Metzler.
1937 PP/07 Catch divide-by-zero in ${eval:...}.
1938 Fixes bugzilla 1102.
1940 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1943 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1944 format-string attack -- SECURITY: remote arbitrary code execution.
1946 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1947 time unintentionally subject to list matching rules, letting the header
1948 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1949 arbitrary string expansion). This allowed for information disclosure.
1951 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1952 INT_MIN/-1 -- value coerced to INT_MAX.
1958 NM/01 Workaround for PCRE version dependency in version reporting
1961 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1962 This fixes portability to compilers other than gcc, notably
1963 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1965 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1966 makefiles for portability to HP-UX and POSIX correctness.
1968 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1969 Also via indented variable definition in the Makefile.
1970 (Debugging by Oliver Heesakkers).
1972 PP/02 Restore caching of spamd results with expanded spamd_address.
1973 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1975 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1976 Improves build reliability. Fix from: Frank Elsner
1978 NM/02 Fix wide character breakage in the rfc2047 coding
1979 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1981 NM/03 Allow underscore in dnslist lookups
1982 Fixes bug 1026. Patch from Graeme Fowler
1984 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1985 Code patches from Adam Ciarcinski of NetBSD.
1987 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1990 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1991 is logged, to avoid truncation. Patch from John Horne.
1993 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1994 Patch from Jakob Hirsch.
1996 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1997 SQL string expansion failure details.
1998 Patch from Andrey Oktyabrski.
2000 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2001 Patch from Simon Arlott.
2003 PP/09 New lookups functionality failed to compile on old gcc which rejects
2004 extern declarations in function scope.
2005 Patch from Oliver Fleischmann
2007 PP/10 Use sig_atomic_t for flags set from signal handlers.
2008 Check getgroups() return and improve debugging.
2009 Fixed developed for diagnosis in bug 927 (which turned out to be
2012 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2013 Patch from Mark Zealey.
2015 PP/12 Bugzilla 1056: Improved spamd server selection.
2016 Patch from Mark Zealey.
2018 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2019 Based on patch from Heiko Schlittermann.
2021 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2022 Patch from Uwe Doering, sign-off by Michael Haardt.
2024 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2025 variable declaration deep within a block. Bug and patch from
2028 PP/15 lookups-Makefile IRIX compatibility coercion.
2030 PP/16 Make DISABLE_DKIM build knob functional.
2032 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2033 Patch by Simon Arlott
2035 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2036 variable argument macros. Our copy now differs from upstream.
2042 TF/01 Failure to get a lock on a hints database can have serious
2043 consequences so log it to the panic log.
2045 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2046 controlled using the smtp_confirmation log selector.
2048 TF/03 Include the error message when we fail to unlink a spool file.
2050 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2051 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2052 for maintaining out-of-tree patches for some time.
2054 PP/01 Bugzilla 139: Documentation and portability issues.
2055 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2056 Handle per-OS dynamic-module compilation flags.
2058 PP/02 Let /dev/null have normal permissions.
2059 The 4.73 fixes were a little too stringent and complained about the
2060 permissions on /dev/null. Exempt it from some checks.
2061 Reported by Andreas M. Kirchwitz.
2063 PP/03 Report version information for many libraries, including
2064 Exim version information for dynamically loaded libraries. Created
2065 version.h, now support a version extension string for distributors
2066 who patch heavily. Dynamic module ABI change.
2068 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2069 privilege escalation vulnerability whereby the Exim run-time user
2070 can cause root to append content of the attacker's choosing to
2073 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2076 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2077 If dropping privileges for untrusted macros, we disabled normal logging
2078 on the basis that it would fail; for the Exim run-time user, this is not
2079 the case, and it resulted in successful deliveries going unlogged.
2080 Fixed. Reported by Andreas Metzler.
2086 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2087 only prepend for the Resent-* case. Fixes regression introduced in
2088 Exim 4.70 by NM/22 for Bugzilla 607.
2090 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2091 increasing numbers of administrators be bitten by this.
2093 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2095 PP/03 Bugzilla 994: added openssl_options main configuration option.
2097 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2099 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2101 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2103 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2104 without a peer certificate, leading to a segfault because of an
2105 assumption that peers always have certificates. Be a little more
2106 paranoid. Problem reported by Martin Tscholak.
2108 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2109 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2110 NB: ClamAV planning to remove STREAM in "middle of 2010".
2111 CL also introduces -bmalware, various -d+acl logging additions and
2112 more caution in buffer sizes.
2114 PP/09 Implemented reverse_ip expansion operator.
2116 PP/10 Bugzilla 937: provide a "debug" ACL control.
2118 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2120 PP/12 Bugzilla 973: Implement --version.
2122 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2124 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2126 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2128 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2129 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2130 ignore trailing whitespace.
2132 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2134 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2137 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2138 Notification from Dr Andrew Aitchison.
2140 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2141 ExtendedDetectionInfo response format.
2142 Notification from John Horne.
2144 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2147 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2148 XSL and documented dependency on system catalogs, with examples of how
2151 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2154 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2155 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2156 configuration file which is writeable by the Exim user or group.
2158 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2159 of configuration files to cover files specified with the -C option if
2160 they are going to be used with root privileges, not just the default
2163 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2164 option (effectively making it always true).
2166 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2167 files to be used while preserving root privileges.
2169 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2170 that rogue child processes cannot use them.
2172 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2173 run-time user, instead of root.
2175 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2176 Exim run-time user without dropping privileges.
2178 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2179 result string, instead of calling string_vformat() twice with the same
2182 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2183 for other users. Others should always drop root privileges if they use
2184 -C on the command line, even for a whitelisted configure file.
2186 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2188 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2194 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2195 $data_path, and $header_path variables; fixed documentation bugs and
2198 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2199 exipick to access non-standard spools, including the "frozen" queue
2202 NM/01 Bugzilla 965: Support mysql stored procedures.
2203 Patch from Alain Williams
2205 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2207 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2208 Patch from Andreas Metzler
2210 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2211 Patch from Kirill Miazine
2213 NM/05 Bugzilla 671: Added umask to procmail example.
2215 JJ/03 installed exipick 20100323.0, fixing doc bug
2217 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2218 directory. Notification and patch from Dan Rosenberg.
2220 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2222 TK/02 Improve log output when DKIM signing operation fails.
2224 MH/01 Treat the transport option dkim_domain as a colon separated
2225 list, not as a single string, and sign the message with each element,
2226 omitting multiple occurences of the same signer.
2228 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2229 Bugzilla 985, 986. Patch by Simon Arlott
2231 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2232 Patch by Simon Arlott
2234 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2235 MBX locking. Notification from Dan Rosenberg.
2241 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2243 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2245 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2247 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2249 NM/04 Bugzilla 915: Flag broken perl installation during build.
2255 TK/01 Added patch by Johannes Berg that expands the main option
2256 "spamd_address" if it starts with a dollar sign.
2258 TK/02 Write list of recipients to X-Envelope-Sender header when building
2259 the mbox-format spool file for content scanning (suggested by Jakob
2262 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2263 (http://www.dcc-servers.net/) support via dccifd. Activated by
2264 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2266 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2267 by Mark Daniel Reidel <mr@df.eu>.
2269 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2270 When building exim an external PCRE library is now needed -
2271 PCRE is a system library on the majority of modern systems.
2272 See entry on PCRE_LIBS in EDITME file.
2274 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2275 conversation. Added nologin parameter to request.
2276 Patch contributed by Kirill Miazine.
2278 TF/01 Do not log submission mode rewrites if they do not change the address.
2280 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2282 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2283 log files in place. Contributed by Roberto Lima.
2285 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2287 TF/03 Bugzilla 615: When checking the local_parts router precondition
2288 after a local_part_suffix or local_part_prefix option, Exim now
2289 does not use the address's named list lookup cache, since this
2290 contains cached lookups for the whole local part.
2292 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2293 Robert Millan. Documentation is in experimental-spec.txt.
2295 TF/04 Bugzilla 668: Fix parallel build (make -j).
2297 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2299 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2300 Patch provided by Jan Srzednicki.
2302 TF/05 Leading white space used to be stripped from $spam_report which
2303 wrecked the formatting. Now it is preserved.
2305 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2306 that they are available at delivery time.
2308 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2310 TF/08 TLS error reporting now respects the incoming_interface and
2311 incoming_port log selectors.
2313 TF/09 Produce a more useful error message if an SMTP transport's hosts
2314 setting expands to an empty string.
2316 NM/06 Bugzilla 744: EXPN did not work under TLS.
2317 Patch provided by Phil Pennock.
2319 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2320 Patch provided by Richard Godbee.
2322 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2323 acl_smtp_notquit, added index entry.
2325 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2326 Patch provided by Eugene Bujak.
2328 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2329 accept(). Patch provided by Maxim Dounin.
2331 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2332 Patch provided by Phil Pennock.
2334 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2336 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2337 Patch provided by Brad "anomie" Jorsch.
2339 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2340 Patch provided by Dean Brooks.
2342 TK/05 Add native DKIM support (does not depend on external libraries).
2344 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2345 Patch provided by Graeme Fowler.
2347 NM/16 Bugzilla 851: Documentation example syntax fix.
2349 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2351 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2354 NM/19 Bugzilla 745: TLS version reporting.
2355 Patch provided by Phil Pennock.
2357 NM/20 Bugzilla 167: bool: condition support.
2358 Patch provided by Phil Pennock.
2360 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2361 clients. Patch provided by Phil Pennock.
2363 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2364 Patch provided by Brad "anomie" Jorsch.
2366 NM/23 Bugzilla 687: Fix misparses in eximstats.
2367 Patch provided by Heiko Schlittermann.
2369 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2370 Patch provided by Heiko Schlittermann.
2372 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2373 plus update to original patch.
2375 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2377 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2378 Patch provided by David Brownlee.
2380 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2382 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2384 NM/30 Bugzilla 888: TLS documentation bugfixes.
2386 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2388 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2389 Unlike the original bugzilla I have changed all shell scripts in src tree.
2391 NM/33 Bugzilla 898: Transport filter timeout fix.
2392 Patch by Todd Rinaldo.
2394 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2395 Patch by Serge Demonchaux.
2397 NM/35 Bugzilla 39: Base64 decode bug fixes.
2398 Patch by Jakob Hirsch.
2400 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2402 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2404 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2406 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2412 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2413 ALT-N's libdkim that I have put here:
2414 http://duncanthrax.net/exim-experimental/
2416 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2417 were called 'true' and it seems that C99 defines that as a reserved
2418 keyword to be used with 'bool' variable types. That means you could
2419 not include C99-style headers which use bools without triggering
2420 build errors in sieve.c.
2422 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2423 as mailq or other aliases. Changed the --help handling significantly
2424 to do whats expected. exim_usage() emits usage/help information.
2426 SC/01 Added the -bylocaldomain option to eximstats.
2428 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2430 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2432 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2438 PH/01 Another patch from the Sieve maintainer.
2440 PH/02 When an IPv6 address is converted to a string for single-key lookup
2441 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2442 dots are used instead of colons so that keys in lsearch files need not
2443 contain colons. This was done some time before quoting was made available
2444 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2445 (notated using the quote facility) so as to distinguish them from IPv4
2446 keys. This meant that lookups for IP addresses in host lists did not work
2447 for iplsearch lookups.
2449 This has been fixed by arranging for IPv6 addresses to be expressed with
2450 colons if the lookup type is iplsearch. This is not incompatible, because
2451 previously such lookups could never work.
2453 The situation is now rather anomalous, since one *can* have colons in
2454 ordinary lsearch keys. However, making the change in all cases is
2455 incompatible and would probably break a number of configurations.
2457 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2460 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2461 conversion specification without a maximum field width, thereby enabling
2462 a rogue spamd server to cause a buffer overflow. While nobody in their
2463 right mind would setup Exim to query an untrusted spamd server, an
2464 attacker that gains access to a server running spamd could potentially
2465 exploit this vulnerability to run arbitrary code as the Exim user.
2467 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2468 $primary_hostname instead of what libspf2 thinks the hosts name is.
2470 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2471 a directory entry by the name of the lookup key. Previously, if a
2472 symlink pointed to a non-existing file or a file in a directory that
2473 Exim lacked permissions to read, a lookup for a key matching that
2474 symlink would fail. Now it is enough that a matching directory entry
2475 exists, symlink or not. (Bugzilla 503.)
2477 PH/03 The body_linecount and body_zerocount variables are now exported in the
2480 PH/04 Added the $dnslist_matched variable.
2482 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2483 This means they are set thereafter only if the connection becomes
2486 PH/06 Added the client_condition to authenticators so that some can be skipped
2487 by clients under certain conditions.
2489 PH/07 The error message for a badly-placed control=no_multiline_responses left
2490 "_responses" off the end of the name.
2492 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2494 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2495 (without spaces) instead of just copying the configuration text.
2497 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2499 PH/11 Added $max_received_linelength.
2501 PH/12 Added +ignore_defer and +include_defer to host lists.
2503 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2504 way in which PCRE > 7.0 is built.
2506 PH/14 Implemented queue_only_load_latch.
2508 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2509 MAIL command. The effect was to mangle the value on 64-bit systems.
2511 PH/16 Another patch from the Sieve maintainer.
2513 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2515 PH/18 If a system quota error occurred while trying to create the file for
2516 a maildir delivery, the message "Mailbox is full" was not appended to the
2517 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2518 only to a quota excession during the actual writing of the file.
2520 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2521 characters?) which causes problems in log lines. The DN values are now
2522 passed through string_printing() before being added to log lines.
2524 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2525 and InterBase are left for another time.)
2527 PH/21 Added message_body_newlines option.
2529 PH/22 Guard against possible overflow in moan_check_errorcopy().
2531 PH/23 POSIX allows open() to be a macro; guard against that.
2533 PH/24 If the recipient of an error message contained an @ in the local part
2534 (suitably quoted, of course), incorrect values were put in $domain and
2535 $local_part during the evaluation of errors_copy.
2541 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2542 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2545 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2546 issue a MAIL command.
2548 PH/02 In an ACL statement such as
2550 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2552 if a client was not listed at all, or was listed with a value other than
2553 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2554 the condition was not true (as it should be), so access was not denied.
2555 The bug was that the ! inversion was incorrectly passed on to the second
2556 item. This has been fixed.
2558 PH/03 Added additional dnslists conditions == and =& which are different from
2559 = and & when the dns lookup returns more than one IP address.
2561 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2562 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2564 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2565 FSYNC, which compiles an option called disable_fsync that allows for
2566 bypassing fsync(). The documentation is heavily laced with warnings.
2568 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2570 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2571 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2572 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2573 including adding "make clean"; (3) Added -fPIC when compiling the test
2574 dynamically loaded module, to get rid of a warning.
2576 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2577 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2578 = 0s. The bug is otherwise harmless.
2580 PH/07 There was a bug in the dovecot authenticator such that the value of
2581 $auth1 could be overwritten, and so not correctly preserved, after a
2582 successful authentication. This usually meant that the value preserved by
2583 the server_setid option was incorrect.
2585 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2587 PH/09 Installed PCRE release 7.0.
2589 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2590 run for batched SMTP input. It is now run at the start of every message
2591 in the batch. While fixing this I discovered that the process information
2592 (output by running exiwhat) was not always getting set for -bs and -bS
2593 input. This is fixed, and it now also says "batched" for BSMTP.
2595 PH/11 Added control=no_pipelining.
2597 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2598 patch, slightly modified), and move the expansion of helo_data till after
2599 the connection is made in the smtp transport (so it can use these
2602 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2604 PH/14 Added log_selector = +pid.
2606 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2608 PH/16 Add ${if forany and ${if forall.
2610 PH/17 Added dsn_from option to vary the From: line in DSNs.
2612 PH/18 Flush SMTP output before performing a callout, unless control =
2613 no_callout_flush is set.
2615 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2616 was true (the default) a successful delivery failed to delete the retry
2617 item, thus causing premature timeout of the address. The bug is now
2620 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2622 PH/21 Long custom messages for fakedefer and fakereject are now split up
2623 into multiline responses in the same way that messages for "deny" and
2624 other ACL rejections are.
2626 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2627 with slight modification.
2629 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2630 draft, changing the syntax and factoring some duplicate code".
2632 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2633 for deliveries of the second and subsequent messages over the same SMTP
2636 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2637 ${reduce, with only minor "tidies".
2639 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2641 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2642 expansion side effects.
2644 PH/27 When a message times out after an over-quota error from an Exim-imposed
2645 quota, the bounce message says "mailbox is full". This message was not
2646 being given when it was a system quota that was exceeded. It now should
2649 MH/03 Made $recipients available in local_scan(). local_scan() already has
2650 better access to the recipient list through recipients_list[], but
2651 $recipients can be useful in postmaster-provided expansion strings.
2653 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2654 in the case of a MAIL command with additional options following the
2655 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2656 were accidentally chopped off.
2658 PH/29 SMTP synchronization checks are implemented when a command is read -
2659 there is a check that no more input is waiting when there shouldn't be
2660 any. However, for some commands, a delay in an ACL can mean that it is
2661 some time before the response is written. In this time, more input might
2662 arrive, invalidly. So now there are extra checks after an ACL has run for
2663 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2664 pipelining has not been advertised.
2666 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2668 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2669 returned in the SMTP response when smtp_return_error_details was false.
2670 This has been fixed.
2672 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2673 instead of the C I/O that was originally supplied, because problems were
2674 reported on Solaris.
2676 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2677 Exim which did not show up earlier: it was assuming that a call to
2678 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2679 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2680 was a macro that became an assignment, so it seemed to work. This has
2681 changed to a proper function call with a void return, hence the compile
2682 error. Exim's code has been fixed.
2684 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2687 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2689 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2691 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2692 because it turns out to be too big or there is a timeout) there is no
2693 "Completed" line in the log. When some messages of this type were
2694 selected by exigrep, they were listed as "not completed". Others were
2695 picked up by some special patterns. I have improved the selection
2696 criteria to be more general.
2698 PH/38 The host_find_failed option in the manualroute router can now be set
2699 to "ignore", to completely ignore a host whose IP address cannot be
2700 found. If all hosts are ignored, the behaviour is controlled by the new
2701 host_all_ignored option.
2703 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2704 homing or because of multiple MX records with /mx) generated more than
2705 one IP address, and the following item turned out to be the local host,
2706 all the secondary addresses of the first item were incorrectly removed
2707 from the list, along with the local host and any following hosts (which
2708 is what is supposed to happen).
2710 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2711 whoever called Exim into the -H file. In the case of the daemon it was
2712 behaving confusingly. When first started, it used values for whoever
2713 started the daemon, but after a SIGHUP it used the Exim user (because it
2714 calls itself on a restart). I have changed the code so that it now always
2717 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2718 message are rejected with the same error (e.g. no authentication or bad
2719 sender address), and a DATA command is nevertheless sent (as can happen
2720 with PIPELINING or a stupid MUA), the error message that was given to the
2721 RCPT commands is included in the rejection of the DATA command. This is
2722 intended to be helpful for MUAs that show only the final error to their
2725 PH/42 Another patch from the Sieve maintainer.
2727 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2728 Eximstats - Fixed some broken HTML links and added missing column headers
2730 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2731 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2733 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2735 PH/43 Yet another patch from the Sieve maintainer.
2737 PH/44 I found a way to check for a TCP/IP connection going away before sending
2738 the response to the final '.' that terminates a message, but only in the
2739 case where the client has not sent further data following the '.'
2740 (unfortunately, this is allowed). However, in many cases there won't be
2741 any further data because there won't be any more messages to send. A call
2742 to select() can be used: if it shows that the input is "ready", there is
2743 either input waiting, or the socket has been closed. An attempt to read
2744 the next input character can distinguish the two cases. Previously, Exim
2745 would have sent an OK response which the client would never have see.
2746 This could lead to message repetition. This fix should cure that, at
2747 least in a lot of common cases.
2749 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2750 advertised in response to EHLO.
2756 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2757 fixed by 4.65/MH/01 (is this a record?) are fixed:
2759 (i) An empty string was always treated as zero by the numeric comparison
2760 operators. This behaviour has been restored.
2762 (ii) It is documented that the numeric comparison operators always treat
2763 their arguments as decimal numbers. This was broken in that numbers
2764 starting with 0 were being interpreted as octal.
2766 While fixing these problems I realized that there was another issue that
2767 hadn't been noticed. Values of message_size_limit (both the global option
2768 and the transport option) were treated as octal if they started with 0.
2769 The documentation was vague. These values are now always treated as
2770 decimal, and I will make that clear in the documentation.
2776 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2777 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2780 MH/01 Don't check that the operands of numeric comparison operators are
2781 integers when their expansion is in "skipping" mode (fixes bug
2782 introduced by 4.64-PH/07).
2784 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2785 child addresses, Exim now panics and dies. Previously, because the count
2786 is held in a short int, deliveries were likely to be lost. As such a
2787 large number of recipients for a single message is ridiculous
2788 (performance will be very, very poor), I have chosen to impose a limit
2789 rather than extend the field.
