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.
92 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
93 When redis_servers is set to a list of > 1 element, and the Redis servers
94 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
95 case of MOVED into a DEFER case instead, thus moving the query onto the
96 next server in the list. For a cluster of N elements, all N servers must
97 be defined in redis_servers.
99 GF/02 Catch and remove uninitialized value warning in exiqsumm
100 Check for existence of @ARGV before looking at $ARGV[0]
102 JH/01 Replace the store_release() internal interface with store_newblock(),
103 which internalises the check required to safely use the old one, plus
104 the allocate and data copy operations duplicated in both (!) of the
105 extant use locations.
107 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
108 modifier. This matches the restriction on the commandline.
110 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
111 Previously only the last row was returned.
113 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
114 we assumed that tags in the header were well-formed, and parsed the
115 element content after inspecting only the first char of the tag.
116 Assumptions at that stage could crash the receive process on malformed
119 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
120 While running the DKIM ACL we operate on the Permanent memory pool so that
121 variables created with "set" persist to the DATA ACL. Also (at any time)
122 DNS lookups that fail create cache records using the Permanent pool. But
123 expansions release any allocations made on the current pool - so a dnsdb
124 lookup expansion done in the DKIM ACL releases the memory used for the
125 DNS negative-cache, and bad things result. Solution is to switch to the
126 Main pool for expansions.
127 While we're in that code, add checks on the DNS cache during store_reset,
128 active in the testsuite.
129 Problem spotted, and debugging aided, by Wolfgang Breyha.
131 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
132 When none of the hosts presented to a transport match an already-open
133 connection, close it and proceed with the list. Previously we would
134 queue the message. Spotted by Lena with Yahoo, probably involving
137 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
138 Previously a spurious "250 OK id=" response was appended to the proper
141 JH/08 The "support for" informational output now, which built with Content
142 Scanning support, has a line for the malware scanner interfaces compiled
143 in. Interface can be individually included or not at build time.
145 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
146 by the template makefile "src/EDITME". The "STREAM" support for an older
147 ClamAV interface method is removed.
149 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
150 rows affected is given instead).
152 JH/11 The runtime Berkeley DB library version is now additionally output by
153 "exim -d -bV". Previously only the compile-time version was shown.
155 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
156 SMTP connection. Previously, when one had more receipients than the
157 first, an abortive onward connection was made. Move to full support for
158 multiple onward connections in sequence, handling cutthrough connection
159 for all multi-message initiating connections.
161 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
162 routers. Previously, a multi-recipient message would fail to match the
163 onward-connection opened for the first recipient, and cause its closure.
165 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
166 a timeout on read on a GnuTLS initiating connection, resulting in the
167 initiating connection being dropped. This mattered most when the callout
168 was marked defer_ok. Fix to keep the two timeout-detection methods
171 JH/15 Relax results from ACL control request to enable cutthrough, in
172 unsupported situations, from error to silently (except under debug)
173 ignoring. This covers use with PRDR, frozen messages, queue-only and
176 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
178 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
179 metadata, resulting in a crash in free().
181 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
182 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
183 Broken also in d185889f4, with init system revamp.
185 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
186 Previously we abruptly closed the connection after reading a malware-
187 found indication; now we go on to read the "scan ok" response line,
190 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
191 ACL. Previously, a crash would result.
193 JH/19 Speed up macro lookups during configuration file read, by skipping non-
194 macro text after a replacement (previously it was only once per line) and
195 by skipping builtin macros when searching for an uppercase lead character.
197 JH/20 DANE support moved from Experimental to mainline. The Makefile control
198 for the build is renamed.
200 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
201 was allocated for every new TLS startup, meaning one per message. Fix
202 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
204 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
205 reported the original. Fix to report (as far as possible) the ACL
206 result replacing the original.
208 JH/23 Fix memory leak during multi-message connections using STARTTLS under
209 OpenSSL. Certificate information is loaded for every new TLS startup,
210 and the resources needed to be freed.
212 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
214 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
217 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
218 DATA response info to the (existing) per-recipient response info for
219 the "C=" log element. It can have useful tracking info from the
220 destination system. Patch from Simon Arlott.
222 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
223 length value. Previously this would segfault.
225 HS/02 Support Avast multiline protoocol, this allows passing flags to
226 newer versions of the scanner.