2795 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2796 leftover -K file (the existence of which was triggered by #402).
2797 While we were at it, introduced process PID as part of the -K
2798 filename. This should rule out race conditions when creating
2801 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2802 processing considerably. Previous code took too long for large mails,
2803 triggering a timeout which in turn triggers #401.
2805 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2806 in the DK code in transports.c. sendfile() is not really portable,
2807 hence the _LINUX specificness.
2809 TF/01 In the add_headers option to the mail command in an Exim filter,
2810 there was a bug that Exim would claim a syntax error in any
2811 header after the first one which had an odd number of characters
2814 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2815 callout verification, Exim cached a "reject" for the entire domain. This
2816 is correct for most verifications, but it is not correct for a recipient
2817 verification with use_sender or use_postmaster set, because in that case
2818 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2819 case of MAIL FROM:<> rejection from other early rejections (e.g.
2820 rejection of HELO). When verifying a recipient using a non-null MAIL
2821 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2822 Whatever the result of the callout, the value of the domain cache is
2823 left unchanged (for any other kind of callout, getting as far as trying
2824 RCPT means that the domain itself is ok).
2826 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2829 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2830 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2831 session, the daemon ignores ECONNECT errors and logs others; it now
2832 ignores EPIPE as well.
2834 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2835 (quoted-printable decoding).
2837 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2838 later the small subsequent patch to fix an introduced bug.
2840 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2842 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2844 PH/08 An error is now given if message_size_limit is specified negative.
2846 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2847 to be given (somewhat) arbitrary names.
2849 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2852 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2853 miscellaneous code fixes
2855 PH/10 Added the log_reject_target ACL modifier to specify where to log
2858 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2859 hostname. This is wrong, because it relates to the incoming message (and
2860 probably the interface on which it is arriving) and not to the outgoing
2861 callout (which could be using a different interface). This has been
2862 changed to use the value of the helo_data option from the smtp transport
2863 instead - this is what is used when a message is actually being sent. If
2864 there is no remote transport (possible with a router that sets up host
2865 addresses), $smtp_active_hostname is used.
2867 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2868 tweaks were necessary in order to get it to work (see also 21 below):
2869 (a) The code assumed that strncpy() returns a negative number on buffer
2870 overflow, which isn't the case. Replaced with Exim's string_format()
2872 (b) There were several signed/unsigned issues. I just did the minimum
2873 hacking in of casts. There is scope for a larger refactoring.
2874 (c) The code used strcasecmp() which is not a standard C function.
2875 Replaced with Exim's strcmpic() function.
2876 (d) The code set only $1; it now sets $auth1 as well.
2877 (e) A simple test gave the error "authentication client didn't specify
2878 service in request". It would seem that Dovecot has changed its
2879 interface. Fortunately there's a specification; I followed it and
2880 changed what the client sends and it appears to be working now.
2882 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2885 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2886 address A is aliased to B and C, where B exists and C does not. Without
2887 -v the output is "A verified" because verification stops after a
2888 successful redirection if more than one address is generated. However,
2889 with -v the child addresses are also verified. Exim was outputting "A
2890 failed to verify" and then showing the successful verification for C,
2891 with its parentage. It now outputs "B failed to verify", showing B's
2892 parentage before showing the successful verification of C.
2894 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2895 look up a TXT record in a specific list after matching in a combined
2898 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2899 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2900 they consult the DNS. I had assumed they would set it the way they
2901 wanted; and indeed my experiments on Linux seem to show that in some
2902 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2903 To be on the safe side, however, I have now made the interface to
2904 host_find_byname() similar to host_find_bydns(), with an argument
2905 containing the DNS resolver options. The host_find_byname() function now
2906 sets these options at its start, just as host_find_bydns() does. The smtp
2907 transport options dns_qualify_single and dns_search_parents are passed to
2908 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2909 of host_find_byname() use the default settings of RES_DEFNAMES
2910 (qualify_single) but not RES_DNSRCH (search_parents).
2912 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2913 spool_read_header() do less string testing, by means of a preliminary
2914 switch on the second character of optional "-foo" lines. (This is
2915 overdue, caused by the large number of possibilities that now exist.
2916 Originally there were few.) While I was there, I also converted the
2917 str(n)cmp tests so they don't re-test the leading "-" and the first
2918 character, in the hope this might squeeze out yet more improvement.
2920 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2921 flag allowing group syntax was set by the header_syntax check but not
2922 turned off, possible causing trouble later; (2) The flag was not being
2923 set at all for the header_verify test, causing "group"-style headers to
2924 be rejected. I have now set it in this case, and also caused header_
2925 verify to ignore an empty address taken from a group. While doing this, I
2926 came across some other cases where the code for allowing group syntax
2927 while scanning a header line wasn't quite right (mostly, not resetting
2928 the flag correctly in the right place). These bugs could have caused
2929 trouble for malformed header lines. I hope it is now all correct.
2931 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2932 with the "reply" argument non-NULL. The code, however (which originally
2933 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2934 but it didn't always do it. This confused somebody who was copying the
2935 code for some other use. I have removed all the tests.
2937 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2938 feature that was used to support insecure browsers during the U.S. crypto
2939 embargo. It requires special client support, and Exim is probably the
2940 only MTA that supported it -- and would never use it because real RSA is
2941 always available. This code has been removed, because it had the bad
2942 effect of slowing Exim down by computing (never used) parameters for the
2943 RSA_EXPORT functionality.
2945 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2946 authenticator to fail if there's a tab character in the incoming data
2947 (there should never be unless someone is messing about, as it's supposed
2948 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2949 if the connection is using TLS or if the remote IP is the same as the
2950 local IP, and the "valid-client-cert option" if a client certificate has
2953 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2954 authenticators. This can be used for authorization after authentication
2955 succeeds. (In the case of plaintext, it servers for both authentication
2958 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2959 if any retry times were supplied.
2961 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2962 connection, where there is no client IP address to check. In this
2963 situation, the verify now always succeeds.
2965 PH/25 Applied John Jetmore's -Mset patch.
2967 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2969 PH/27 In a string expansion for a processed (not raw) header when multiple
2970 headers of the same name were present, leading whitespace was being
2971 removed from all of them, but trailing whitespace was being removed only
2972 from the last one. Now trailing whitespace is removed from each header
2973 before concatenation. Completely empty headers in a concatenation (as
2974 before) are ignored.
2976 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2977 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2979 PH/29 [Removed. This was a change that I later backed out, and forgot to
2980 correct the ChangeLog entry (that I had efficiently created) before
2981 committing the later change.]
2983 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2984 address errors (4xx response to RCPT) over the same connection as other
2985 messages routed to the same hosts. Such deliveries are always "forced",
2986 so retry times are not inspected. This resulted in far too many retries
2987 for the affected addresses. The effect occurred only when there were more
2988 hosts than the hosts_max_try setting in the smtp transport when it had
2989 the 4xx errors. Those hosts that it had tried were not added to the list
2990 of hosts for which the message was waiting, so if all were tried, there
2991 was no problem. Two fixes have been applied:
2993 (i) If there are any address or message errors in an SMTP delivery, none
2994 of the hosts (tried or untried) are now added to the list of hosts
2995 for which the message is waiting, so the message should not be a
2996 candidate for sending over the same connection that was used for a
2997 successful delivery of some other message. This seems entirely
2998 reasonable: after all the message is NOT "waiting for some host".
2999 This is so "obvious" that I'm not sure why it wasn't done
3000 previously. Hope I haven't missed anything, but it can't do any
3001 harm, as the worst effect is to miss an optimization.
3003 (ii) If, despite (i), such a delivery is accidentally attempted, the
3004 routing retry time is respected, so at least it doesn't keep
3005 hammering the server.
3007 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3008 in ${readsocket because some servers need this prod.
3010 PH/32 Added some extra debug output when updating a wait-xxx database.
3012 PH/33 The hint "could be header name not terminated by colon", which has been
3013 given for certain expansion errors for a long time, was not being given
3014 for the ${if def:h_colon_omitted{... case.
3016 PH/34 The spec says: "With one important exception, whenever a domain list is
3017 being scanned, $domain contains the subject domain." There was at least
3018 one case where this was not true.
3020 PH/35 The error "getsockname() failed: connection reset by peer" was being
3021 written to the panic log as well as the main log, but it isn't really
3022 panic-worthy as it just means the connection died rather early on. I have
3023 removed the panic log writing for the ECONNRESET error when getsockname()
3026 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3027 runs only) independently of the message's sender address. This meant
3028 that, if the 4xx error was in fact related to the sender, a different
3029 message to the same recipient with a different sender could confuse
3030 things. In particular, this can happen when sending to a greylisting
3031 server, but other circumstances could also provoke similar problems.
3032 I have changed the default so that the retry time for these errors is now
3033 based a combination of the sender and recipient addresses. This change
3034 can be overridden by setting address_retry_include_sender=false in the
3037 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3038 remote server are returned as part of bounce messages. This was not
3039 happening for LMTP over a pipe (the lmtp transport), but now it is the
3040 same for both kinds of LMTP.
3042 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3043 in header lines that were in fact CNAMEs. This is no longer the case.
3045 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3046 and queue runs started by the daemon processed all messages. This has
3047 been fixed so that -R and -S can now usefully be given with -q<time>.
3049 PH/40 Import PCRE release 6.7 (fixes some bugs).
3051 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3053 PH/42 Give an error if -q is specified more than once.
3055 PH/43 Renamed the variables $interface_address and $interface_port as
3056 $received_ip_address and $received_port, to make it clear that these
3057 values apply to message reception, and not to the outgoing interface when
3058 a message is delivered. (The old names remain recognized, of course.)
3060 PH/44 There was no timeout on the connect() call when using a Unix domain
3061 socket in the ${readsocket expansion. There now is.
3063 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3064 be meaningful with "accept".
3066 SC/01 Eximstats V1.43
3067 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3069 SC/02 Eximstats V1.44
3070 Use a glob alias rather than an array ref in the generated
3071 parser. This improves both readability and performance.
3073 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3074 Collect SpamAssassin and rejection statistics.
3075 Don't display local sender or destination tables unless
3076 there is data to show.
3077 Added average volumes into the top table text output.
3079 SC/04 Eximstats V1.46
3080 Collect data on the number of addresses (recipients)
3081 as well as the number of messages.
3083 SC/05 Eximstats V1.47
3084 Added 'Message too big' to the list of mail rejection
3085 reasons (thanks to Marco Gaiarin).
3087 SC/06 Eximstats V1.48
3088 Mainlog lines which have GMT offsets and are too short to
3089 have a flag are now skipped.
3091 SC/07 Eximstats V1.49 (Alain Williams)
3092 Added the -emptyok flag.
3094 SC/08 Eximstats V1.50
3095 Fixes for obtaining the IP address from reject messages.
3097 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3098 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3099 whitespace changes from 4.64-PH/27
3101 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3104 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3105 are found, allow negative numbers in numeric criteria)
3107 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3109 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3110 to match changes made in 4.64-PH/43
3112 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3114 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3115 "domain" is actually an IP address, and force a failure. This locks out
3116 those revolvers/nameservers that support "A-for-A" lookups, in
3117 contravention of the specifications.
3119 PH/48 When a host name was looked up from an IP address, and the subsequent
3120 forward lookup of the name timed out, the host name was left in
3121 $sender_host_name, contrary to the specification.
3123 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3124 restricted to single-key lookups, Exim was not diagnosing an error if
3125 * or *@ was used with a query-style lookup.
3127 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3129 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3130 long ago, but noone interested enough thought of it. Let's just say that
3131 the "1.1" means that there are some new functions that weren't there at
3132 some point in the past.
3134 PH/51 Error processing for expansion failure of helo_data from an smtp
3135 transport during callout processing was broken.
3137 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3138 tested/used via the -bh/-bhc/-bs options.
3140 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3141 bug, fixed in subsequent PCRE releases).
3143 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3144 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3146 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3152 SC/01 Use a glob alias rather than an array ref in eximstats generated
3153 parser. This improves both readability and performance.
3155 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3156 Don't display local sender or destination tables in eximstats unless
3157 there is data to show.
3158 Added average volumes into the eximstats top table text output.
3160 SC/03 Collect data on the number of addresses (recipients) as well
3161 as the number of messages in eximstats.
3163 TF/01 Correct an error in the documentation for the redirect router. Exim
3164 does (usually) call initgroups() when daemonizing.
3166 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3167 with consistent privilege compared to when running as a daemon.
3169 TF/03 Note in the spec that $authenticated_id is not set for local
3170 submissions from trusted users.
3172 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3173 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3175 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3176 by adding some example configuration directives to the default
3177 configuration file. A little bit of work is required to uncomment the
3178 directives and define how usernames and passwords are checked, but
3179 there is now a framework to start from.
3181 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3182 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3183 without this. I don't know how relevant this is to other LDAP libraries.
3185 PH/02 Add the verb name to the "unknown ACL verb" error.
3187 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3189 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3191 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3192 directory not expanded when it should be if an expanded home directory
3193 was set for the address (which is overridden by the transport).
3195 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3198 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3199 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3200 because it is too late at that time, and has no effect.
3202 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3203 security issue with \' (bugzilla #107). I could not use the
3204 PQescapeStringConn() function, because it needs a PGconn value as one of
3207 PH/08 When testing addresses using -bt, indicate those final addresses that
3208 are duplicates that would not cause an additional delivery. At least one
3209 person was confused, thinking that -bt output corresponded to deliveries.
3210 (Suppressing duplicates isn't a good idea as you lose the information
3211 about possibly different redirections that led to the duplicates.)
3213 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3214 systems where poll() doesn't work, in particular OS X.
3216 PH/10 Added more information to debugging output for retry time not reached.
3218 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3219 operations in malware.c.
3221 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3224 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3225 syslog_duplication set false, log lines that would normally be written
3226 both the the main log and to the reject log were not written to syslog at
3229 PH/14 In the default configuration, change the use of "message" in ACL warn
3230 statements to "add_header".
3232 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3233 not followed by a command (e.g. "seen endif").
3235 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3236 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3239 PH/17 Added extra conditions to the default value of delay_warning_condition
3243 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3244 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3245 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3248 The Auto-Submitted: and various List- headers are standardised, whereas I
3249 don't think Precedence: ever was.
3251 PH/18 Refactored debugging code in route_finduser() to show more information,
3252 in particular, the error code if getpwnam() issues one.
3254 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3255 This is apparently needed in addition to the PH/07 change above to avoid
3256 any possible encoding problems.
3258 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3259 but not after initializing Perl.
3261 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3262 output them only if debugging. By default they are written stderr,
3263 apparently, which is not desirable.
3265 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3268 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3271 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3273 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3274 authenticated or an ident call has been made. Suppress the default
3275 values for $authenticated_id and $authenticated_sender (but permit -oMai
3276 and -oMas) when testing with -bh.
3278 PH/24 Re-jigged the order of the tests in the default configuration so that the
3279 tests for valid domains and recipients precede the DNS black list and CSA
3280 tests, on the grounds that those ones are more expensive.
3282 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3283 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3284 This bug exists in every version of Exim that I still have, right back to
3287 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3288 However, an attempt to turn on case-sensitivity in a regex key by
3289 including (?-i) didn't work because the subject string was already
3290 lowercased, and the effects were non-intuitive. It turns out that a
3291 one-line patch can be used to allow (?-i) to work as expected.
3297 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3298 other effects) broke the use of negated acl sub-conditions.
3300 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3303 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3304 "Deny" causes Exim to reject the incoming connection with a 554 error.
3305 Unfortunately, if there is a major crisis, such as a disk failure,
3306 tcp-wrappers gives "deny", whereas what one would like would be some
3307 kind of temporary error. A kludge has been added to help with this.
3308 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3309 554 error is used if errno is still zero or contains ENOENT (which occurs
3310 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3313 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3315 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3316 errors. Otherwise a message that provokes a temporary error (when other
3317 messages do not) can cause a whole host to time out.
3319 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3320 addresses were routed directly to files or pipes from a redirect router.
3321 File deliveries just didn't batch; pipe deliveries might have suffered
3324 PH/06 A failure to get a lock for a hints database would erroneously always say
3325 "Failed to get write lock", even when it was really a read lock.
3327 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3328 of 0600. This has been changed to use the value of the lockfile_mode
3329 option (which defaults to 0600).
3331 PH/08 Applied small patch from the Sieve maintainer.
3333 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3334 folder from quota calculations, a direct delivery into this folder messed
3335 up the contents of the maildirsize file. This was because the regex was
3336 used only to exclude .Trash (or whatever) when the size of the mailbox
3337 was calculated. There was no check that a delivery was happening into an
3338 excluded directory. This bug has been fixed by ignoring all quota
3339 processing for deliveries into excluded directories.
3341 PH/10 Added the maildirfolder_create_regex option to appendfile.
3347 PH/01 The code for finding all the local interface addresses on a FreeBSD
3348 system running IPv6 was broken. This may well have applied to all BSD
3349 systems, as well as to others that have similar system calls. The broken
3350 code found IPv4 interfaces correctly, but gave incorrect values for the
3351 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3352 that it would not match correctly against @[] and not recognize the IPv6
3355 PH/02 The ipliteral router was not recognizing addresses of the form user@
3356 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3358 PH/03 Added disable_ipv6.
3360 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3361 decoded form, because it is most often used to construct To: headers
3362 lines in autoreplies, and the decoded form may well be syntactically
3363 invalid. However, $reply_address has leading white space removed, and all
3364 newlines turned into spaces so that the autoreply transport does not
3367 PH/05 If group was specified without a user on a router, and no group or user
3368 was specified on a transport, the group from the router was ignored.
3370 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3371 for visible compile-time settings that can be used to change these
3372 numbers, for those that want even more. Backwards compatibility with old
3373 spool files has been maintained. However, going back to a previous Exim
3374 release will lost any variables that are in spool files.
3376 PH/07 Two small changes when running in the test harness: increase delay when
3377 passing a TCP/IP connection to a new process, in case the original
3378 process has to generate a bounce, and remove special handling of
3379 127.0.0.2 (sic), which is no longer necessary.
3381 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3382 be the same on different OS.
3384 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3387 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3388 whether --show-vars was specified or not
3390 JJ/02 exipick: Added support for new ACL variable spool format introduced
3393 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3394 syntactically invalid From: or Reply-to: line, and a filter used this to
3395 generate an autoreply, and therefore failed to obtain an address for the
3396 autoreply, Exim could try to deliver to a non-existent relative file
3397 name, causing unrelated and misleading errors. What now happens is that
3398 it logs this as a hard delivery error, but does not attempt to create a
3401 PH/11 The exinext utility has a -C option for testing purposes, but although
3402 the given file was scanned by exinext itself; it wasn't being passed on
3403 when Exim was called.
3405 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3406 an end-of-file indication when reading a command response.
3408 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3409 compiled. In many other places in Exim, IPv6 addresses are always
3410 recognized, so I have changed this. It also means that IPv4 domain
3411 literals of the form [IPV4:n.n.n.n] are now always recognized.
3413 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3414 used if the router is not running as root, for example, when verifying at
3415 ACL time, or when using -bh. The debugging output from this situation was
3416 non-existent - all you got was a failure to exec. I have made two
3419 (a) Failures to set uid/gid, the current directory, or a process leader
3420 in a subprocess such as that created by queryprogram now generate
3421 suitable debugging output when -d is set.
3423 (b) The queryprogram router detects when it is not running as root,
3424 outputs suitable debugging information if -d is set, and then runs
3425 the subprocess without attempting to change uid/gid.
3427 PH/15 Minor change to Makefile for building test_host (undocumented testing
3430 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3431 additional section of a DNS packet that returns MX or SRV records.
3432 Instead, it always explicitly searches for A/AAAA records. This avoids
3433 major problems that occur when a DNS server includes only records of one
3434 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3435 fixed another bug: if SRV records were looked up and the corresponding
3436 address records were *not* found in the additional section, the port
3437 values from the SRV records were lost.
3439 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3440 using the correct key (the original address) when searching the retry
3441 rules in order to find which one to use for generating the retry hint.
3443 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3444 adding the default one. Similarly, if it contains a Reply-To: header, the
3445 errors_reply_to option, if set, is not used.
3447 PH/19 When calculating a retry time, Exim used to measure the "time since
3448 failure" by looking at the "first failed" field in the retry record. Now
3449 it does not use this if it is later than than the arrival time of the
3450 message. Instead it uses the arrival time. This makes for better
3451 behaviour in cases where some deliveries succeed, thus re-setting the
3452 "first failed" field. An example is a quota failure for a huge message
3453 when small messages continue to be delivered. Without this change, the
3454 "time since failure" will always be short, possible causing more frequent
3455 delivery attempts for the huge message than are intended.
3456 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3458 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3459 $1, $2, $3) because the numerical variables can be reset during some
3460 expansion items (e.g. "match"), thereby losing the authentication data.
3462 PH/21 Make -bV show the size of off_t variables so that the test suite can
3463 decide whether to run tests for quotas > 2G.