228 JH/28 Ensure that variables possibly set during message acceptance are marked
229 dead before release of memory in the daemon loop. This stops complaints
230 about them when the debug_store option is enabled. Discovered specifically
231 for sender_rate_period, but applies to a whole set of variables.
232 Do the same for the queue-runner and queue-list loops, for variables set
233 from spool message files. Do the same for the SMTP per-message loop, for
234 certain variables indirectly set in ACL operations.
236 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
237 as a multi-recipient message from a mailinglist manager). The coding had
238 an arbitrary cutoff number of characters while checking for more input;
239 enforced by writing a NUL into the buffer. This corrupted long / fast
240 input. The problem was exposed more widely when more pipelineing of SMTP
241 responses was introduced, and one Exim system was feeding another.
242 The symptom is log complaints of SMTP syntax error (NUL chars) on the
243 receiving system, and refused recipients seen by the sending system
244 (propating to people being dropped from mailing lists).
245 Discovered and pinpointed by David Carter.
247 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
248 replaced by the ${authresults } expansion.
250 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
252 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
253 allows proper process termination in container environments.
255 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
256 Previously the "final dot" had a newline after it; ensure it is CR,LF.
258 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
259 and "err_perm", deprecated since 4.83 when the RFC-defined words
260 "temperror" and "permerror" were introduced.
262 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
263 transport-filters or DKIM-signing. The restriction was lost in the
264 consolidation of verify-callout and delivery SMTP handling.
265 Extend the restriction to also cover ARC-signing.
267 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
268 in defer=pass mode supply a 450 to the initiator. Previously the message
271 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
272 tls_require_ciphers is used as before.
274 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
275 "pass_unscanned". Only tmpfails from the scanner are written to
276 the paniclog, as they may require admin intervention (permission
277 denied, license issues). Other scanner errors (like decompression
278 bombs) do not cause a paniclog entry.
280 JH/36 Fix reinitialisation of DKIM logging variable between messages.
281 Previously it was possible to log spurious information in receive log
284 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
285 triggered odd behaviour from Outlook Express clients.
287 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
290 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
291 since the IETF WG has not yet settled on that versus the original
292 "bare" representation.
294 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
295 Previously the millisecond value corrupted the output.
296 Fix also for syslog_pid=no and log_selector +pid, for which the pid
297 corrupted the output.
303 JH/01 Rework error string handling in TLS interface so that the caller in
304 more cases is responsible for logging. This permits library-sourced
305 string to be attached to addresses during delivery, and collapses
306 pairs of long lines into single ones.
308 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
309 during configuration. Wildcards are allowed and expanded.
311 JH/02 Rework error string handling in DKIM to pass more info back to callers.
312 This permits better logging.
314 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
315 do not close it down and have the child transport start it up again on
316 the passed-on TCP connection. Instead, proxy the child (and any
317 subsequent ones) for TLS via a unix-domain socket channel. Logging is
318 affected: the continued delivery log lines do not have any DNSSEC, TLS
319 Certificate or OCSP information. TLS cipher information is still logged.
321 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
322 identical IP addresses on different listening ports. Will also affect
325 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
326 add noisy ifdef guards to special-case this sillyness.
327 Patch from Bernd Kuhls.
329 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
330 than 255 are no longer allowed.
332 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
333 Disable the session-cache too, which might reduce our load. Since we
334 currrectly use a new context for every connection, both as server and
335 client, there is no benefit for these.
336 GnuTLS appears to not support tickets server-side by default (we don't
337 call gnutls_session_ticket_enable_server()) but client side is enabled
338 by default on recent versions (3.1.3 +) unless the PFS priority string
341 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
342 <https://reproducible-builds.org/specs/source-date-epoch/>.
344 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
345 the check for any unsuccessful recipients did not notice the limit, and
346 erroneously found still-pending ones.
348 JH/08 Pipeline CHUNKING command and data together, on kernels that support
349 MSG_MORE. Only in-clear (not on TLS connections).
351 JH/09 Avoid using a temporary file during transport using dkim. Unless a
352 transport-filter is involved we can buffer the headers in memory for
353 creating the signature, and read the spool data file once for the
354 signature and again for transmission.
356 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
357 4.77 as the kernel support then wasn't solid, having issues in 64bit
358 mode. Now, it's been long enough. Add support for FreeBSD also.