3465 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3466 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3467 filecount value is greater than 2G or if a quota value is greater than 2G
3468 on a system where the size of off_t is not greater than 4, a panic error
3471 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3472 never match. The debug and -bh output now contains an explicit error
3473 message indicating a malformed IPv4 address or mask.
3475 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3476 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3477 PH/23 above applies.
3479 PH/25 Do not write to syslog when running in the test harness. The only
3480 occasion when this arises is a failure to open the main or panic logs
3481 (for which there is an explicit test).
3483 PH/26 Added the /no_tell option to "control=freeze".
3485 PH/27 If a host name lookup failed very early in a connection, for example, if
3486 the IP address matched host_lookup and the reverse lookup yielded a name
3487 that did not have a forward lookup, an error message of the form "no IP
3488 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3489 could be logged. Now it outputs the IP address instead of "NULL".
3491 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3492 allows the sender and the authenticated sender to be set when
3493 submitting a message from within Exim. Since child_open_exim() is
3494 documented for local_scan(), the new function should be too.
3496 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3497 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3498 results in an empty string is now treated as unset.
3500 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3502 PH/31 Added sender_verify_fail logging option.
3504 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3505 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3506 tidied the source and removed it altogether.
3508 PH/33 When a queue run was abandoned because the load average was too high, a
3509 log line was always written; now it is written only if the queue_run log
3510 selector is set. In addition, the log line for abandonment now contains
3511 information about the queue run such as the pid. This is always present
3512 in "start" and "stop" lines but was omitted from the "abandon" line.
3514 PH/34 Omit spaces between a header name and the colon in the error message that
3515 is given when verify = headers_syntax fails (if there are lots of them,
3516 the message gets confusing).
3518 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3519 names, as there are now some PTR records that contain slashes. This check
3520 is only to protect against broken name servers that fall over on strange
3521 characters, so the fact that it applies to all lookups doesn't matter.
3523 PH/36 Now that the new test suite is complete, we can remove some of the
3524 special code in Exim that was needed for the old test suite. For example,
3525 sorting DNS records because real resolvers return them in an arbitrary
3526 order. The new test suite's fake resolver always returns records in the
3529 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3530 bounces) except when queue_only is set, to avoid logging races between
3531 the different processes.
3533 PH/38 Panic-die if .include specifies a non-absolute path.
3535 PH/39 A tweak to the "H" retry rule from its user.
3537 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3538 a label. They prevented compilation on older perls.
3540 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3541 a warning to be raised on newish perls.
3543 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3544 on queue. Changes to match documented behaviour of showing count of
3545 messages matching specified criteria.
3547 PH/40 Changed the default ident timeout from 30s to 5s.
3549 PH/41 Added support for the use of login_cap features, on those BSD systems
3550 that have them, for controlling the resources used by pipe deliveries.
3552 PH/42 The content-scanning code uses fopen() to create files in which to put
3553 message data. Previously it was not paying any attention to the mode of
3554 the files. Exim runs with umask(0) because the rest of the code creates
3555 files with open(), and sets the required mode explicitly. Thus, these
3556 files were ending up world-writeable. This was not a big issue, because,
3557 being within the spool directory, they were not world-accessible. I have
3558 created a function called modefopen, which takes an additional mode
3559 argument. It sets umask(777), creates the file, chmods it to the required
3560 mode, then resets the umask. All the relevant calls to fopen() in the
3561 content scanning code have been changed to use this function.
3563 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3564 to 24 hours. This avoids potential overflow problems when processing G
3565 and H retry rules. I suspect nobody ever tinkers with this value.
3567 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3569 PH/45 When the plaintext authenticator is running as a client, the server's
3570 challenges are checked to ensure they are valid base64 strings. By
3571 default, the authentication attempt is cancelled if an invalid string is
3572 received. Setting client_ignore_invalid_base64 true ignores these errors.
3573 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3574 they are received. Thus, the responses can be made to depend on the
3575 challenges. If an invalid string is ignored, an empty string is placed in
3578 PH/46 Messages that are created by the autoreply transport now contains a
3579 References: header, in accordance with RFCs 2822 and 3834.
3581 PH/47 Added authenticated_sender_force to the smtp transport.
3583 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3585 PH/49 Installed latest patch from the Sieve maintainer.
3587 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3588 was also set, the appendfile transport was unnecessarily scanning a
3589 directory of message files (e.g. for maildir delivery) to find the count
3590 of files (along with the size), even though it did not need this
3591 information. It now does the scan only if it needs to find either the
3592 size of the count of files.
3594 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3596 PH/52 Two bugs concerned with error handling when the smtp transport is
3599 (i) Exim was not creating retry information for temporary errors given
3600 for individual recipients after the DATA command when the smtp transport
3601 was used in LMTP mode. This meant that they could be retried too
3602 frequently, and not timed out correctly.
3604 (ii) Exim was setting the flag that allows error details to be returned
3605 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3606 recipients that were returned after the DATA command.
3608 PH/53 This is related to PH/52, but is more general: for any failing address,
3609 when detailed error information was permitted to be returned to the
3610 sender, but the error was temporary, then after the final timeout, only
3611 "retry timeout exceeded" was returned. Now it returns the full error as
3612 well as "retry timeout exceeded".
3614 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3615 do this, and (what is worse) MTAs that accept it.
3617 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3618 will now be deprecated.
3620 PH/56 New os.c-cygwin from the Cygwin maintainer.
3622 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3623 formats (previously only available in exim formats via -bpr, -bpru,
3624 and -bpra. Now also available in native and exiqgrep formats)
3626 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3627 with very large, slow to parse queues
3629 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3631 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3633 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3634 responses to authentication challenges, though it was showing the
3635 challenges; (ii) I've removed the CR characters from the debug output for
3638 PH/58 Allow for the insertion of a newline as well as a space when a string
3639 is turned into more than one encoded-word during RFC 2047 encoding. The
3640 Sieve code now uses this.
3642 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3643 data_4xx, lost_connection, tls_required.
3645 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3646 message was being sent as an SMTP response.
3648 PH/61 Add -l and -k options to exicyclog.
3650 PH/62 When verifying, if an address was redirected to one new address, so that
3651 verification continued, and the new address failed or deferred after
3652 having set something in $address_data, the value of $address_data was not
3653 passed back to the ACL. This was different to the case when no
3654 redirection occurred. The value is now passed back in both cases.
3656 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3657 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3658 use login_cap.h, so on its own it isn't the distinguishing feature. The
3659 new name refers directly to the setclassresources() function.
3661 PH/65 Added configuration files for NetBSD3.
3663 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3665 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3666 is preferred over IPv4.
3668 PH/68 The bounce_return_message and bounce_return_body options were not being
3669 honoured for bounces generated during the reception of non-SMTP messages.
3670 In particular, this applied to messages rejected by the ACL. This bug has
3671 been fixed. However, if bounce_return_message is true and bounce_return_
3672 body is false, the headers that are returned for a non-SMTP message
3673 include only those that have been read before the error was detected.
3674 (In the case of an ACL rejection, they have all been read.)
3676 PH/69 The HTML version of the specification is now built in a directory called
3677 spec_html instead of spec.html, because the latter looks like a path with
3678 a MIME-type, and this confuses some software.
3680 PH/70 Catch two compiler warnings in sieve.c.
3682 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3683 function verify_get_ident() calls ip_connect() to connect a socket, but
3684 if the "connect()" function timed out, ip_connect() used to close the
3685 socket. However, verify_get_ident() also closes the socket later, and in
3686 between Exim writes to the log, which may get opened at this point. When
3687 the socket was closed in ip_connect(), the log could get the same file
3688 descriptor number as the socket. This naturally causes chaos. The fix is
3689 not to close the socket in ip_connect(); the socket should be closed by
3690 the function that creates it. There was only one place in the code where
3691 this was missing, in the iplookup router, which I don't think anybody now
3692 uses, but I've fixed it anyway.
3694 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3695 well as to direct DNS lookups. Otherwise the handling of names in host
3696 lists is inconsistent and therefore confusing.
3702 PH/01 Two changes to the default runtime configuration:
3704 (1) Move the checks for relay_from_hosts and authenticated clients from
3705 after to before the (commented out) DNS black list checks.
3707 (2) Add control=submission to the relay_from_hosts and authenticated
3708 clients checks, on the grounds that messages accepted by these
3709 statements are most likely to be submissions.
3711 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3713 (1) Generate an error if the third argument for the ${prvs expansion is
3716 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3719 (3) Reset the variables that are obtained from the first argument of
3720 ${prvscheck and used in the second argument before leaving the code,
3721 because their memory is reclaimed, so using them afterwards may do
3724 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3725 one (it's much easier than Tom thought :-).
3727 (5) Because of (4), we can now allow for the use of $prvscheck_result
3728 inside the third argument.
3730 PH/03 For some reason, the default setting of PATH when running a command from
3731 a pipe transport was just "/usr/bin". I have changed it to
3734 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3735 anything to be listed in the output from -bV.
3737 PH/05 When a filter generated an autoreply, the entire To: header line was
3738 quoted in the delivery log line, like this:
3740 => >A.N.Other <ano@some.domain> <original@ddress> ...
3742 This has been changed so that it extracts the operative address. There
3743 may be more than one such address. If so, they are comma-separated, like
3746 => >ano@some.domain,ona@other.domain <original@ddress> ...
3748 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3749 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3750 not being looked up in the rDNS to get a host name, Exim was showing the
3751 IP address twice in Received: lines, even though the IP addresses were
3752 identical. For example:
3754 Received: from [1.2.3.4] (helo=[1.2.3.4])
3756 However, if the real host name was known, it was omitting the HELO data
3757 if it matched the actual IP address. This has been tidied up so that it
3758 doesn't show the same IP address twice.
3760 PH/07 When both +timestamp and +memory debugging was on, the value given by
3761 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3762 called by the debug printing, thereby overwriting the timestamp buffer.
3763 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3765 PH/08 When the original message was included in an autoreply transport, it
3766 always said "this is a copy of the message, including all the headers",
3767 even if body_only or headers_only was set. It now gives an appropriate
3770 PH/09 Applied a patch from the Sieve maintainer which:
3772 o fixes some comments
3773 o adds the (disabled) notify extension core
3774 o adds some debug output for the result of if/elsif tests
3775 o points to the current vacation draft in the documentation
3776 and documents the missing references header update
3780 o fixes a bug in processing the envelope test (when testing
3781 multiple envelope elements, the last element determined the
3784 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3785 Electronic Mail") by including:
3787 Auto-submitted: auto-generated
3789 in the messages that it generates (bounce messages and others, such as
3790 warnings). In the case of bounce messages for non-SMTP messages, there was
3791 also a typo: it was using "Auto_submitted" (underscore instead of
3792 hyphen). Since every message generated by Exim is necessarily in response
3793 to another message, thes have all been changed to:
3795 Auto-Submitted: auto-replied
3797 in accordance with these statements in the RFC:
3799 The auto-replied keyword:
3801 - SHOULD be used on messages sent in direct response to another
3802 message by an automatic process,
3804 - MUST NOT be used on manually-generated messages,
3806 - MAY be used on Delivery Status Notifications (DSNs) and Message
3807 Disposition Notifications (MDNs),
3809 - MUST NOT be used on messages generated by automatic or periodic
3810 processes, except for messages which are automatic responses to
3813 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3814 to the default Received: header definition.
3816 PH/12 Added log selector acl_warn_skipped (default on).
3818 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3819 variables because (a) they are in the wrong storage pool and (b) even if
3820 they were copied, it wouldn't work properly because of the caching.
3822 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3823 checking when decoding. Apparently there are clients that generate
3824 overlong encoded strings. Why am I not surprised?
3826 PH/15 If the first argument of "${if match_address" was not empty, but did not
3827 contain an "@" character, Exim crashed. Now it writes a panic log message
3828 and treats the condition as false.
3830 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3832 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3833 extension "envelope-auth". The code is finished and in agreement with
3834 other implementations, but there is no documentation so far and in fact,
3835 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3836 not changing the active code.
3838 Print executed "if" and "elsif" statements when debugging is used. This
3839 helps a great deal to understand what a filter does.
3841 Document more things not specified clearly in RFC3028. I had all this
3842 sorted out, when out of a sudden new issues came to my mind. Oops."
3844 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3847 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3848 canonical form (as documented). However, after a host name lookup from
3849 the IP address, check_host() was doing a simple string comparison with
3850 addresses acquired from the DNS when checking that the found name did
3851 have the original IP as one of its addresses. Since any found IPv6
3852 addresses are likely to be in abbreviated form, the comparison could
3853 fail. Luckily, there already exists a function for doing the comparison
3854 by converting both addresses to binary, so now that is used instead of
3855 the text comparison.
3857 PH/20 There was another similar case to PH/19, when a complete host name was
3858 given in a host list; looking up its IP address could give an abbreviated
3859 form, whereas the current host's name might or might not be abbreviated.
3860 The same fix has been applied.
3866 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3867 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3870 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3871 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3873 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3875 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3876 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3877 Also, for "Auto-Submitted", treat anything other than "no" as
3878 non-personal, in accordance with RFC 3834. (Previously it treated
3879 anything starting "auto-" as non-personal.)
3881 TF/01 The control=submission/name=... option had a problem with syntax
3882 errors if the name included a slash character. The /name= option
3883 now slurps the rest of the string, so it can include any characters
3884 but it must come last in the list of options (after /sender_retain
3887 PH/05 Some modifications to the interface to the fake nameserver for the new
3895 TK/01 Added the "success_on_redirect" address verification option. See
3896 NewStuff for rationale and an example.
3898 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3900 PH/02 Patch to exigrep to allow it to work on syslog lines.
3902 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3903 fread() to skip over the body file's header line, because in Cygwin the
3904 header line is locked and is inaccessible.
3906 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3907 co-exist for some time) to make it clear that it is the Exim ID that is
3908 referenced, not the Message-ID: header line.
3910 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3911 string_format() function, because snprintf() does not exist on all
3914 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3915 ${stat: expansion item.
3917 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3918 protocol synchronization error", to keep the pedants happy.
3920 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3921 well as for IRIX systems, when gcc is being used. See the host.c source
3924 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3926 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3929 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3930 transport and to the smtp transport in LMTP mode.
3932 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3934 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3935 run a filter in a subprocess. This could lead to confusion in subsequent
3936 lookups in the parent process. There should also be a search_tidyup() at
3937 the end of the subprocess.
3939 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3940 only if the host matched helo_try_verify_hosts, which caused the
3941 verification to occur when the EHLO/HELO command was issued. The ACL just
3942 tested the remembered result. Now, if a previous verification attempt has
3943 not happened, "verify = helo" does it there and then.
3945 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3947 TK/03 Fix log output including CR from clamd.
3949 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3950 exist provoked a memory error which could cause a segfault.
3952 PH/15 Installed PCRE 6.2
3954 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3956 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3957 of the problem. Specifically, suggested +O2 rather than +O1 for the
3960 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3962 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3963 record, the port was not correctly logged when the outgoing_port log
3964 selector was set (it logged the transort's default port).
3966 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3967 fallback_hosts, and "hosts" in the smtp transport.
3969 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3970 host errors such as "Connection refused".
3972 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3973 authentication with radiusclient 0.4.9:
3975 - Error returned from rc_read_config was caught wrongly
3976 - Username/password not passed on to radius server due to wrong length.
3978 The presumption is that some radiusclient API changes for 4.51/PH/17
3979 were not taken care of correctly. The code is still untested by me (my
3980 Linux distribution still has 0.3.2 of radiusclient), but it was
3981 contributed by a Radius user.
3983 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3984 expanding the "port" option of the smtp transport.
3986 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3987 while reading a MIME header. Thanks to Tom Hughes for a patch.
3989 PH/24 Include config.h inside local_scan.h so that configuration settings are
3992 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3993 that in an ACL for RCPT (for example), you can examine exactly what was
3996 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3997 commands, but it was not correctly comparing the address with the actual
3998 client host address. Thus, it would show the EHLO address in Received:
3999 header lines when this was not necessary.
4001 PH/27 Added the % operator to ${eval:}.
4003 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4004 it should be ignoring failures (because with -C, for example, it has lost
4005 privilege). It wasn't ignoring creation failures other than "already
4008 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4011 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4012 because an input line was too long, either on its own, or by virtue of
4013 too many continuations, the temporary file was not being removed, and the
4014 return code was incorrect.
4016 PH/31 Missing "BOOL" in function definition in filtertest.c.
4018 PH/32 Applied Sieve patches from the maintainer.
4020 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4022 PH/33 Added "verify = not_blind".
4024 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4025 Local/Makefile (with some defaults set). These are used in built scripts
4026 such as exicyclog, but they have never been used in the exim_install
4027 script (though there are many overriding facilities there). I have
4028 arranged that the exim_install script now takes note of these two
4031 PH/35 Installed configuration files for Dragonfly.
4033 PH/36 When a locally submitted message by a trusted user did not contain a
4034 From: header, and the sender address was obtained from -f or from an SMTP
4035 MAIL command, and the trusted user did not use -F to supply a sender
4036 name, $originator_name was incorrectly used when constructing a From:
4037 header. Furthermore, $originator_name was used for submission mode
4038 messages from external hosts without From: headers in a similar way,
4039 which is clearly wrong.
4041 PH/37 Added control=suppress_local_fixups.
4043 PH/38 When log_selector = +received_sender was set, and the addition of the
4044 sender made the log line's construction buffer exactly full, or one byte
4045 less than full, an overflow happened when the terminating "\n" was
4048 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4049 when the result of a list match is failure because a DNS lookup failed.
4051 PH/40 RM_COMMAND is now used in the building process.
4053 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4054 the "build-* directories that it finds.
4056 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4057 address in a domain literal was a prefix of an interface address.
4059 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4060 when verifying a sender address, unless rewrite_headers is false.
4062 PH/44 Wrote a long comment about why errors_to addresses are verified as
4063 recipients, not senders.
4065 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4066 the ratelimit ACL was added.
4068 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4070 PH/46 Added extra information about PostgreSQL errors to the error string.
4072 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4073 avoiding the need to install special zones in a real server. This is
4074 backwards compatible; if it can't find the fake resolver, it drops back.
4075 Thus, both old and new test suites can be run.
4077 TF/02 Added util/ratelimit.pl
4079 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4080 clock is set back in time.
4082 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4083 Candler <B.Candler@pobox.com>.
4085 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4086 OK for addresses that are the result of redirections.
4088 PH/48 A number of further additions for the benefit of the new test suite,
4089 including a fake gethostbyname() that interfaces to the fake DNS resolver
4092 TF/06 The fix for widen_domains has also been applied to qualify_single and
4093 search_parents which are the other dnslookup options that can cause
4096 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4099 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4101 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4102 service" if it encountered an MX record with an empty target hostname.
4103 The message is now "an MX or SRV record indicated no SMTP service".
4105 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4106 if the DNS of the sending site is misconfigured. This is quite a
4107 common situation. This change restores the behaviour of treating a
4108 helo verification defer as a failure.
4110 PH/51 If self=fail was set on a router, the bounce message did not include the
4111 actual error message.
4117 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4119 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4120 command itself ended in error, the underlying message about the transport
4121 filter timeout was being overwritten with the pipe command error. Now the
4122 underlying error message should be appended to the second error message.
4124 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4126 PH/02 Reduce the amount of output that "make" produces by default. Full output
4127 can still be requested.
4129 PH/03 The warning log line about a condition test deferring for a "warn" verb
4130 was being output only once per connection, rather than after each
4131 occurrence (because it was using the same function as for successful
4132 "warn" verbs). This seems wrong, so I have changed it.
4134 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4135 it should not have, which might have caused a crash in the right
4136 circumstances, but probably never did.
4138 PH/04 Installed a modified version of Tony Finch's patch to make submission
4139 mode fix the return path as well as the Sender: header line, and to
4140 add a /name= option so that you can make the user's friendly name appear
4143 TF/03 Added the control = fakedefer ACL modifier.
4145 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4146 Mark Lowes for thorough testing.
4148 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4150 TK/03 Merged latest SRS patch from Miles Wilton.
4152 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4153 with the definition in sysexits.h (which is #included earlier).
4154 Fortunately, Exim does not actually use EX_OK. The code used to try to
4155 preserve the sysexits.h value, by assuming that macro definitions were
4156 scanned for macro replacements. I have been disabused of this notion,
4157 so now the code just undefines EX_OK before #including unistd.h.
4159 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4160 in the smtp transport. When a block could not be written in a single
4161 write() function, the timeout was being re-applied to each part-write.
4162 This seems wrong - if the receiver was accepting one byte at a time it
4163 would take for ever. The timeout is now adjusted when this happens. It
4164 doesn't have to be particularly precise.
4166 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4167 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4169 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4170 without a domain if the check to <postmaster@domain> fails.
4172 SC/01 Eximstats: added -xls and the ability to specify output files
4173 (patch written by Frank Heydlauf).