360 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
361 case where the routing stage had gathered several addresses to send to
362 a host before calling the transport for the first, we previously failed
363 to close down TLS in the old transport process before passing the TCP
364 connection to the new process. The new one sent a STARTTLS command
365 which naturally failed, giving a failed delivery and bloating the retry
366 database. Investigation and fix prototype from Wolfgang Breyha.
368 JH/12 Fix check on SMTP command input synchronisation. Previously there were
369 false-negatives in the check that the sender had not preempted a response
370 or prompt from Exim (running as a server), due to that code's lack of
371 awareness of the SMTP input buffering.
373 PP/04 Add commandline_checks_require_admin option.
374 Exim drops privileges sanely, various checks such as -be aren't a
375 security problem, as long as you trust local users with access to their
376 own account. When invoked by services which pass untrusted data to
377 Exim, this might be an issue. Set this option in main configuration
378 AND make fixes to the calling application, such as using `--` to stop
381 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
382 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
383 responses to those, into a single TLS record each way (this usually means
384 a single packet). As a side issue, smtp_enforce_sync now works on TLS
387 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
388 affects you only if you're dancing at the edge of the param size limits.
389 If you are, and this message makes sense to you, then: raise the
390 configured limit or use OpenSSL 1.1. Nothing we can do for older
393 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
394 cmdline element to get the documented default one. Previously it was
397 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
398 are unsuable for non-message-specific reasons - eg. connection timeout,
399 banner-time rejection.
401 JH/16 Fix logging of delivery remote port, when specified by router, under
404 PP/06 Repair manualroute's ability to take options in any order, even if one
405 is the name of a transport.
408 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
410 JH/17 Change the list-building routines interface to use the expanding-string
411 triplet model, for better allocation and copying behaviour.
413 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
414 Previously it was constructed the first time a possibly-matching string
415 was met in the configuration file input during startup; now it is done
418 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
419 compatible one, to avoid the (poorly documented) possibility of a config
420 file in the working directory redirecting the DB files, possibly correpting
421 some existing file. CVE-2017-10140 assigned for BDB.
423 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
424 cache-hot. Previously, although the result was properly cached, the
425 initial verify call returned a defer.
427 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
428 the main verify for receipient in uncached-mode.
430 JH/22 Retire historical build files to an "unsupported" subdir. These are
431 defined as "ones for which we have no current evidence of testing".
433 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
434 if present. Previously it was ignored.
436 JH/24 Start using specified-initialisers in C structure init coding. This is
437 a C99 feature (it's 2017, so now considered safe).
439 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
440 if was a fixed-sized field and bitmask ops via macros; it is now more
443 PP/07 GitHub PR 56: Apply MariaDB build fix.
444 Patch provided by Jaroslav Škarvada.
446 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
447 during Coverity cleanups [4.87 JH/47]
448 Diagnosis and fix provided by Michael Fischer v. Mollard.
450 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
451 the right size to place the terminating semicolon on its own folded
452 line, the header hash was calculated to an incorrect value thanks to
453 the (relaxed) space the fold became.
455 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
456 and confused the parent.
458 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
459 which could crash as a result. This could lead to undeliverable messages.
461 JH/28 Logging: "next input sent too soon" now shows where input was truncated
464 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
465 matters on fast-turnover and PID-randomising systems, which were getting
466 out-of-order delivery.
468 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
469 a possibly-overlapping copy. The symptom was that "Remote host closed
470 connection in response to HELO" was logged instead of the actual 4xx
473 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
474 Previously only that bufferd was discarded, resulting in SYMTP command
477 JH/32 DKIM: when a message has multiple signatures matching an identity given
478 in dkim_verify_signers, run the dkim acl once for each. Previously only
479 one run was done. Bug 2189.
481 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
482 "panic the current process" to "deliberately defer". The panic log is
483 still written with the problem list name; the mail and reject logs now
484 get a temp-reject line for the message that was being handled, saying
485 something like "domains check lookup or other defer". The SMTP 451
486 message is still "Temporary local problem".
488 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
489 A crafted sequence of BDAT commands could result in in-use memory beeing
490 freed. CVE-2017-16943.
492 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
493 from SMTP input. Previously it was always done; now only done for DATA
494 and not BDAT commands. CVE-2017-16944.
496 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
497 to the message (such as an overlong header line). Previously this was
498