4175 SC/02 Eximstats: use FileHandles for outputting results.
4177 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4179 SC/04 Eximstats: fixed display of large numbers with -nvr option
4181 SC/05 Eximstats: fixed merging of reports with empty tables.
4183 SC/06 Eximstats: added the -include_original_destination flag
4185 SC/07 Eximstats: removed tabs and trailing whitespace.
4187 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4189 TK/06 MBOX spool code: Add real "From " MBOX separator line
4190 so the .eml file is really in mbox format (even though
4191 most programs do not really care). Patch from Alex Miller.
4193 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4194 The latter is generated from $received_to and is only set if the
4195 message has one envelope recipient. SA can use these headers,
4196 obviously out-of-the-box. Patch from Alex Miller.
4198 PH/08 The ${def test on a variable was returning false if the variable's
4199 value was "0", contrary to what the specification has always said!
4200 The result should be true unless the variable is empty.
4202 PH/09 The syntax error of a character other than { following "${if
4203 def:variable_name" (after optional whitespace) was not being diagnosed.
4204 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4205 accidental colon was present, for example, could give incorrect results.
4207 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4208 result is used (not including appendfile, where other changes are about
4211 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4212 This involved changing a lot of size variables from int to off_t. It
4213 should work with maildirs and everything.
4215 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4216 spamd dying while we are connected to it.
4218 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4221 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4222 function for BDB 4.3.
4224 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4226 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4227 the sources for PCRE have been re-arranged and more files are now
4230 PH/15 The code I had for printing potentially long long variables in PH/11
4231 above was not the best (it lost precision). The length of off_t variables
4232 is now inspected at build time, and an appropriate printing format (%ld
4233 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4234 to be "long long int" or "long int". This is needed for the internal
4235 formatting function string_vformat().
4237 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4238 the configuration file to be ":syslog", then the script "guesses" where
4239 the logs files are, rather than using the compiled in default. In our
4240 case the guess is not the same as the compiled default, so the script
4241 suddenly stopped working when I started to use syslog. The patch checks
4242 to see if log_file_path is "". If so, it attempts to read it from exim
4243 with no configuration file to get the compiled in version, before it
4244 falls back to the previous guessing code."
4246 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4247 implementing BATV in an Exim configuration. See NewStuff for the gory
4250 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4251 Makefile that are specific to HP-UX.
4253 PH/18 If the "use_postmaster" option was set for a recipient callout together
4254 with the "random" option, the postmaster address was used as the MAIL
4255 FROM address for the random test, but not for the subsequent recipient
4256 test. It is now used for both.
4258 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4259 patch removes a few documentation additions to RFC 3028, because the
4260 latest draft now contains them. It adds the new en;ascii-case comparator
4261 and a new error check for 8bit text in MIME parts. Comparator and
4262 require names are now matched exactly. I enabled the subaddress
4263 extension, but it is not well tested yet (read: it works for me)."
4265 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4266 rework some of the code of TK/09 above to avoid the hardwired use of
4267 "%lld" and "long long". Replaced the call to snprintf() with a call to
4270 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4271 records point to non-existent hosts", "retry timeout exceeded", and
4272 "retry time not reached for any host after a long failure period".
4274 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4275 experimental DomainKeys support:
4277 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4278 (2) On an error such as an illegally used "control", the wrong name for
4279 the control was given.
4281 These problems did NOT occur unless DomainKeys support was compiled.
4283 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4285 PH/24 Added ${if match_ip condition.
4287 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4288 databases so that it will be absolutely obvious if a crash occurs in the
4289 DB library. This is a regular occurrence (often caused by mis-matched
4292 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4293 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4294 picked up on a user's system that detects such things. There doesn't seem
4295 to be a gcc warning option for this - only an attribute that has to be
4296 put on the function's prototype. It seems that in Fedora Core 4 they have
4297 set this on a number of new functions. No doubt there will be more in due
4300 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4301 specify a transport. However, if an address that was verified by such a
4302 router was the subject of a callout, Exim crashed because it tried to
4303 read the rcpt_include_affixes from the non-existent transport. Now it
4304 just assumes that the setting of that option is false. This bug was
4305 introduced by 4.51/PH/31.
4307 PH/28 Changed -d+all to exclude +memory, because that information is very
4308 rarely of interest, but it makes the output a lot bigger. People tend to
4309 do -d+all out of habit.
4311 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4312 code in os-type was giving problems when libc.so lives in lib64, like on
4315 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4316 aren't the modern standard, and it seems that some systems' include files
4317 don't always have them. Exim was already checking for some of the newer
4318 ones like T_AAAA, and defining it itself. I've added checks for all the
4319 record types that Exim uses.
4321 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4322 not automatically generating a new one, as it is supposed to. This
4323 prevented TLS from working. If the file did exist, but contained invalid
4324 data, a new version was generated, as expected. It was only the case of a
4325 non-existent file that was broken.
4327 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4328 with a change in libdomainkeys > 0.64.
4330 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4331 from DNS. If the selector record carries the flag, it now has
4332 precedence over the domain-wide flag.
4334 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4336 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4337 the use of a port name that isn't defined in /etc/services) occurred, the
4338 message was deferred as in a normal delivery, and thus remained on the
4339 spool, instead of being failed because of the mua_wrapper setting. This
4340 is now fixed, and I tidied up some of the mua_wrapper messages at the
4343 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4344 the messages in a hash of arrays rather than using individual hashes.
4345 This is a bit cleaner and results in dramatic memory savings, albeit
4346 at a slight CPU cost.
4348 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4349 as requested by Marc Sherman.
4351 SC/10 Eximstats: added histograms for user specified patterns as requested
4354 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4356 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4357 fopen() in the content-scanning modules that did not already have it.
4363 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4364 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4366 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4368 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4370 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4371 to test Sieve filters that use "vacation".
4373 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4374 that changes the way the GnuTLS parameters are stored in the cache file.
4375 The new format can be generated externally. For backward compatibility,
4376 if the data in the cache doesn't make sense, Exim assumes it has read an
4377 old-format file, and it generates new data and writes a new file. This
4378 means that you can't go back to an older release without removing the
4381 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4382 work if there are any delivery delays because "one_time" forces the
4383 parent to be marked "delivered", so its unseen clone is never tried
4384 again. For this reason, Exim now forbids the simultaneous setting of
4387 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4388 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4389 entry if you want to know the details. The fix, however, neglected to
4390 consider the case when local delivery batching is involved. The test for
4391 "previously delivered" was not happening when checking to see if an
4392 address could be batched with a previous (undelivered) one; under
4393 certain circumstances this could lead to multiple deliveries to the same
4396 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4397 in its include files, and this causes problems building Exim.
4399 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4400 header_syntax) but Exim was just ignoring anything given after a slash.
4401 In particular, this caused confusion with an attempt to use "verify =
4402 reverse_host_lookup/defer_ok". An error is now given when options are
4403 supplied for verify items that do not have them. (Maybe reverse_host_
4404 lookup should have a defer_ok option, but that's a different point.)
4406 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4407 defined by RFC 821) to 2048, because there were problems with some AUTH
4408 commands, and RFC 1869 says the size should be increased for extended
4409 SMTP commands that take arguments.
4411 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4414 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4415 "unknown" error; now it says that the functionality isn't in the binary.
4417 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4418 an address' error message when a string expansion fails (syntax or
4419 whatever). Otherwise the password may appear in the log. Following change
4420 PH/42 below, there is no longer a chance of it appearing in a bounce
4423 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4425 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4426 of its addresses was ever tried. (Bugzilla bug #2.)
4428 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4429 the result incorrectly in the debug output. (It correctly added a newline
4430 to what was transported.)
4432 TF/01 Added $received_time.
4434 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4435 commented out examples of how to interface to a virus scanner and to
4436 SpamAssassin. Also added commented examples of av_scanner and
4437 spamd_address settings.
4439 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4440 and controls are allowed in which ACLs. There were a couple of minor
4441 errors. Some of the entries in the conditions table (which is a table of
4442 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4443 negation of where the condition IS allowed.
4445 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4447 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4448 header file does not have a version number, so I've had to invent a new
4449 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4450 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4451 radiusclient), but it was contributed by a Radius user.
4453 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4454 files or directories, for OpenSSL.
4456 PH/19 When an Exim process that is running as root has to create an Exim log
4457 file, it does so in a subprocess that runs as exim:exim so as to get the
4458 ownership right at creation (otherwise, other Exim processes might see
4459 the file with the wrong ownership). There was no test for failure of this
4460 fork() call, which would lead to the process getting stuck as it waited
4461 for a non-existent subprocess. Forks do occasionally fail when resources
4462 run out. I reviewed all the other calls to fork(); they all seem to check
4465 PH/20 When checking for unexpected SMTP input at connect time (before writing
4466 the banner), Exim was not dealing correctly with a non-positive return
4467 from the read() function. If the client had disconnected by this time,
4468 the result was a log entry for a synchronization error with an empty
4469 string after "input=" when read() returned zero. If read() returned -1
4470 (an event I could not check), uninitialized data bytes were printed.
4471 There were reports of junk text (parts of files, etc) appearing after
4474 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4476 PH/22 Added support for macro redefinition, and (re)definition in between
4477 driver and ACL definitions.
4479 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4480 forgetting to use the resulting value; it was using the unexpanded value.
4482 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4483 hadn't been configured. The fix is from Juergen Kreileder, who
4484 understands it better than I do:
4486 "Here's what I see happening with three configured cyrus_sasl
4487 authenticators configured (plain, login, cram-md5):
4489 On startup auth_cyrus_sasl_init() gets called for each of these.
4490 This means three calls to sasl_listmech() without a specified mech_list.
4491 => SASL tests which mechs of all available mechs actually work
4492 => three warnings about OTP not working
4493 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4495 With the patch, sasl_listmech() also gets called three times. But now
4496 SASL's mech_list option is set to the server_mech specified in the the
4497 authenticator. Or in other words, the answer from sasl_listmech()
4498 gets limited to just the mech you're testing for (which is different
4500 => the return list contains just 'plain' or 'login', 'cram-md5' or
4501 nothing depending on the value of ob->server_mech.
4503 I've just tested the patch: Authentication still works fine,
4504 unavailable mechs specified in the exim configuration are still
4505 caught, and the auth.log warnings about OTP are gone."
4507 PH/25 When debugging is enabled, the contents of the command line are added
4508 to the debugging output, even when log_selector=+arguments is not
4511 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4512 answer is "GNU", and only if the return is "GNU/something" is the answer
4515 PH/27 $acl_verify_message is now set immediately after the failure of a
4516 verification in an ACL, and so is available in subsequent modifiers. In
4517 particular, the message can be preserved by coding like this:
4519 warn !verify = sender
4520 set acl_m0 = $acl_verify_message
4522 Previously, $acl_verify_message was set only while expanding "message"
4523 and "log_message" when a very denied access.
4525 PH/28 Modified OS/os.c-Linux with
4527 -#ifndef OS_LOAD_AVERAGE
4528 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4530 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4531 nomenclature these days.)
4533 PH/29 Installed patch from the Sieve maintainer that adds the options
4534 sieve_useraddress and sieve_subaddress to the redirect router.
4536 PH/30 In these circumstances:
4537 . Two addresses routed to the same list of hosts;
4538 . First host does not offer TLS;
4539 . First host accepts first address;
4540 . First host gives temporary error to second address;
4541 . Second host offers TLS and a TLS session is established;
4542 . Second host accepts second address.
4543 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4544 and peerdn, if requested) that were in fact used only for the second
4547 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4548 attention to any local part prefix or suffix that was matched by the
4549 router that accepted the address. It now behaves in the same way as it
4550 does for delivery: the affixes are removed from the local part unless
4551 rcpt_include_affixes is set on the transport.
4553 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4554 timeout during the DATA phase of an incoming message.
4556 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4557 applied a patch sanctioned by the Sieve maintainer.
4559 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4560 the uid or gid is negative. A case of a negative gid caused this to be
4561 noticed. The fix allows for either to be negative.
4563 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4564 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4567 PH/36 The change PH/12 above was broken. Fixed it.
4569 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4570 the grounds that routing the same address twice would always produce the
4571 same answer. This might have been true once, but it is certainly no
4572 longer true now. Routing a child address may depend on the previous
4573 routing that produced that child. Some complicated redirection strategies
4574 went wrong when messages had multiple recipients, and made Exim's
4575 behaviour dependent on the order in which the addresses were given.
4577 I have moved the duplicate checking until after the routing is complete.
4578 Exim scans the addresses that are assigned to local and remote
4579 transports, and removes any duplicates. This means that more work will be
4580 done, as duplicates will always all be routed, but duplicates are
4581 presumably rare, so I don't expect this is of any significance.
4583 For deliveries to pipes, files, and autoreplies, the duplicate checking
4584 still happens during the routing process, since they are not going to be
4587 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4588 It corrects a timeout issue with spamd. This is Ian's comment: "The
4589 background is that sometimes spamd either never reads data from a
4590 connection it has accepted, or it never writes response data. The exiscan
4591 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4592 blindly assumes that writes won't block so it may never time out."
4594 PH/39 Allow G after quota size as well as K and M.
4596 PH/40 The value set for $authenticated_id in an authenticator may not contain
4597 binary zeroes or newlines because the value is written to log lines and
4598 to spool files. There was no check on this. Now the value is run through
4599 the string_printing() function so that such characters are converted to
4600 printable escape sequences.
4602 PH/41 $message_linecount is a new variable that contains the total number of
4603 lines in the message. Compare $body_linecount, which is the count for the
4606 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4607 bounce and delay warning messages, except in certain special cases, which
4610 (a) An SMTP error message from a remote host;
4611 (b) A message specified in a :fail: redirection;
4612 (c) A message specified in a "fail" command in a system filter;
4613 (d) A message specified in a FAIL return from the queryprogram router;
4614 (e) A message specified by the cannot_route_message router option.
4616 In these cases only, Exim does include the error details in bounce and
4617 warning messages. There are also a few cases where bland messages such
4618 as "unrouteable address" or "local delivery error" are given.
4620 PH/43 $value is now also set for the "else" part of a ${run expansion.
4622 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4623 being worked on, but at least Exim now implements the latest version to
4626 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4627 process to complete was treated as a delivery failure, a timeout while
4628 writing the message to the pipe was logged, but erroneously treated as a
4629 successful delivery. Such timeouts include transport filter timeouts. For
4630 consistency with the overall process timeout, these timeouts are now
4631 treated as errors, giving rise to delivery failures by default. However,
4632 there is now a new Boolean option for the pipe transport called
4633 timeout_defer, which, if set TRUE, converts the failures into defers for
4634 both kinds of timeout. A transport filter timeout is now identified in
4637 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4638 systems where "make" and "gmake" are different, calling "gmake" at top
4639 level broke things. I've arranged for the value of $(MAKE) to be passed
4640 from the Makefile to this script so that it can call the same version of
4644 A note about Exim versions 4.44 and 4.50
4645 ----------------------------------------
4647 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4648 changes of various kinds. As a consequence, a big documentation update was
4649 needed. This delayed the release for rather longer than seemed good, especially
4650 in the light of a couple of (minor) security issues. Therefore, the changes
4651 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4652 release. So 4.44 and 4.50 are in effect two different branches that both start
4655 I have left the 4.50 change log unchanged; it contains all the changes since
4656 4.43. The change log for 4.44 is below; many of its items are identical to
4657 those for 4.50. This seems to be the most sensible way to preserve the
4658 historical information.
4664 1. Minor wording change to the doc/README.SIEVE file.
4666 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4667 computation of the current number of files was incorrect.
4669 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4670 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4673 4. Give more explanation in the error message when the command for a transport
4674 filter fails to execute.
4676 5. There are several places where Exim runs a non-Exim command in a
4677 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4678 was being done only for the command run by the queryprogram router. It is
4679 now done for all such subprocesses. The other cases are: ${run, transport
4680 filters, and the commands run by the lmtp and pipe transports.
4682 6. Added CONFIGURE_GROUP build-time option.
4684 7. Some older OS have a limit of 256 on the maximum number of file
4685 descriptors. Exim was using setrlimit() to set 1000 as a large value
4686 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4687 systems. I've change it so that if it can't get 1000, it tries for 256.
4689 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4690 was an oversight, and furthermore, ever since the addition of extra
4691 controls (e.g. 4.43/32), the checks on when to allow different forms of
4692 "control" were broken. There should now be diagnostics for all cases when a
4693 control that does not make sense is encountered.
4695 9. Added the /retain_sender option to "control=submission".
4697 10. $recipients is now available in the predata ACL (oversight).
4699 11. Tidy the search cache before the fork to do a delivery from a message
4700 received from the command line. Otherwise the child will trigger a lookup
4701 failure and thereby defer the delivery if it tries to use (for example) a
4702 cached ldap connection that the parent has called unbind on.
4704 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4705 of $address_data from the recipient verification was clobbered by the
4706 sender verification.
4708 13. The value of address_data from a sender verification is now available in
4709 $sender_address_data in subsequent conditions in the ACL statement.
4711 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4713 15. Added a new option "connect=<time>" to callout options, to set a different
4716 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4717 was its contents. (It was OK if the option was not defined at all.)
4719 17. A "Completed" log line is now written for messages that are removed from
4720 the spool by the -Mrm option.
4722 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4723 information about exactly what failed.
4725 19. Added -dd to debug only the daemon process.
4727 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4728 handles timeouts, both on the server side and network timeouts. Renamed the
4729 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4731 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4732 It is now set to "smtps".
4734 22. $host_address is now set to the target address during the checking of
4735 ignore_target_hosts.
4737 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4738 being passed; this would have caused $sender_host_name to have been used if
4739 matching the list had actually called for a host name (not very likely,
4740 since this list is usually IP addresses). A host name is now passed as
4743 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4744 code that specifies a non-restarting handler (typically sigaction() in
4745 modern systems) in an attempt to fix a rare and obscure crash bug.
4747 25. Narrowed the window for a race in the daemon that could cause it to ignore
4748 SIGCHLD signals. This is not a major problem, because they are used only to
4749 wake it up if nothing else does.
4751 26. A malformed maildirsize file could cause Exim to calculate negative values
4752 for the mailbox size or file count. Odd effects could occur as a result.
4753 The maildirsize information is now recalculated if the size or filecount
4756 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4757 support for a long time. Removed HAVE_SYS_VFS_H.
4759 28. Installed the latest version of exipick from John Jetmore.
4761 29. In an address list, if the pattern was not a regular expression, an empty
4762 subject address (from a bounce message) matched only if the pattern was an
4763 empty string. Non-empty patterns were not even tested. This was the wrong
4764 because it is perfectly reasonable to use an empty address as part of a
4765 database query. An empty address is now tested by patterns that are
4766 lookups. However, all the other forms of pattern expect the subject to
4767 contain a local part and a domain, and therefore, for them, an empty
4768 address still always fails if the pattern is not itself empty.
4770 30. Exim went into a mad DNS loop when attempting to do a callout where the
4771 host was specified on an smtp transport, and looking it up yielded more
4772 than one IP address.
4774 31. Re-factored the code for checking spool and log partition space into a
4775 function that finds that data and another that does the check. The former
4776 is then used to implement four new variables: $spool_space, $log_space,
4777 $spool_inodes, and $log_inodes.
4779 32. The RFC2047 encoding function was originally intended for short strings
4780 such as real names; it was not keeping to the 75-character limit for
4781 encoded words that the RFC imposes. It now respects the limit, and
4782 generates multiple encoded words if necessary. To be on the safe side, I
4783 have increased the buffer size for the ${rfc2047: expansion operator from
4786 33. It is now permitted to omit both strings after an "if" condition; if the
4787 condition is true, the result is "true". As before, when the second string
4788 is omitted, a false condition yields an empty string. This makes it less
4789 cumbersome to write custom ACL and router conditions.
4791 34. Failure to deliver a bounce message always caused it to be frozen, even if
4792 there was an errors_to setting on the router. The errors_to setting is now
4795 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4796 canonical form (fully expanded) before being placed in
4797 $sender_host_address.
4799 36. The table in the code that translates DNS record types into text (T_A to
4800 "A" for instance) was missing entries for NS and CNAME. It is just possible
4801 that this could have caused confusion if both these types were looked up
4802 for the same domain, because the text type is used as part of Exim's
4803 per-process caching. But the chance of anyone hitting this buglet seems
4806 37. The dnsdb lookup has been extended in a number of ways.
4808 (1) There is a new type, "zns", which walks up the domain tree until it
4809 finds some nameserver records. It should be used with care.
4811 (2) There is a new type, "mxh", which is like "mx" except that it returns
4812 just the host names, not the priorities.
4814 (3) It is now possible to give a list of domains (or IP addresses) to be
4815 looked up. The behaviour when one of the lookups defers can be
4816 controlled by a keyword.
4818 (4) It is now possible to specify the separator character for use when
4819 multiple records are returned.
4821 38. The dnslists ACL condition has been extended: it is now possible to supply
4822 a list of IP addresses and/or domains to be looked up in a particular DNS
4825 39. Added log_selector=+queue_time_overall.
4827 40. When running the queue in the test harness, wait just a tad after forking a
4828 delivery process, to get repeatability of debugging output.
4830 41. Include certificate and key file names in error message when GnuTLS fails
4831 to set them up, because the GnuTLS error message doesn't include the name
4832 of the failing file when there is a problem reading it.
4834 42. Allow both -bf and -bF in the same test run.
4836 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4838 44. The "Exiscan patch" is now merged into the mainline Exim source.
4840 45. Sometimes the final signoff response after QUIT could fail to get
4841 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4842 < 0 before doing a fflush(). This bug looks as though it goes back to the
4843 introduction of TLS in release 3.20, but "sometimes" must have been rare
4844 because the tests only now provoked it.
4846 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4847 (this can affect the format of dates).
4849 47. exim_tidydb, when checking for the continued existence of a message for
4850 which it has found a message-specific retry record, was not finding
4851 messages that were in split spool directories. Consequently, it was
4852 deleting retry records that should have stayed in existence.
4854 48. Steve fixed some bugs in eximstats.
4856 49. The SPA authentication driver was not abandoning authentication and moving
4857 on to the next authenticator when an expansion was forced to fail,
4858 contradicting the general specification for all authenticators. Instead it
4859 was generating a temporary error. It now behaves as specified.
4861 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4862 (the order specifies the preference for clients). The order is now AES256,
4863 AES128, 3DES, ARCFOUR128.
4865 51. Small patch to Sieve code - explicitly set From: when generating an
4868 52. Exim crashed if a remote delivery caused a very long error message to be
4869 recorded - for instance if somebody sent an entire SpamAssassin report back
4870 as a large number of 550 error lines. This bug was coincidentally fixed by
4871 increasing the size of one of Exim's internal buffers (big_buffer) that
4872 happened as part of the Exiscan merge. However, to be on the safe side, I
4873 have made the code more robust (and fixed the comments that describe what
4876 53. Now that there can be additional text after "Completed" in log lines (if
4877 the queue_time_overall log selector is set), a one-byte patch to exigrep
4878 was needed to allow it to recognize "Completed" as not the last thing in
4881 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4882 patch that reportedly fixes this has been added. I am not expert enough to
4883 create a test for it. This is what the patch creator wrote:
4885 "I found a little strange behaviour of ldap code when working with
4886 Windows 2003 AD Domain, where users was placed in more than one
4887 Organization Units. When I tried to give exim partial DN, the exit code
4888 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4889 But simultaneously result of request was absolutely normal ldap result,
4890 so I produce this patch..."
4892 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4893 REFERENCE, so I have modified the code to exclude the patch when that macro
4896 55. Some experimental protocols are using DNS PTR records for new purposes. The
4897 keys for these records are domain names, not reversed IP addresses. The
4898 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4899 leaves it alone. Component reversal etc. now happens only for IP addresses.
4902 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4904 57. Double the size of the debug message buffer (to 2048) so that more of very
4905 long debug lines gets shown.
4907 58. The exicyclog utility now does better if the number of log files to keep
4908 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4910 59. Two changes related to the smtp_active_hostname option:
4912 (1) $smtp_active_hostname is now available as a variable.
4913 (2) The default for smtp_banner uses $smtp_active_hostname instead
4914 of $primary_hostname.
4916 60. The host_aton() function is supposed to be passed a string that is known
4917 to be a valid IP address. However, in the case of IPv6 addresses, it was
4918 not checking this. This is a hostage to fortune. Exim now panics and dies
4919 if the condition is not met. A case was found where this could be provoked
4920 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4921 components; fortuitously, this particular loophole had already been fixed
4922 by change 4.50/55 above.
4924 If there are any other similar loopholes, the new check in host_aton()
4925 itself should stop them being exploited. The report I received stated that
4926 data on the command line could provoke the exploit when Exim was running as
4927 exim, but did not say which command line option was involved. All I could
4928 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4929 running as the user.
4932 61. There was a buffer overflow vulnerability in the SPA authentication code
4933 (which came originally from the Samba project). I have added a test to the
4934 spa_base64_to_bits() function which I hope fixes it.
4937 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4938 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4940 63. The daemon start-up calls getloadavg() while still root for those OS that
4941 need the first call to be done as root, but it missed one case: when
4942 deliver_queue_load_max is set with deliver_drop_privilege. This is
4943 necessary for the benefit of the queue runner, because there is no re-exec
4944 when deliver_drop_privilege is set.
4946 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4947 This has been fixed.
4949 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4950 was in use, was not putting the data itself into the right store pool;
4951 consequently, it could be overwritten for a subsequent message in the same
4952 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4955 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4957 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4958 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4959 the function were treating the return as a boolean value, which happened to
4960 work because 0=false and not-0=true, but is not correct code.
4962 68. The host_aton() function was not handling scoped IPv6 addresses (those
4963 with, for example, "%eth0" on the end) correctly.
4965 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4966 negated items (that is, ~something) in unsigned ints. Some compilers
4967 apparently mutter when there is no cast.
4969 70. If an address verification called from an ACL failed, and did not produce a
4970 user-specific message (i.e. there was only a "system" message), nothing was
4971 put in $acl_verify_message. In this situation, it now puts the system
4974 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4975 session; change 4.31/43 added the unwanted input to the log line - except
4976 that it did not do this in the start of session case. It now does.
4978 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4979 This is wrong and can cause the other end to generate a synchronization
4980 error if it is another Exim or anything else that does the synchronization
4981 check. A QUIT command is no longer sent after a timeout.
4983 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4984 during host lookups.
4986 74. The defer_ok option of callout verification was not working if it was used
4987 when verifying addresses in header lines, that is, for this case:
4989 verify = header_sender/callout=defer_ok
4991 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4992 those file descriptors could be used for SMTP connections. If anything
4993 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4994 could be sent to the SMTP client, causing chaos. The daemon now opens
4995 stdin, stdout, and stderr to /dev/null when it puts itself into the
4998 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4999 log by default. The user can override this with suitable Perl magic.
5001 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5002 the log message when discard triggers, was not working for the DATA ACL or
5003 for the non-SMTP ACL.
5005 78. Error message wording change in sieve.c.
5007 79. If smtp_accept_max_per_host was set, the number of connections could be
5008 restricted to fewer than expected, because the daemon was trying to set up
5009 a new connection before checking whether the processes handling previous
5010 connections had finished. The check for completed processes is now done
5011 earlier. On busy systems, this bug wouldn't be noticed because something
5012 else would have woken the daemon, and it would have reaped the completed
5015 80. If a message was submitted locally by a user whose login name contained one
5016 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5017 It caused a spool format error. I have fixed the spool reading code. A
5018 related problem was that the "from" clause in the Received: line became
5019 illegal because of the space(s). It is now covered by ${quote_local_part.
5021 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5024 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5026 83. Patch from Sieve maintainer to fix unterminated string problem in
5027 "vacation" handling.
5029 84. Some minor changes to the Linux configuration files to help with other
5030 OS variants using glibc.
5032 85. One more patch for Sieve to update vacation handling to latest spec.
5035 ----------------------------------------------------
5036 See the note above about the 4.44 and 4.50 releases.
5037 ----------------------------------------------------
5043 1. Change 4.43/35 introduced a bug that caused file counts to be
5044 incorrectly computed when quota_filecount was set in an appendfile
5047 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5048 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5051 3. Give more explanation in the error message when the command for a transport
5052 filter fails to execute.
5054 4. There are several places where Exim runs a non-Exim command in a
5055 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5056 was being done only for the command run by the queryprogram router. It is
5057 now done for all such subprocesses. The other cases are: ${run, transport
5058 filters, and the commands run by the lmtp and pipe transports.
5060 5. Some older OS have a limit of 256 on the maximum number of file
5061 descriptors. Exim was using setrlimit() to set 1000 as a large value
5062 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5063 systems. I've change it so that if it can't get 1000, it tries for 256.
5065 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5066 was an oversight, and furthermore, ever since the addition of extra
5067 controls (e.g. 4.43/32), the checks on when to allow different forms of
5068 "control" were broken. There should now be diagnostics for all cases when a
5069 control that does not make sense is encountered.
5071 7. $recipients is now available in the predata ACL (oversight).
5073 8. Tidy the search cache before the fork to do a delivery from a message
5074 received from the command line. Otherwise the child will trigger a lookup
5075 failure and thereby defer the delivery if it tries to use (for example) a
5076 cached ldap connection that the parent has called unbind on.
5078 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5079 of $address_data from the recipient verification was clobbered by the
5080 sender verification.
5082 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5083 was its contents. (It was OK if the option was not defined at all.)
5085 11. A "Completed" log line is now written for messages that are removed from
5086 the spool by the -Mrm option.
5088 12. $host_address is now set to the target address during the checking of
5089 ignore_target_hosts.
5091 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5092 being passed; this would have caused $sender_host_name to have been used if
5093 matching the list had actually called for a host name (not very likely,
5094 since this list is usually IP addresses). A host name is now passed as
5097 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5098 code that specifies a non-restarting handler (typically sigaction() in
5099 modern systems) in an attempt to fix a rare and obscure crash bug.
5101 15. Narrowed the window for a race in the daemon that could cause it to ignore
5102 SIGCHLD signals. This is not a major problem, because they are used only to
5103 wake it up if nothing else does.
5105 16. A malformed maildirsize file could cause Exim to calculate negative values
5106 for the mailbox size or file count. Odd effects could occur as a result.
5107 The maildirsize information is now recalculated if the size or filecount
5110 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5111 support for a long time. Removed HAVE_SYS_VFS_H.
5113 18. Updated exipick to current release from John Jetmore.
5115 19. Allow an empty sender to be matched against a lookup in an address list.
5116 Previously the only cases considered were a regular expression, or an
5119 20. Exim went into a mad DNS lookup loop when doing a callout where the
5120 host was specified on the transport, if the DNS lookup yielded more than
5123 21. The RFC2047 encoding function was originally intended for short strings
5124 such as real names; it was not keeping to the 75-character limit for
5125 encoded words that the RFC imposes. It now respects the limit, and
5126 generates multiple encoded words if necessary. To be on the safe side, I
5127 have increased the buffer size for the ${rfc2047: expansion operator from
5130 22. Failure to deliver a bounce message always caused it to be frozen, even if
5131 there was an errors_to setting on the router. The errors_to setting is now
5134 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5135 canonical form (fully expanded) before being placed in
5136 $sender_host_address.
5138 24. Updated eximstats to version 1.33
5140 25. Include certificate and key file names in error message when GnuTLS fails
5141 to set them up, because the GnuTLS error message doesn't include the name
5142 of the failing file when there is a problem reading it.
5144 26. Expand error message when OpenSSL has problems setting up cert/key files.
5147 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5148 (this can affect the format of dates).
5150 28. exim_tidydb, when checking for the continued existence of a message for
5151 which it has found a message-specific retry record, was not finding
5152 messages that were in split spool directories. Consequently, it was
5153 deleting retry records that should have stayed in existence.
5155 29. eximstats updated to version 1.35
5156 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5157 1.35 - bugfix such that pie charts by volume are generated correctly
5159 30. The SPA authentication driver was not abandoning authentication and moving
5160 on to the next authenticator when an expansion was forced to fail,
5161 contradicting the general specification for all authenticators. Instead it
5162 was generating a temporary error. It now behaves as specified.
5164 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5165 (the order specifies the preference for clients). The order is now AES256,
5166 AES128, 3DES, ARCFOUR128.
5168 31. Small patch to Sieve code - explicitly set From: when generating an
5171 32. Exim crashed if a remote delivery caused a very long error message to be
5172 recorded - for instance if somebody sent an entire SpamAssassin report back
5173 as a large number of 550 error lines. This bug was coincidentally fixed by
5174 increasing the size of one of Exim's internal buffers (big_buffer) that
5175 happened as part of the Exiscan merge. However, to be on the safe side, I
5176 have made the code more robust (and fixed the comments that describe what
5179 33. Some experimental protocols are using DNS PTR records for new purposes. The
5180 keys for these records are domain names, not reversed IP addresses. The
5181 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5182 leaves it alone. Component reversal etc. now happens only for IP addresses.
5185 34. The host_aton() function is supposed to be passed a string that is known
5186 to be a valid IP address. However, in the case of IPv6 addresses, it was
5187 not checking this. This is a hostage to fortune. Exim now panics and dies
5188 if the condition is not met. A case was found where this could be provoked
5189 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5190 components; fortuitously, this particular loophole had already been fixed
5191 by change 4.50/55 or 4.44/33 above.
5193 If there are any other similar loopholes, the new check in host_aton()
5194 itself should stop them being exploited. The report I received stated that
5195 data on the command line could provoke the exploit when Exim was running as
5196 exim, but did not say which command line option was involved. All I could
5197 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5198 running as the user.
5201 35. There was a buffer overflow vulnerability in the SPA authentication code
5202 (which came originally from the Samba project). I have added a test to the
5203 spa_base64_to_bits() function which I hope fixes it.
5206 36. The daemon start-up calls getloadavg() while still root for those OS that
5207 need the first call to be done as root, but it missed one case: when
5208 deliver_queue_load_max is set with deliver_drop_privilege. This is
5209 necessary for the benefit of the queue runner, because there is no re-exec
5210 when deliver_drop_privilege is set.
5212 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5213 was in use, was not putting the data itself into the right store pool;
5214 consequently, it could be overwritten for a subsequent message in the same
5215 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5218 38. Sometimes the final signoff response after QUIT could fail to get
5219 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5220 < 0 before doing a fflush(). This bug looks as though it goes back to the
5221 introduction of TLS in release 3.20, but "sometimes" must have been rare
5222 because the tests only now provoked it.
5228 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5229 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5230 Now it returns an integer. A number of calls were still expecting a T/F
5231 return. Fortuitously, in all cases, the tests worked in OK situations,
5232 which is the norm. However, things would have gone wrong on any write
5233 failures on the smtp file descriptor. This function is used when sending
5234 messages over SMTP and also when doing verify callouts.
5236 2. When Exim is called to do synchronous delivery of a locally submitted
5237 message (the -odf or -odi options), it no longer closes stderr before doing
5240 3. Implemented the mua_wrapper option.
5242 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5244 5. Implemented the functions header_remove(), header_testname(),
5245 header_add_at_position(), and receive_remove_recipient(), and exported them
5248 6. If an ACL "warn" statement specified the addition of headers, Exim already
5249 inserted X-ACL-Warn: at the start if there was no header name. However, it
5250 was not making this test for the second and subsequent header lines if
5251 there were newlines in the string. This meant that an invalid header could
5252 be inserted if Exim was badly configured.
5254 7. Allow an ACL "warn" statement to add header lines at the start or after all
5255 the Received: headers, as well as at the end.
5257 8. Added the rcpt_4xx retry error code.
5259 9. Added postmaster_mailfrom=xxx to callout verification option.
5261 10. Added mailfrom=xxxx to the callout verification option, for verify=
5264 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5265 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5267 12. Inserted some casts to stop certain compilers warning when using pointer
5268 differences as field lengths or precisions in printf-type calls (mostly
5269 affecting debugging statements).
5271 13. Added optional readline() support for -be (dynamically loaded).
5273 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5274 same clock tick as a message's arrival, so that its received time was the
5275 same as the "first fail" time on the retry record, and that message
5276 remained on the queue past the ultimate address timeout, every queue runner
5277 would try a delivery (because it was past the ultimate address timeout) but
5278 after another failure, the ultimate address timeout, which should have then
5279 bounced the address, did not kick in. This was a "< instead of <=" error;
5280 in most cases the first failure would have been in the next clock tick
5281 after the received time, and all would be well.
5283 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5284 being recognized when the domain list was tested by the match_domain
5285 condition in an expansion string.
5287 16. Added the ${str2b64: operator.
5289 17. Exim was always calling setrlimit() to set a large limit for the number of
5290 processes, without checking whether the existing limit was already
5291 adequate. (It did check for the limit on file descriptors.) Furthermore,
5292 errors from getrlimit() and setrlimit() were being ignored. Now they are
5293 logged to the main and panic logs, but Exim does carry on, to try to do its
5294 job under whatever limits there are.
5296 18. Imported PCRE 5.0.
5298 19. Trivial typo in log message " temporarily refused connection" (the leading
5301 20. If the log selector return_path_on_delivery was set and an address was
5302 redirected to /dev/null, the delivery process crashed because it assumed
5303 that a return path would always be set for a "successful" delivery. In this
5304 case, the whole delivery is bypassed as an optimization, and therefore no
5307 21. Internal re-arrangement: the function for sending a challenge and reading
5308 a response while authentication was assuming a zero-terminated challenge
5309 string. It's now changed to take a pointer and a length, to allow for
5310 binary data in such strings.
5312 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5314 23. Exim was not respecting finduser_retries when seeking the login of the
5315 uid under which it was called; it was always trying 10 times. (The default
5316 setting of finduser_retries is zero.) Also, it was sleeping after the final
5317 failure, which is pointless.
5319 24. Implemented tls_on_connect_ports.
5321 25. Implemented acl_smtp_predata.
5323 26. If the domain in control=submission is set empty, Exim assumes that the
5324 authenticated id is a complete email address when it generates From: or
5325 Sender: header lines.
5327 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5328 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5329 chown and chgrp in /bin and hostname in /usr/bin.
5331 28. Exim was keeping the "process log" file open after each use, just as it
5332 does for the main log. This opens the possibility of it remaining open for
5333 long periods when the USR1 signal hits a daemon. Occasional processlog
5334 errors were reported, that could have been caused by this. Anyway, it seems
5335 much more sensible not to leave this file open at all, so that is what now
5338 29. The long-running daemon process does not normally write to the log once it
5339 has entered its main loop, and it closes the log before doing so. This is
5340 so that log files can straightforwardly be renamed and moved. However,
5341 there are a couple of unusual error situations where the daemon does write
5342 log entries, and I had neglected to close the log afterwards.
5344 30. The text of an SMTP error response that was received during a remote
5345 delivery was being truncated at 512 bytes. This is too short for some of
5346 the long messages that one sometimes sees. I've increased the limit to
5349 31. It is now possible to make retry rules that apply only when a message has a
5350 specific sender, in particular, an empty sender.
5352 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5353 it possible to be selective about when SMTP synchronization is enforced.
5355 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5357 32. Implemented hosts_connection_nolog.
5359 33. Added an ACL for QUIT.
5361 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5364 35. Added mailbox_size and mailbox_filecount to appendfile.
5366 36. Added control = no_multiline_responses to ACLs.
5368 37. There was a bug in the logic of the code that waits for the clock to tick
5369 in the case where the clock went backwards by a substantial amount such
5370 that the microsecond fraction of "now" was more than the microsecond
5371 fraction of "then" (but the whole seconds number was less).
5373 38. Added support for the libradius Radius client library this is found on
5374 FreeBSD (previously only the radiusclient library was supported).
5380 1. When certain lookups returned multiple values in the form name=value, the
5381 quoting of the values was not always being done properly. Specifically:
5382 (a) If the value started with a double quote, but contained no whitespace,
5384 (b) If the value contained whitespace other than a space character (i.e.
5385 tabs or newlines or carriage returns) it was not quoted.
5386 This fix has been applied to the mysql and pgsql lookups by writing a
5387 separate quoting function and calling it from the lookup code. The fix
5388 should probably also be applied to nisplus, ibase and oracle lookups, but
5389 since I cannot test any of those, I have not disturbed their existing code.
5391 2. A hit in the callout cache for a specific address caused a log line with no
5392 reason for rejecting RCPT. Now it says "Previous (cached) callout
5393 verification failure".
5395 3. There was an off-by-one bug in the queryprogram router. An over-long
5396 return line was truncated at 256 instead of 255 characters, thereby
5397 overflowing its buffer with the terminating zero. As well as fixing this, I
5398 have increased the buffer size to 1024 (and made a note to document this).
5400 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5401 when Exim is waiting for an SMTP response from a remote server, Exim
5402 restarts its select() call on the socket, thereby resetting its timeout.
5403 This is not a problem when such interrupts are rare. Somebody set up a cron
5404 job to run exiwhat every 2 minutes, which is less than the normal select()
5405 timeout (5 or 10 minutes). This meant that the select() timeout never
5406 kicked in because it was always reset. I have fixed this by comparing the
5407 time when an interrupt arrives with the time at the start of the first call
5408 to select(). If more time than the timeout has elapsed, the interrupt is
5409 treated as a timeout.
5411 5. Some internal re-factoring in preparation for the addition of Sieve
5412 extensions (by MH). In particular, the "personal" test is moved to a
5413 separate function, and given an option for scanning Cc: and Bcc: (which is
5414 not set for Exim filters).
5416 6. When Exim created an email address using the login of the caller as the
5417 local part (e.g. when creating a From: or Sender: header line), it was not
5418 quoting the local part when it contained special characters such as @.
5420 7. Installed new OpenBSD configuration files.
5422 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5423 try to make them clearer.
5425 9. Callout options, other than the timeout value, were being ignored when
5426 verifying sender addresses in header lines. For example, when using
5428 verify = header_sender/callout=no_cache
5430 the cache was (incorrectly) being used.
5432 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5433 only the Cygwin environment.
5435 11. When return_path_on_delivery was set as a log selector, if different remote
5436 addresses in the same message used different return paths and parallel
5437 remote delivery occurred, the wrong values would sometimes be logged.
5438 (Whenever a remote delivery process finished, the return path value from
5439 the most recently started remote delivery process was logged.)
5441 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5442 lines when AUTH and/or TLS are in use. This is the "received protocol"
5443 field. Exim used to use "asmtp" for authenticated SMTP, without any
5444 indication (in the protocol name) for TLS use. Now it follows the RFC and
5445 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5446 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5447 names appear in log lines as well as in Received: header lines.
5449 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5450 extensions, and comparison tests, and to fix some bugs.
5452 14. Changes to the "personal" filter test:
5454 (1) The test was buggy in that it was just doing the equivalent of
5455 "contains" tests on header lines. For example, if a user's address was
5456 anne@some.where, the "personal" test would incorrectly be true for
5458 To: susanne@some.where
5460 This test is now done by extracting each address from the header in turn,
5461 and checking the entire address. Other tests that are part of "personal"
5462 are now done using regular expressions (for example, to check local parts
5463 of addresses in From: header lines).
5465 (2) The list of non-personal local parts in From: addresses has been
5466 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5467 taken from the Sieve specification recommendations.
5469 (3) If the message contains any header line starting with "List-" it is
5470 treated as non-personal.
5472 (4) The test for "circular" in the Subject: header line has been removed
5473 because it now seems ill-conceived.
5475 15. Minor typos in src/EDITME comments corrected.
5477 16. Installed latest exipick from John Jetmore.
5479 17. If headers_add on a router specified a text string that was too long for
5480 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5481 of string_sprintf() is now avoided.
5483 18. $message_body_size was not set (it was always zero) when running the DATA
5484 ACL and the local_scan() function.
5486 19. For the "mail" command in an Exim filter, no default was being set for
5487 the once_repeat time, causing a random time value to be used if "once" was
5488 specified. (If the value happened to be <= 0, no repeat happened.) The
5489 default is now 0s, meaning "never repeat". The "vacation" command was OK
5490 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5491 (I found it when inspecting the code).
5493 20. There is now an overall timeout for performing a callout verification. It
5494 defaults to 4 times the callout timeout, which applies to individual SMTP
5495 commands during the callout. The overall timeout applies when there is more
5496 than one host that can be tried. The timeout is checked before trying the
5497 next host. This prevents very long delays if there are a large number of
5498 hosts and all are timing out (e.g. when the network connections are timing
5499 out). The value of the overall timeout can be changed by specifying an
5500 additional sub-option for "callout", called "maxwait". For example:
5502 verify = sender/callout=5s,maxwait=20s
5504 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5505 to the end before writing, but this should make it even safer).
5507 22. Exim was forgetting that it had advertised PIPELINING for the second and
5508 subsequent messages on an SMTP connection. It was also not resetting its
5509 memory on STARTTLS and an internal HELO.
5511 23. When Exim logs an SMTP synchronization error within a session, it now
5512 records whether PIPELINING has been advertised or not.
5514 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5515 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5516 rather than long int.
5518 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5520 26. Added the never_mail option to autoreply.
5526 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5527 crash if the getsockname() call failed; this can happen if a connection is
5528 closed very soon after it is established. The problem was simply in the
5529 order in which certain operations were done, causing Exim to try to write
5530 to the SMTP stream before it had set up the file descriptor. The bug has
5531 been fixed by making things happen in the correct order.
5537 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5538 before the connection was closed, thus losing the rejection response.
5540 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5541 some early Solaris releases, but causes trouble in current releases where
5542 socklen_t is defined.
5544 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5547 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5550 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5551 that releasing the top of it at the end releases what was used for sub-
5552 expansions (unless the block got too big). However, discard this block if
5553 the first thing is a variable or header, so that we can use its block when
5554 it is dynamic (useful for very large $message_headers, for example).
5556 6. Lookups now cache *every* query, not just the most recent. A new, separate
5557 store pool is used for this. It can be recovered when all lookup caches are
5558 flushed. Lookups now release memory at the end of their result strings.
5559 This has involved some general refactoring of the lookup sources.
5561 7. Some code has been added to the store_xxx() functions to reduce the amount
5562 of flapping under certain conditions.
5564 8. log_incoming_interface used to affect only the <= reception log lines. Now
5565 it causes the local interface and port to be added to several more SMTP log
5566 lines, for example "SMTP connection from", and rejection lines.
5568 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5570 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5572 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5573 could be overwritten at the end of the current message (or the start of a
5574 new message if it was set in a HELO ACL). The value is now preserved for
5575 the duration of the SMTP connection.
5577 12. If a transport had a headers_rewrite setting, and a matching header line
5578 contained an unqualified address, that address was qualified, even if it
5579 did not match any rewriting rules. The underlying bug was that the values
5580 of the flags that permit the existence of unqualified sender and recipient
5581 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5582 non-local messages, and by -bnq for local messages) were not being
5583 preserved with the message after it was received.
5585 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5586 "next input=" as part of the text comprising the host identity instead of
5587 the correct text. The code was using the same buffer for two different
5588 strings. However, depending on which order the printing function evaluated
5589 its arguments, the bug did not always show up. Under Linux, for example, my
5590 test suite worked just fine.
5592 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5593 doesn't work with some older versions of Perl. It has been changed to "my",
5594 which in any case is probably the better facility to use.
5596 15. A really picky compiler found some instances of statements for creating
5597 error messages that either had too many or two few arguments for the format
5600 16. The size of the buffer for calls to the DNS resolver has been increased
5601 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5602 for addresses that have a lot of PTR records. This alleviates a problem; it
5603 does not fully solve it.
5605 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5606 buffer now truncates the list and logs the incident, which is the same
5607 action as happens when Exim is looking up a host name and its aliases.
5608 Previously in this situation something unpredictable would happen;
5609 sometimes it was "internal error: store_reset failed".
5611 18. If a server dropped the connection unexpectedly when an Exim client was
5612 using GnuTLS and trying to read a response, the client delivery process
5613 crashed while trying to generate an error log message.
5615 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5616 string, for example:
5618 warn message = H1: something\nH2: something
5620 the text was added as a single header line from Exim's point of view
5621 though it ended up OK in the delivered message. However, searching for the
5622 second and subsequent header lines using $h_h2: did not work. This has been
5623 fixed. Similarly, if a system filter added multiple headers in this way,
5624 the routers could not see them.
5626 20. Expanded the error message when iplsearch is called with an invalid key to
5627 suggest using net-iplsearch in a host list.
5629 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5630 ACLs are no longer actually imposed (and a message to that effect is
5633 22. If a "gecos" field in a passwd entry contained escaped characters, in
5634 particular, if it contained a \" sequence, Exim got it wrong when building
5635 a From: or a Sender: header from that name. A second bug also caused
5636 incorrect handling when an unquoted " was present following a character
5637 that needed quoting.
5639 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5640 was not being matched caselessly.
5642 24. Arranged for all hyphens in the exim.8 source to be escaped with
5645 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5646 properly. Recipient callout cache records were still being keyed to include
5647 the sender, even when use_sender was set false. This led to far more
5648 callouts that were necessary. The sender is no longer included in the key
5649 when use_sender is false.
5651 26. Added "control = submission" modifier to ACLs.
5653 27. Added the ${base62d: operator to decode base 62 numbers.
5655 28. dnsdb lookups can now access SRV records.
5657 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5658 the configuration file.
5660 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5661 (-v) mode. This makes the output for a verbose queue run more intelligible.
5663 31. Added a use_postmaster feature to recipient callouts.
5665 32. Added the $body_zerocount variable, containing the number of binary zero
5666 bytes in the message body.
5668 33. The time of last modification of the "new" subdirectory is now used as the
5669 "mailbox time last read" when there is a quota error for a maildir
5672 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5674 35. Added +ignore_unknown as a special item in host lists.
5676 36. Code for decoding IPv6 addresses in host lists is now included, even if
5677 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5678 address was recognized as an IP address, but was then not correctly decoded
5679 into binary, causing unexpected and incorrect effects when compared with
5686 1. Very minor rewording of debugging text in manualroute to say "list of
5687 hosts" instead of "hostlist".
5689 2. If verify=header_syntax was set, and a header line with an unqualified
5690 address (no domain) and a large number of spaces between the end of the
5691 name and the colon was received, the reception process suffered a buffer
5692 overflow, and (when I tested it) crashed. This was caused by some obsolete
5693 code that should have been removed. The fix is to remove it!
5695 3. When running in the test harness, delay a bit after writing a bounce
5696 message to get a bit more predictability in the log output.
5698 4. Added a call to search_tidyup() just before forking a reception process. In
5699 theory, someone could use a lookup in the expansion of smtp_accept_max_
5700 per_host which, without the tidyup, could leave open a database connection.
5702 5. Added the variables $recipient_data and $sender_data which get set from a
5703 lookup success in an ACL "recipients" or "senders" condition, or a router
5704 "senders" option, similar to $domain_data and $local_part_data.
5706 6. Moved the writing of debug_print from before to after the "senders" test
5709 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5710 problems for message scanning, either using a data ACL, or using
5711 local_scan() because the Received: header was not generated till after they
5712 were called (in order to set the time as the time of reception completion).
5713 I have revised the way this works. The header is now generated after the
5714 body is received, but before the ACL or local_scan() are called. After they
5715 are run, the timestamp in the header is updated.
5721 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5722 before starting a queue runner without re-exec. This happened only when
5723 deliver_drop_privilege was set or when the Exim user was set to root. The
5724 effect of the bug was that timeouts during subsequent deliveries caused
5725 crashes instead of being properly handled. The handler is now left at its
5726 default (and expected) setting.
5728 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5729 message, again when deliver_drop_privilege is set or Exim is run as root.
5730 The bug described in (1) was not present in this case, but the tidying up
5731 of the other signals was missing. I have made the two cases consistent.
5733 3. The ignore_target_hosts setting on a manualroute router was being ignored
5734 for hosts that were looked up using the /MX notation.
5736 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5739 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5740 operated on the sender address. After changing the $sender_address to <>
5741 for the sender address verify, Exim was re-instated it as the original
5742 (before rewriting) address, but remembering that it had rewritten it, so it
5743 wasn't rewriting it again. This bug also had the effect of breaking the
5744 sender address verification caching when the sender address was rewritten.
5746 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5747 This has been changed so that if the ip literal address matches
5748 ignore_target_hosts, the router declines.
5750 7. Added expansion conditions match_domain, match_address, and match_local_
5751 part (NOT match_host).
5753 8. The placeholder for the Received: header didn't have a length field set.
5755 9. Added code to Exim itself and to exim_lock to test for a specific race
5756 condition that could lead to file corruption when using MBX delivery. The
5757 issue is with the lockfile that is created in /tmp. If this file is removed
5758 after a process has opened it but before that process has acquired a lock,
5759 there is the potential for a second process to recreate the file and also
5760 acquire a lock. This could lead to two Exim processes writing to the file
5761 at the same time. The added code performs the same test as UW imapd; it
5762 checks after acquiring the lock that its file descriptor still refers to
5763 the same named file.
5765 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5766 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5769 11. Added the smtp_active_hostname option. If used, this will typically be made
5770 to depend on the incoming interface address. Because $interface_address is
5771 not set up until the daemon has forked a reception process, error responses
5772 that can happen earlier (such as "too many connections") no longer contain
5775 12. If an expansion in a condition on a "warn" statement fails because a lookup
5776 defers, the "warn" statement is abandoned, and the next ACL statement is
5777 processed. Previously this caused the whole ACL to be aborted.
5779 13. Added the iplsearch lookup type.
5781 14. Added ident_timeout as a log selector.
5783 15. Added tls_certificate_verified as a log selector.
5785 16. Added a global option tls_require_ciphers (compare the smtp transport
5786 option of the same name). This controls incoming TLS connections.
5788 17. I finally figured out how to make tls_require_ciphers do a similar thing
5789 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5790 before starting the TLS session.
5792 18. Tabs are now shown as \t in -bP output.
5794 19. If the log selector return_path_on_delivery was set, Exim crashed when
5795 bouncing a message because it had too many Received: header lines.
5797 20. If two routers both had headers_remove settings, and the first one included
5798 a superfluous trailing colon, the final name in the first list and the
5799 first name in the second list were incorrectly joined into one item (with a
5800 colon in the middle).
5806 1. Added -C and -D options to the exinext utility, mainly to make it easier
5807 to include in the automated testing, but these could be helpful when
5808 multiple configurations are in use.
5810 2. The exinext utility was not formatting the output nicely when there was
5811 an alternate port involved in the retry record key, nor when there was a
5812 message id as well (for retries that were specific to a specific message
5813 and a specific host). It was also confused by IPv6 addresses, because of
5814 the additional colons they contain. I have fixed the IPv4 problem, and
5815 patched it up to do a reasonable job for IPv6.
5817 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5818 delivery, the log line now contains "pipelined" if PIPELINING was used.
5820 4. An SMTP transport process used to panic and die if the bind() call to set
5821 an explicit outgoing interface failed. This has been changed; it is now
5822 treated in the same way as a connect() failure.
5824 5. A reference to $sender_host_name in the part of a conditional expansion
5825 that was being skipped was still causing a DNS lookup. This no longer
5828 6. The def: expansion condition was not recognizing references to header lines
5829 that used bh_ and bheader_.
5831 7. Added the _cache feature to named lists.
5833 8. The code for checking quota_filecount in the appendfile transport was
5834 allowing one more file than it should have been.
5836 9. For compatibility with Sendmail, the command line option
5844 and sets the incoming protocol and host name (for trusted callers). The
5845 host name and its colon can be omitted when only the protocol is to be set.
5846 Note the Exim already has two private options, -pd and -ps, that refer to
5847 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5848 "s", but I don't think that's a major issue.
5850 10. A number of refactoring changes to the code, none of which should affect
5853 (a) The number of logging options was getting close to filling up the
5854 32-bit word that was used as a bit map. I have split them into two classes:
5855 those that are passed in the argument to log_write(), and those that are
5856 only ever tested independently outside of that function. These are now in
5857 separate 32-bit words, so there is plenty of room for expansion again.
5858 There is no change in the user interface or the logging behaviour.
5860 (b) When building, for example, log lines, the code previously used a
5861 macro that called string_cat() twice, in order to add two strings. This is
5862 not really sufficiently general. Furthermore, there was one instance where
5863 it was actually wrong because one of the argument was used twice, and in
5864 one call a function was used. (As it happened, calling the function twice
5865 did not affect the overall behaviour.) The macro has been replaced by a
5866 function that can join an arbitrary number of extra strings onto a growing
5869 (c) The code for expansion conditions now uses a table and a binary chop
5870 instead of a serial search (which was left over from when there were very
5871 few conditions). Also, it now recognizes conditions like "pam" even when
5872 the relevant support is not compiled in: a suitably worded error message is
5873 given if an attempt is made to use such a condition.
5875 11. Added ${time_interval:xxxxx}.
5877 12. A bug was causing one of the ddress fields not to be passed back correctly
5878 from remote delivery subprocesses. The field in question was not being
5879 subsequently used, so this caused to problems in practice.
5881 13. Added new log selectors queue_time and deliver_time.
5883 14. Might have fixed a bug in maildirsizefile handling that threw up
5884 "unexpected character" debug warnings, and recalculated the data
5885 unnecessarily. In any case, I expanded the warning message to give more
5888 15. Added the message "Restricted characters in address" to the statements in
5889 the default ACL that block characters like @ and % in local parts.
5891 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5892 Three changes have been made:
5894 (a) There was a serious bug; a negative response to MAIL caused the whole
5895 recipient domain to be cached as invalid, thereby blocking all messages
5896 to all local parts at the same domain, from all senders. This bug has
5897 been fixed. The domain is no longer cached after a negative response to
5898 MAIL if the sender used is not empty.
5900 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5903 (c) A new callout option, "use_sender" has been added for people who want
5904 the modified behaviour.
5910 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5913 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5914 indeed breaks things for older releases.
5916 3. Added additional logging to the case where there is a problem reading data
5917 from a filter that is running in a subprocess using a pipe, in order to
5918 try to track down a specific problem.
5920 4. Testing facility fudge: when running in the test harness and attempting
5921 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5922 getting "No route to host". Convert this to a timeout.
5924 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5927 6. Some OS don't have socklen_t but use size_t instead. This affects the
5928 fifth argument of getsockopt() amongst other things. This is now
5929 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5930 can be set for individual OS. I have set it for SunOS5, OSF1, and
5931 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5932 some earlier ones do not.
5934 7. Change 4.30/15 was not doing the test caselessly.
5936 8. The standard form for an IPv6 address literal was being rejected by address
5937 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5938 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5939 this, as well as the form without the "IPv6" on the front (but only when
5940 address literals are enabled, of course).
5942 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5944 10. Exim crashed if a message with an empty sender address specified by -f
5945 encountered a router with an errors_to setting. This could be provoked only
5946 by a command such as
5950 where an empty string was supplied; "<>" did not hit this bug.
5952 11. Installed PCRE release 4.5.
5954 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5955 remained set. It is now erased.
5957 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5958 times from message ids (which are base 36 rather than the normal 62).
5960 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5961 were being counted as actual protocol errors, and logged if the log
5962 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5963 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5964 and DATA following a set of rejected RCPTs do not count as protocol errors.
5965 In other words, Exim assumes they were pipelined, though this may not
5966 actually be the case. Of course, in all cases the client gets an
5967 appropriate error code.
5969 15. If a lookup fails in an ACL condition, a message about the failure may
5970 be available; it is used if testing the ACL cannot continue, because most
5971 such messages specify what the cause of the deferral is. However, some
5972 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5973 that caused an old message to be retained and used if a later statement
5974 caused a defer, replacing the real cause of the deferral.
5976 16. If an IP address had so many PTR records that the DNS lookup buffer
5977 was not large enough to hold them, Exim could crash while trying to process
5978 the truncated data. It now detects and logs this case.
5980 17. Further to 4.21/58, another change has been made: if (and only if) the
5981 first line of a message (the first header line) ends with CRLF, a bare LF
5982 in a subsequent header line has a space inserted after it, so as not to
5983 terminate the header.
5985 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5986 unnecessarily, used atoi() instead of strtol(), and didn't check the
5987 termination when getting file sizes from file names by regex.
5989 19. Completely re-implemented the support for maildirsize files, in the light
5990 of a number of problems with the previous contributed implementation
5991 (4.30/29). In particular:
5993 . If the quota is zero, the maildirsize file is maintained, but no quota is
5996 . If the maildir directory does not exist, it is created before any attempt
5997 to write a maildirsize file.
5999 . The quota value in the file is just a cache; if the quota is changed in
6000 the transport, the new value overrides.
6002 . A regular expression is available for excluding directories from the
6005 20. The autoreply transport checks the characters in options that define the
6006 message's headers; it allows continued headers, but it was checking with
6007 isspace() after an embedded newline instead of explicitly looking for a
6010 21. If all the "regular" hosts to which an address was routed had passed their
6011 expiry times, and had not reached their retry times, the address was
6012 bounced, even if fallback hosts were defined. Now Exim should go on to try
6015 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6016 equivalent code in the SMTP transport. Some hosts send humungous responses
6017 to HELO/EHLO, more than 1024 it seems.
6019 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6020 gives compiler warnings in some environments. I've now done it "properly",
6023 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6024 (because of problems with the built-in one) was declared to return uschar *
6025 instead of char *, causing compiler failure.
6027 25. Fixed a file descriptor leak when processing alias/forward files.
6029 26. Fixed a minor format string issue in dbfn.c.
6031 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6033 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6034 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6037 29. When a sender address is verified, it is cached, to save repeating the test
6038 when there is more than one recipient in a message. However, when the
6039 verification involves a callout, it is possible for different callout
6040 options to be set for different recipients. It is too complicated to keep
6041 track of this in the cache, so now Exim always runs a verification when a
6042 callout is required, relying on the callout cache for the optimization.
6043 The overhead is duplication of the address routing, but this should not be
6046 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6047 to be verified with callout=postmaster, and the main callout worked but the
6048 postmaster check failed, the verification correctly failed. However, if a
6049 subsequent RCPT command asked for sender verification *without* the
6050 postmaster check, incorrect caching caused this verification also to fail,
6053 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6054 it was not caching the DNS options (qualify_single, search_parents) that
6055 were used when the lookup failed. A subsequent lookup with different
6056 options therefore always gave the same answer, though there were cases
6057 where it should not have. (Example: a "domains = !$mx_any" option on a
6058 dnslookup router: the "domains" option is always processed without any
6059 widening, but the router might have qualify_single set.) Now Exim uses the
6060 cached value only when the same options are set.
6062 32. Added John Jetmore's "exipick" utility to the distribution.
6064 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6065 than a timeout (e.g. a certificate is required, and is not provided), an
6066 Exim server now closes the connection immediately. Previously it waited for
6067 the client to close - but if the client is SSL, it seems that they each
6068 wait for each other, leading to a delay before one of them times out.
6070 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6071 maintained 0.8.x compatibility because I don't think many are using it, and
6072 it is clearly obsolete.
6074 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6077 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6078 client certificate was expired. A simple patch fixes this, though I don't
6079 understand the full logic of why the verify callback is called multiple
6082 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6083 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6084 which causes problems with some clients (such as the Certicom SSL Plus
6085 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6086 disables the coutermeasure allowing Eudora to connect."
6088 38. Exim was not checking that a write() to a log file succeeded. This could
6089 lead to Bad Things if a log got too big, in particular if it hit a file
6090 size limit. Exim now panics and dies if it cannot write to a log file, just
6091 as it does if it cannot open a log file.
6093 39. Modified OS/Makefile-Linux so that it now contains
6095 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6097 The two -D definitions ensure that Exim is compiled with large file
6098 support, which makes it possible to handle log files that are bigger than
6101 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6102 instance) a domain was checked against a named list that involved a lookup,
6103 causing $domain_data to be set, then another domain was checked against the
6104 same list, then the first domain was re-checked, the value of $domain_data
6105 after the final check could be wrong. In particular, if the second check
6106 failed, it could be set empty. This bug probably also applied to
6109 41. The strip_trailing_dot option was not being applied to the address given
6110 with the -f command-line option.
6112 42. The code for reading a message's header from the spool was incrementing
6113 $received_count, but never initializing it. This meant that the value was
6114 incorrect (doubled) while delivering a message in the same process in which
6115 it was received. In the most common configuration of Exim, this never
6116 happens - a fresh exec is done - but it can happen when
6117 deliver_drop_privilege is set.
6119 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6120 it now includes up to 150 characters of the unexpected data in the log
6123 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6124 and building data strings. The size of both of these buffers was 10 000
6125 bytes - far larger than anybody would *ever* want, thought I. Needless to
6126 say, somebody hit the limit. I have increased the maximum line length to
6127 20 000 and the maximum data length of concatenated lines to 100 000. I have
6128 also fixed two bugs, because there was no checking on these buffers. Tsk,
6129 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6130 buffer is too small.
6132 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6133 lsearch lookups. Now it does.
6135 46. When parsing a route_list item in a manualroute router, a fixed-length
6136 buffer was used for the list of hosts. I made this 1024 bytes long,
6137 thinking that nobody would ever have a list of hosts that long. Wrong.
6138 Somebody had a whole pile of complicated expansion conditions, and the
6139 string was silently truncated, leading to an expansion error. It turns out
6140 that it is easier to change to an unlimited length (owing to other changes
6141 that have happened since this code was originally written) than to build
6142 structure for giving a limitation error. The length of the item that
6143 expands into the list of hosts is now unlimited.
6145 47. The lsearch lookup could not handle data where the length of text line was
6146 more than 4095 characters. Such lines were truncated, leading to shortened
6147 data being returned. It should now handle lines of any length.
6149 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6150 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6153 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6154 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6155 no longer happens in comment lines. A list of replacements is now placed
6156 at the head of all of the source files, except those whose only change is
6157 to replace PERL_COMMAND in the very first #! line.
6159 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6160 messages on the queue, with a bottom-up merge sort, using code contributed
6161 by Michael Haardt. This should make operations like -bp somewhat faster on
6162 large queues. It won't affect queue runners, except when queue_run_in_order
6165 51. Installed eximstats 1.31 in the distribution.
6167 52. Added support for SRV lookups to the dnslookup router.
6169 53. If an ACL referred to $message_body or $message_body_end, the value was not
6170 reset for any messages that followed in the same SMTP session.
6172 54. The store-handling optimization for building very long strings was not
6173 differentiating between the different store pools. I don't think this
6174 actually made any difference in practice, but I've tidied it.
6176 55. While running the routers to verify a sender address, $sender_address
6177 was still set to the sender address. This is wrong, because when routing to
6178 send a bounce to the sender, it would be empty. Therefore, I have changed
6179 it so that, while verifying a sender address, $sender_address is set to <>.
6180 (There is no change to what happens when verifying a recipient address.)
6182 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6183 target A or AAAA records (if not already returned) without resetting the
6184 qualify_single or search_parents options of the DNS resolver. These are
6185 inappropriate in this case because the targets of MX and SRV records must
6186 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6187 target that, when qualified, matched something in the local domain. These
6188 two options are now turned off when doing these lookups.
6190 57. It seems that at least some releases of Reiserfs (which does not have the
6191 concept of a fixed number of inodes) returns zero and not -1 for the
6192 number of available inodes. This interacted badly with check_spool_inodes,
6193 which assumed that -1 was the "no such thing" setting. What I have done is
6194 to check that the total number of inodes is greater than zero before doing
6195 the test of how many are available.
6197 58. When a "warn" ACL statement has a log_message modifier, the message is
6198 remembered, and not repeated. This is to avoid a lot of repetition when a
6199 message has many recipients that cause the same warning to be written.
6200 However, Exim was preserving the list of already written lines for an
6201 entire SMTP session, which doesn't seem right. The memory is now reset if a
6202 new message is started.
6204 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6205 debugging output unless log_rewrite was also set.
6207 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6208 of (int)(handle) when we know that handle contains (void *)(-1).
6210 61. The Exim daemon panic-logs an error return when it closes the incoming
6211 connection. However "connection reset by peer" seems to be common, and
6212 isn't really an error worthy of noting specially, so that particular error
6215 62. When Exim is trying to find all the local interfaces, it used to panic and
6216 die if the ioctl to get the interface flags failed. However, it seems that
6217 on at least one OS (Solaris 9) it is possible to have an interface that is
6218 included in the list of interfaces, but for which you get a failure error
6219 for this call. This happens when the interface is not "plumbed" into a
6220 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6221 failure of the "get flags" call assumes that the interface is down.
6223 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6224 makes life easier for people who are doing arithmetic on fields extracted
6225 from dates, where you often get leading zeros that should not be
6226 interpreted as octal.
6228 64. Added qualify_domain to the redirect router, to override the global
6231 65. If a pathologically long header line contained very many addresses (the
6232 report of this problem mentioned 10 000) and each of them was rewritten,
6233 Exim could use up a very large amount of memory. (It kept on making new
6234 copies of the header line as it rewrote, and never released the old ones.)
6235 At the expense of a bit more processing, the header rewriting function has
6236 been changed so that it no longer eats memory in this way.
6238 66. The generation of the Received: header has been moved from the time that a
6239 message starts to be received, to the time that it finishes. The timestamp
6240 in the Received: header should now be very close to that of the <= log
6241 line. There are two side-effects of this change:
6243 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6244 logged header lines no longer include the local Received: line, because
6245 it has not yet been created. The same applies to a copy of the message
6246 that is returned to a non-SMTP sender when a message is rejected.
6248 (b) When a filter file is tested using -bf, no additional Received: header
6249 is added to the test message. After some thought, I decided that this
6252 This change does not affect the value of $received_for. It is still set
6253 after address rewriting, but before local_scan() is called.
6255 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6257 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6258 gave an unhelpful panic error message, and a defer error. I have managed to
6259 change this behaviour so that it now rejects any supplied certificate,
6260 which seems right, as the list of acceptable certificates is empty.
6262 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6263 gave an unhelpful defer error. I have not managed to make this reject any
6264 supplied certificates, but the error message it gives is "no certificate
6265 supplied", which is not helpful.
6267 70. exigrep's output now also includes lines that are not associated with any
6268 message, but which match the given pattern. Implemented by a patch from
6269 Martin Sluka, which also tidied up the Perl a bit.
6271 71. Recipient callout verification, like sender verification, was using <> in
6272 the MAIL FROM command. This isn't really the right thing, since the actual
6273 sender may affect whether the remote host accepts the recipient or not. I
6274 have changed it to use the actual sender in the callout; this means that
6275 the cache record is now keyed on a recipient/sender pair, not just the
6276 recipient address. There doesn't seem to be a real danger of callout loops,
6277 since a callout by the remote host to check the sender would use <>.
6278 [SEE ABOVE: changed after hitting problems.]
6280 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6281 temporary errors. However, in the case of such a code being given after
6282 the end of a data transmission (i.e. after ".") Exim was failing to write
6283 a retry record for the message. (Yes, there was some broken host that was
6284 actually sending 8xx at this point.)
6286 73. An unknown lookup type in a host list could cause Exim to panic-die when
6287 the list was checked. (An example that provoked this was putting <; in the
6288 middle of a list instead of at the start.) If this happened during a DATA
6289 ACL check, a -D file could be left lying around. This kind of configuration
6290 error no longer causes Exim to die; instead it causes a defer error. The
6291 incident is still logged to the main and panic logs.
6293 74. Buglet left over from Exim 3 conversion. The message "too many messages
6294 in one connection" was written to the rejectlog but not the mainlog, except
6295 when address rewriting (yes!) was being logged.
6297 75. Added write_rejectlog option.
6299 76. When a system filter was run not as root (that is, when system_filter_user
6300 was set), the values of the $n variables were not being returned to the
6301 main process; thus, they were not subsequently available in the $sn
6304 77. Added +return_path_on_delivery log selector.
6306 78. A connection timeout was being treated differently from recipients deferred
6307 when testing hosts_max_try with a message that was older than the host's
6308 retry timeout. (The host should not be counted, thus allowing all hosts to
6309 be tried at least once before bouncing.) This may have been the cause of an
6310 occasionally reported bug whereby a message would remain on the queue
6311 longer than the retry timeout, but would be bounced if a delivery was
6312 forced. I say "may" because I never totally pinned down the problem;
6313 setting up timeout/retry tests is difficult. See also the next item.
6315 79. The ultimate address timeout was not being applied to errors that involved
6316 a combination of host plus message (for example, a timeout on a MAIL
6317 command). When an address resolved to a number of possible hosts, and they
6318 were not all tried for each delivery (e.g. because of hosts_max_try), a
6319 message could remain on the queue longer than the retry timeout.
6321 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6324 81. Fixed an obscure SMTP outgoing bug which required at least the following
6325 conditions: (a) there was another message waiting for the same server;
6326 (b) the server returned 5xx to all RCPT commands in the first message so
6327 that the message was not completed; (c) the server dropped the connection
6328 or gave a negative response to the RSET that Exim sends to abort the
6329 transaction. The observed case was a dropped connection after DATA that had
6330 been sent in pipelining mode. That is, the server had advertised PIPELINING
6331 but was not implementing it correctly. The effect of the bug was incorrect
6332 behaviour, such as trying another host, and this could lead to a crash.
6338 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6339 and daemon.c were passed as pointers to ints; they should have been
6340 pointers to socklen_t variables (which are typically unsigned ints).
6342 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6345 3. Fixed a really odd bug that affected only the testing scheme; patching a
6346 certain fixed string in the binary changed the value of another string that
6347 happened to be identical to the end of the original first string.
6349 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6350 name", it returns that address as the IP address. On some operating
6351 systems (e.g. Solaris), it also passes back the IP address string as the
6352 "host name". However, on others (e.g. Linux), it passes back an empty
6353 string. Exim wasn't checking for this, and was changing the host name to an
6354 empty string, assuming it had been canonicalized.
6356 5. Although rare, it is permitted to have more than one PTR record for a given
6357 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6358 all the names associated with an address, because they do in Solaris.
6359 However, it seems that they do not in Linux for data that comes from the
6360 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6361 I found this out when I moved to a new Linux workstation and tried to run
6362 the Exim test suite.
6364 To get round this problem I have changed the code so that it now does its
6365 own call to the DNS to look up PTR records when searching for a host name.
6366 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6367 addresses that are only in /etc/hosts are still found.
6369 This behaviour is, however, controlled by an option called host_lookup_
6370 order, which defaults to "bydns:byaddr". If people want to use the other
6371 order, or indeed, just use one or the other means of lookup, they can
6372 specify it in this variable.
6374 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6375 some operating systems, this comes back from gethostbyaddr() as an empty
6376 string, and this is what Exim used to test for. However, it seems that in
6377 other systems, "." is yielded. Exim now tests for this case too.
6379 7. The values of check_spool_space and check_log_space are now held internally
6380 as a number of kilobytes instead of an absolute number of bytes. If a
6381 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6382 kilobyte. This means that much larger values can be stored.
6384 8. Exim monitor: an attempt to get the action menu when not actually pointing
6385 at a message produces an empty menu entitled "No message selected". This
6386 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6387 no entries in it ("Shell widget menu has zero width and/or height"). So I
6388 have added a single, blank menu entry in this case.
6390 9. Added ${quote_local_part.
6392 10. MIME decoding is now applied to the contents of Subject: header lines when
6395 11. Now that a reference to $sender_host_address automatically causes a reverse
6396 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6397 host lookup before query-style lookups in lists that might use this
6398 variable. This has therefore been abolished, and the "net-" prefix is no
6399 longer necessary for query-style lookups.
6401 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6402 have been a typo for LFLAGS, so it has been changed.
6404 13. The install script calls Exim with "-C /dev/null" in order to find the
6405 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6406 to be output. However, since Exim outputs its version number before the
6407 error, it didn't break the script. It just looked ugly. I fixed this by
6408 always allowing "-C /dev/null" if the caller is root.
6410 14. Ignore overlarge ACL variable number when reading spool file - insurance
6411 against a later release with more variables having written the file.
6413 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6414 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6415 this, as well as the form without the "IPv6" on the front.
6417 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6418 OS/Makefile-Darwin file.
6420 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6421 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6423 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6424 variable. However, some OS use ints for inodes. I've added cast to long int
6425 to get rid of the compiler warning.
6427 19. I had forgotten to lock out "/../" in configuration file names when
6428 ALT_CONFIG_PREFIX was set.
6430 20. Routers used for verification do not need to specify transports. However,
6431 if such a router generated a host list, and callout was configured, Exim
6432 crashed, because it could not find a port number from the (non-existent)
6433 transport. It now assumes port 25 in this circumstance.
6435 21. Added the -t option to exigrep.
6437 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6438 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6439 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6440 them is set, LOOKUP_LSEARCH is forced.
6442 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6444 24. Added sender and host information to the "rejected by local_scan()" log
6445 line; previously there was no indication of these.
6447 25. Added .include_if_exists.
6449 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6450 Linux. It turns out that not all file systems support this. Apparently some
6451 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6452 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6453 not supported on the file descriptor, is now ignored when Exim is trying to
6454 sync a directory. This applies only to Linux.
6456 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6458 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6459 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6462 29. Added support for maildirsize files from supplied patch (modified a bit).
6464 30. The use of :fail: followed by an empty string could lead Exim to respond to
6465 sender verification failures with (e.g.):
6467 550 Verification failed for <xxx>
6468 550 Sender verify failed
6470 where the first response line was missing the '-' that indicates it is not
6471 the final line of the response.
6473 31. The loop for finding the name of the user that called Exim had a hardwired
6474 limit of 10; it now uses the value of finduser_retries, which is used for
6475 all other user lookups.
6477 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6480 33. Exim was neglecting to zero errno before one call of strtol() when
6481 expanding a string and expecting an integer value. On some systems this
6482 resulted in spurious "integer overflow" errors. Also, it was casting the
6483 result into an int without checking.
6485 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6486 did not work. The code looks as if it has *never* worked, though it appears
6487 to have been documented since at least release 1.62. I have made it work.
6489 35. The "timeout_DNS" error in retry rules, also documented since at least
6490 1.62, also never worked. As it isn't clear exactly what this means, and
6491 clearly it isn't a major issue, I have abolished the feature by treating it
6492 as "timeout", and writing a warning to the main and panic logs.
6494 36. The display of retry rules for -brt wasn't always showing the error code
6497 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6498 timeout_connect_A, timeout_connect_MX.
6500 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6501 to the empty sender.
6503 39. The daemon was not analysing the content of -oX till after it had closed
6504 stderr and disconnected from the controlling terminal. This meant that any
6505 syntax errors were only noted on the panic log, and the return code from
6506 the command was 0. By re-arranging the code a little, I've made the
6507 decoding happen first, so such errors now appear on stderr, and the return
6508 code is 1. However, the actual setting up of the sockets still happens in
6509 the disconnected process, so errors there are still only recorded on the
6512 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6513 connections (as happens on some IP stacks) was logged at start up time as
6514 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6515 it from "IPv6 and IPv4", which means that two separate sockets are being
6518 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6519 says whether AF_INET or AF_INET6 was passed as an argument.
6521 42. Exiwhat output was messed up when time zones were included in log
6524 43. Exiwhat now gives more information about the daemon's listening ports,
6525 and whether -tls-on-connect was used.
6527 44. The "port" option of the smtp transport is now expanded.
6529 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6530 silently ignored. Now an error message is written to the main and panic
6533 46. There's a new ACL modifier called "logwrite" which writes to a log file
6534 as soon as it is encountered.
6536 47. Added $local_user_uid and $local_user_gid at routing time.
6538 48. Exim crashed when trying to verify a sender address that was being
6541 49. Exim was recognizing only a space character after ".include". It now also
6542 recognizes a tab character.
6544 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6545 extracting the relevant information from the specification. The man page no
6546 longer contains scrambled data for the -d option, and I've added a section
6547 at the front about calling Exim under different names.
6549 51. Added "extra_headers" argument to the "mail" command in filter files.
6551 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6554 53. Installed eximstats 1.29.
6556 54. Added transport_filter_timeout as a generic transport option.
6558 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6559 Cc: header lines. This was required by RFC 822, but it not required by RFC
6562 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6563 incoming messages that did not have them. Now it does so only if the
6564 message originates locally, that is, if there is no associated remote host
6565 address. When Resent- header lines are present, this applies to the Resent-
6566 lines rather than the non-Resent- lines.
6568 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6569 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6571 58. Messages for configuration errors now include the name of the main
6572 configuration file - useful now that there may be more than one file in a
6573 list (.included file names were always shown).
6575 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6576 for those rare installations that do not start the daemon as root or run it
6577 setuid root. I've cut out the call to initgroups() if the daemon is not
6580 60. The Exim user and group can now be bound into the binary as text strings
6581 that are looked up at the start of Exim's processing.
6583 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6585 62. Added $mailstore_basename variable.
6587 63. Installed patch to sieve.c from Michael Haardt.
6589 64. When Exim failed to open the panic log after failing to open the main log,
6590 the original message it was trying to log was written to stderr and debug
6591 output, but if they were not available (the usual case in production), it
6592 was lost. Now it is written to syslog before the two lines that record the
6593 failures to open the logs.
6595 65. Users' Exim filters run in subprocesses under the user's uid. It is
6596 possible for a "deliver" command or an alias in a "personal" command to
6597 provoke an address rewrite. If logging of address rewriting is configured,
6598 this fails because the process is not running as root or exim. There may be
6599 a better way of dealing with this, but for the moment (because 4.30 needs
6600 to be released), I have disabled address rewrite logging when running a
6601 filter in a non-root, non-exim process.
6607 1. The buildconfig auxiliary program wasn't quoting the value set for
6608 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6609 not defined. This bug was masked in 4.22 by the effect that was fixed in
6612 2. Some messages that were rejected after a message id was allocated were
6613 shown as "incomplete" by exigrep. It no longer does this for messages that
6614 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6616 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6617 have allow_domain_literals set, the ID did not get logged in the <= line.
6618 Domain literals are now always recognized in Message-ID: header lines.
6620 4. The first argument for a ${extract expansion item is the key name or field
6621 number. Leading and trailing spaces in this item were not being ignored,
6622 causing some misleading effects.
6624 5. When deliver_drop_privilege was set, single queue runner processes started
6625 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6626 same command in the process it spins off) were not dropping privilege.
6628 6. When the daemon running as "exim" started a queue runner, it always
6629 re-executed Exim in the spun-off process. This is a waste of effort when
6630 deliver_drop_privilege is set. The new process now just calls the
6631 queue-runner function directly.
6637 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6640 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6641 ignored. Though the use of -oP was forcing the writing of a pid file, it
6642 was always written to the default place.
6644 3. If the message "no IP address found for host xxxx" is generated during
6645 incoming verification, it is now followed by identification of the incoming
6646 connection (so you can more easily find what provoked it).
6648 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6650 5. Added some features to "harden" Exim a bit more against certain attacks:
6652 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6653 be put in Local/Makefile. This is like the never_users runtime option,
6654 but it cannot be overridden. The default setting is "root".
6656 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6657 prefix string with which any file named in a -C command line option
6660 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6661 is retained for -C and -D only if the caller of Exim is root. Without
6662 it, the exim user may also use -C and -D and retain privilege.
6664 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6665 command line option is disabled.
6667 6. Macro names set by the -D option must start with an upper case letter, just
6668 like macro names defined in the configuration file.
6670 7. Added "dereference=" facility to LDAP.
6672 8. Two instances of the typo "uknown" in the source files are fixed.
6674 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6675 the Configure-Makefile script screwed up while processing it.
6677 10. Incorporated PCRE 4.4.
6679 11. The SMTP synchronization check was not operating right at the start of an
6680 SMTP session. For example, it could not catch a HELO sent before the client
6681 waited for the greeting. There is now a check for outstanding input at the
6682 point when the greeting is written. Because of the duplex, asynchronous
6683 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6684 way, but not yet received, when the check is performed.
6686 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6687 on TCP/IP sockets, because this apparently causes some broken clients to
6690 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6691 unchanged) from the Cygwin maintainer.
6693 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6694 was supported instead of testing for mbx. Effectively a typo.
6696 15. The spa authenticator server code was not checking that the input it
6697 received was valid base64.
6699 16. The debug output line for the "set" modifier in ACLs was not showing the
6700 name of the variable that was being set.
6702 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6704 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6705 lookup on its own. Something else was needed to trigger the lookup. For
6706 example, a match in host_lookup or the need for a host name in a host list.
6707 Now, if $sender_host_name is referenced and the host name has not yet been
6708 looked up, a lookup is performed. If the lookup fails, the variable remains
6709 empty, and $host_lookup_failed is set to "1".
6711 19. Added "eqi" as a case-independent comparison operator.
6713 20. The saslauthd authentication condition could segfault if neither service
6714 nor realm was specified.
6716 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6717 error message that was logged was misleading, and incoming SMTP
6718 connections were dropped. The message is now more accurate, and temporary
6719 errors are given to SMTP connections.
6721 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6722 (except RSET) with a 421 error, until QUIT is received. However, it was
6723 failing to send a response to QUIT.
6725 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6726 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6727 ACL is now run after the helo_try_verify_hosts code.
6729 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6730 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6731 case-independent, so other case variants are also recognized). Apparently
6732 some systems use these upper case variants.
6734 25. If more than two messages were waiting for the same host, and a transport
6735 filter was specified for the transport, Exim sent two messages over the
6736 same TCP/IP connection, and then failed with "socket operation on non-
6737 socket" when it tried to send the third.
6739 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6741 27. The extern definition of crypt16() in expand.c was not being excluded when
6742 the OS had its own crypt16() function.
6744 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6745 as a preferred synonym for return_size_limit, both as an option and as an
6748 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6750 30. Changed the default configuration ACL to relax the local part checking rule
6751 for addresses that are not in any local domains. For these addresses,
6752 slashes and pipe symbols are allowed within local parts, but the sequence
6753 /../ is explicitly forbidden.
6755 31. SPA server authentication was not clearing the challenge buffer before
6758 32. log_message in a "warn" ACL statement was writing to the reject log as
6759 well as to the main log, which contradicts the documentation and doesn't
6760 seem right (because no rejection is happening). So I have stopped it.
6762 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6763 However, I am unable to do any testing of this.
6765 34. Fixed an infelicity in the appendfile transport. When checking directories
6766 for a mailbox, to see if any needed to be created, it was accidentally
6767 using path names with one or more superfluous leading slashes; tracing
6768 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6770 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6771 contents are added to the log line that is written for every discarded
6772 recipient. (Previously a log_message setting was ignored.)
6774 36. The ${quote: operator now quotes the string if it is empty.
6776 37. The install script runs exim in order to find its version number. If for
6777 some reason other than non-existence or emptiness, which it checks, it
6778 could not run './exim', it was installing it with an empty version number,
6779 i.e. as "exim-". This error state is now caught, and the installation is
6782 38. An argument was missing from the function that creates an error message
6783 when Exim fails to connect to the socket for saslauthd authentication.
6784 This could cause Exim to crash, or give a corrupted message.
6786 39. Added isip, isip4, and isip6 to ${if conditions.
6788 40. The ACL variables $acl_xx are now saved with the message, and can be
6789 accessed later in routers, transports, and filters.
6791 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6792 strings in the file are not string-expanded.
6794 42. If a MAIL command specified a SIZE value that was too large to fit into an
6795 int variable, the check against message_size_limit failed. Such values are
6796 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6797 day this will have to be increased, but I don't think I want to be around
6798 when emails are that large.
6805 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6806 iconv() is not standard in FreeBSD.
6808 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6809 IPv6 enabled. The observed symptom was a segmentation fault on return from
6810 the function os_common_find_running_interfaces() in src/os.c.
6812 3. In the check_special_case() function in daemon.c I had used "errno" as an
6813 argument name, which causes warnings on some systems. This was basically a
6814 typo, since it was named "eno" in the comments!
6816 4. The code that waits for the clock to tick (at a resolution of some fraction
6817 of a second) so as to ensure message-id uniqueness was always waiting for
6818 at least one whole tick, when it could have waited for less. [This is
6819 almost certainly not relevant at current processor speeds, where it is
6820 unlikely to ever wait at all. But we try to future-proof.]
6822 5. The function that sleeps for a time interval that includes fractions of a
6823 second contained a race. It did not block SIGALRM between setting the
6824 timer, and suspending (a couple of lines later). If the interval was short
6825 and the sigsuspend() was delayed until after it had expired, the suspension
6826 never ended. On busy systems this could lead to processes getting stuck for
6829 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6830 process, before it forks any delivery processes. The open lookup caching
6831 mechanism meant that the open file or database connection was passed into
6832 the delivery process. The problem was that delivery processes always tidy
6833 up cached lookup data. This could cause a problem for the next delivery
6834 process started by the queue runner, because the external queue runner
6835 process does not know about the closure. So the next delivery process
6836 still has data in the lookup cache. In the case of a file lookup, there was
6837 no problem because closing a file descriptor in a subprocess doesn't affect
6838 the parent. However, if the lookup was caching a connection to a database,
6839 the connection was closed, and the second delivery process was likely to
6840 see errors such as "PGSQL: query failed: server closed the connection
6841 unexpectedly". The problem has been fixed by closing all cached lookups
6842 in a queue runner before running a delivery process.
6844 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6845 seem to have the "const" qualifier which it has on other OS. I've
6848 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6849 *of the same type* (client or server) with the same public name that an
6850 error should be diagnosed.
6852 9. When Exim looked up a host name for an IP address, but failed to find the
6853 original IP address when looking up the host name (a safety check), it
6854 output the message "<ip address> does not match any IP for NULL", which was
6855 confusing, to say the least. The bug was that the host name should have
6856 appeared instead of "NULL".
6858 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6859 user that is built into the binary, and the -C or -D options is used, root
6860 privilege is dropped before the configuration file is read. In addition,
6861 logging is switched to stderr instead of the normal log files. If the
6862 configuration then re-defines the Exim user, the unprivileged environment
6863 is probably not what is expected, so Exim logs a panic warning message (but
6866 However, if deliver_drop_privilege is set, the unprivileged state may well
6867 be exactly what is intended, so the warning has been cut out in that case,
6868 and Exim is allowed to try to write to its normal log files.
6874 1. smtp_return_error_details was not giving details for temporary sender
6875 or receiver verification errors.
6877 2. Diagnose a configuration error if two authenticators have the same public
6880 3. Exim used not to create the message log file for a message until the first
6881 delivery attempt. This could be confusing when incoming messages were held
6882 for policy or load reasons. The message log file is now created at the time
6883 the message is received, and an initial "Received" line is written to it.
6885 4. The automatically generated man page for command line options had a minor
6886 bug that caused no ill effects; however, a more serious problem was that
6887 the procedure for building the man page automatically didn't always
6888 operate. Consequently, release 4.20 contains an out-of-date version. This
6889 shouldn't happen again.
6891 5. When building Exim with embedded Perl support, the script that builds the
6892 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6893 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6895 6. The freeze_tell option was not being used for messages that were frozen on
6896 arrival, either by an ACL or by local_scan().
6898 7. Added the smtp_incomplete_transaction log selector.
6900 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6901 was accepting AUTH without a new EHLO.
6903 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6904 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6907 10. Logging of TCP/IP connections (when configured) now happens in the main
6908 daemon process instead of the child process, so that the TCP/IP connection
6909 count is more accurate (but it can never be perfect).
6911 11. The use of "drop" in a nested ACL was not being handled correctly in the
6912 outer ACL. Now, if condition failure induced by the nested "drop" causes
6913 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6914 or "require"), the connection is dropped.
6916 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6917 that yield "discard" can now be used with an "accept" or a "discard" verb,
6918 but an error is generated for any others (because I can't see a useful way
6919 to define what should happen).
6921 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6922 are now processed in the same way as lines in the Exim configuration file.
6923 In particular, continuation lines are supported.
6925 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6927 15. Added -ti meaning -t -i.
6929 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6930 domains, and warn by logging if others are found.
6932 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6933 returned from GIFCONF when Exim is trying to find the list of interfaces on
6934 a host. The code in os.c has been modified to copy each ifreq to an aligned
6935 structure in all cases.
6937 Also, in some cases, the returned ifreq's were being copied to a 'struct
6938 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6939 means the last couple of bytes of an IPv6 address could be chopped if the
6940 ifreq contained only a normal sockaddr (14 bytes storage).
6942 18. Named domain lists were not supported in the hosts_treat_as_local option.
6943 An entry such as +xxxx was not recognized, and was treated as a literal
6946 19. Ensure that header lines added by a DATA ACL are included in the reject log
6947 if the ACL subsequently rejects the message.
6949 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6950 MD5 (which is deprecated).
6952 21. When testing a filter file using -bf, Exim was writing a message when it
6953 took the sender from a "From " line in the message, but it was not doing so
6954 when it took $return_path from a Return-Path: header line. It now does.
6956 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6957 with a valid header line field name (a series of printing characters
6958 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6960 23. Changed "disc" in the source to "disk" to conform to the documentation and
6961 the book and for uniformity.
6963 24. Ignore Sendmail's -Ooption=value command line item.
6965 25. When execve() failed while trying to run a command in a pipe transport,
6966 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6967 could be confused with a return value of 69 from the command itself. This
6968 has been changed to 127, the value the shell returns if it is asked to run
6969 a non-existent command. The wording for the related log line suggests a
6970 non-existent command as the problem.
6972 26. If received_header_text expands to an empty string, do not add a Received:
6973 header line to the message. (Well, it adds a token one on the spool, but
6974 marks it "old" so that it doesn't get used or transmitted.)
6976 27. Installed eximstats 1.28 (addition of -nt option).
6978 28. There was no check for failure on the call to getsockname() in the daemon
6979 code. This can fail if there is a shortage of resources on the system, with
6980 ENOMEM, for example. A temporary error is now given on failure.
6982 29. Contrary to the C standard, it seems that in some environments, the
6983 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6984 program. Exim now does this explicitly; it affects the formatting of
6985 timestamps using strftime().
6987 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6988 complaints. I've now initialized all the variables, to avoid this.
6990 32. Header lines added by a system filter were not being "seen" during
6991 transport-time rewrites.
6993 33. The info_callback() function passed to OpenSSL is set up with type void
6994 (*)(SSL *, int, int), as described somewhere. However, when calling the
6995 function (actually a macro) that sets it up, the type void(*)() is
6996 expected. I've put in a cast to prevent warnings from picky compilers.
6998 34. If a DNS black list lookup found a CNAME record, but there were no A
6999 records associated with the domain it pointed at, Exim crashed.
7001 35. If a DNS black list lookup returned more than one A record, Exim ignored
7002 all but the first. It now scans all returned addresses if a particular IP
7003 value is being sought. In this situation, the contents of the
7004 $dnslist_value variable are a list of all the addresses, separated by a
7007 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7008 to accept a host name and all its aliases if the forward lookup for any of
7009 them yielded the IP address of the incoming connection. Now it accepts only
7010 those names whose forward lookup yields the correct IP address. Any other
7011 names are discarded. This closes a loophole whereby a rogue DNS
7012 administrator could create reverse DNS records to break through a
7013 wildcarded host restriction in an ACL.
7015 37. If a user filter or a system filter that ran in a subprocess used any of
7016 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7017 the wrong values were passed to the pipe command ($thisaddress had the
7018 value of $0, $0 had the value of $1, etc). This bug was introduced by
7019 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7021 38. Improved the line breaking for long SMTP error messages from ACLs.
7022 Previously, if there was no break point between 40 and 75 characters, Exim
7023 left the rest of the message alone. Two changes have been made: (a) I've
7024 reduced the minimum length to 35 characters; (b) if it can't find a break
7025 point between 35 and 75 characters, it looks ahead and uses the first one
7026 that it finds. This may give the occasional overlong line, but at least the
7027 remaining text gets split now.
7029 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7030 file descriptors might be low, and that setting 1000 would always raise it.
7031 It turns out that in some environments, the limit is already over 1000 and
7032 that lowering it causes trouble. So now Exim takes care not to decrease it.
7034 40. When delivering a message, the value of $return_path is set to $sender_
7035 address at the start of routing (routers may change the value). By an
7036 oversight, this default was not being set up when an address was tested by
7037 -bt or -bv, which affected the outcome if any router or filter referred to
7040 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7041 log only when "message" or "log_message" are set. However, if one of the
7042 conditions was an address verification, or a call to a nested ACL, the
7043 messages generated by the underlying test were being passed through. This
7044 no longer happens. The underlying message is available in $acl_verify_
7045 message for both "message" and "log_message" expansions, so it can be
7046 passed through if needed.
7048 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7049 new expansion $bh_ to give the encoded byte string without charset
7050 translation. Translation happens only if iconv() is available; HAVE_ICONV
7051 indicates this at build time. HEADERS_CHARSET gives the charset to
7052 translate to; headers_charset can change it in the configuration, and
7053 "headers charset" can change it in an individual filter file.
7055 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7056 that creates RFC 2047 encoded "words" labels them as that charset instead
7057 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7058 expansion operator; (ii) when Exim creates a From: line for a local
7059 message; (iii) when a header line is rewritten to include a "phrase" part.
7061 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7062 buggy, causing it to skip the first lines of messages whose message ID
7063 ended in 'D'. This would not have bitten before Exim release 4.14, because
7064 message IDs were unlikely to end in 'D' before then. The effect was to have
7065 incorrect size information for certain domains.
7067 45. #include "config.h" was missing at the start of the crypt16.c module. This
7068 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7071 46. If there was a timeout during a "random" callout check, Exim treated it as
7072 a failure of the random address, and carried on sending RSET and the real
7073 address. If the delay was just some slowness somewhere, the response to the
7074 original RCPT would be taken as a response to RSET and so on, causing
7075 mayhem of various kinds.
7077 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7078 when I implemented it. It didn't allow for the fact that some option values
7079 may legitimately be negative (e.g. size_addition), and it didn't even do
7080 the right test for positive values.
7082 48. Domain names in DNS records are case-independent. Exim always looks them up
7083 in lower case. Some resolvers return domain names in exactly the case they
7084 appear in the zone file, that is, they may contain uppercase letters. Not
7085 all resolvers do this - some return always lower case. Exim was treating a
7086 change of case by a resolver as a change of domain, similar to a widening
7087 of a domain abbreviation. This triggered its re-routing code and so it was
7088 trying to route what was effectively the same domain again. This normally
7089 caused routing to fail (because the router wouldn't handle the domain
7090 twice). Now Exim checks for this case specially, and just changes the
7091 casing of the domain that it ultimately uses when it transmits the message
7094 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7097 50. If a filter generated a file delivery with a non-absolute name (possible if
7098 no home directory exists for the router), the forbid_file option was not
7101 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7102 the existing equality matching.
7104 52. Exim was using ints instead of ino_t variables in some places where it was
7105 dealing with inode numbers.
7107 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7108 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7109 called TMPDIR, and if it finds it is different, it changes its value.
7111 54. The smtp_printf() function is now made available to local_scan() so
7112 additional output lines can be written before returning. There is also an
7113 smtp_fflush() function to enable the detection of a dropped connection.
7114 The variables smtp_input and smtp_batched_input are exported to
7117 55. Changed the default runtime configuration: the message "Unknown user"
7118 has been removed from the ACL, and instead placed on the localuser router,
7119 using the cannot_route_message feature. This means that any verification
7120 failures that generate their own messages won't get overridden. Similarly,
7121 the "Unrouteable address" message that was in the ACL for unverifiable
7122 relay addresses has also been removed.
7124 56. Added hosts_avoid_esmtp to the smtp transport.
7126 57. The exicyclog script was not checking for the esoteric option
7127 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7128 will work only if exicyclog is run under the appropriate euid.
7130 58. Following a discussion on the list, the rules by which Exim recognises line
7131 endings on incoming messages have been changed. The -dropcr and drop_cr
7132 options are now no-ops, retained only for backwards compatibility. The
7133 following line terminators are recognized: LF CRLF CR. However, special
7134 processing applies to CR:
7136 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7137 nor a local message in the state where . is a terminator.
7139 (ii) If a bare CR is encountered in a header line, an extra space is added
7140 after the line terminator so as not to end the header. The reasoning
7141 behind this is that bare CRs in header lines are most likely either
7142 to be mistakes, or people trying to play silly games.
7144 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7145 was being incorrectly given as 18 bytes larger than it should have been.
7146 This is a VOB (very old bug).
7148 60. This may never have affected anything current, but just in case it has:
7149 When the local host is found other than at the start of a list of hosts,
7150 the local host, those with the same MX, and any that follow, are discarded.
7151 When the list in question was part of a longer list of hosts, the following
7152 hosts (not currently being processed) were also being discarded. This no
7153 longer happens. I'm not sure if this situation could ever has previously
7156 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7159 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7160 header. This is something that is recommended in a new Internet Draft, and
7161 is something that is documented as being done by Sendmail. There are two
7162 possible values. For messages generated by the autoreply transport, Exim
7165 Auto-Submitted: auto-replied
7167 whereas for all other generated messages (e.g. bounces) it adds
7169 Auto-Submitted: auto-generated
7171 63. The "personal" condition in filters now includes a test for the
7172 Auto-Submitted: header. If it contains the string "auto-" the message it
7173 not considered personal.
7175 64. Added rcpt_include_affixes as a generic transport option.
7177 65. Added queue_only_override (default true).
7179 66. Added the syslog_duplication option.
7181 67. If what should have been the first header line of a message consisted of
7182 a space followed by a colon, Exim was mis-interpreting it as a header line.
7183 It isn't of course - it is syntactically invalid and should therefore be
7184 treated as the start of the message body. The misbehaviour could have
7185 caused a number of strange effects, including loss of data in subsequent
7186 header lines, and spool format errors.
7188 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7189 client host had authenticated. This control can now be exercised by an ACL
7190 for more flexibility.
7192 69. By default, callouts do not happen when testing with -bh. There is now a
7193 variant, -bhc, which does actually run the callout code, including
7194 consulting and updating the callout cache.
7196 70. Added support for saslauthd authentication, courtesy of Alexander
7199 71. If statvfs() failed on the spool or log directories while checking their
7200 size for availability, Exim confusingly gave the error "space shortage".
7201 Furthermore, in debugging mode it crashed with a floating point exception.
7202 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7203 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7204 serious problem, Exim now writes to the main and panic logs when this
7205 happens, with details of the failure. It then refuses to accept the
7206 incoming message, giving the message "spool directory problem" or "log
7207 directory problem" with a 421 code for SMTP messages.
7209 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7210 0, 1, and 2 exist, because some OS complain for execs without them (see
7211 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7212 descriptors. However, the code omitted to check that the open succeeded,
7213 causing mysterious errors if for some reason the permissions on /dev/null
7214 got screwed. Now Exim writes a message to the main and panic logs, and
7215 bombs out if it can't open /dev/null.
7217 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7218 interact so that it is all more flexible. It is supposed to remain
7219 backwards compatible. Also added extra_local_interfaces.
7221 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7222 to bomb out with an assertion failure - to the client this appears as a
7223 connection drop. This problem occurs in the part of the code that was taken
7224 from the Samba project. Fortunately, the assertion is in a very simple
7225 function, so I have fixed this by reproducing the function inline in the
7226 one place where it is called, and arranging for authentication to fail
7227 instead of killing the process with assert().
7229 75. The SPA client code was not working when the server requested OEM rather
7230 than Unicode encoding.
7232 76. Added code to make require_files with a specific uid setting more usable in
7233 the case where statting the file as root fails - usually a non-root-mounted
7234 NFS file system. When this happens and the failure is EACCES, Exim now
7235 forks a subprocess and does the per-uid checking as the relevant uid.
7237 77. Added process_log_path.
7239 78. If log_file_path was not explicitly set, a setting of check_log_space or
7240 check_log_inodes was ignored.
7242 79. If a space check for the spool or log partitions fails, the incident is now
7243 logged. Of course, in the latter case the data may get lost...
7245 80. Added the %p formatting code to string_format() so that it can be used to
7246 print addresses in debug_print(). Adjusted all the address printing in the
7247 debugging in store.c to use %p rather than %d.
7249 81. There was a concern that a line of code in smtp_in.c could overflow a
7250 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7251 initially expressed, the concern was not well-founded, because trailing
7252 spaces are removed early. However, if the trailing spaces were followed by
7253 a NULL, they did not get removed, so the overflow was possible. Two fixes
7256 (a) I re-wrote the offending code in a cleaner fashion.
7257 (b) If an incoming SMTP command contains a NULL character, it is rejected
7260 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7261 runs initgroups(), so that if the Exim user is in any additional groups,
7262 they will be used during message reception.
7268 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob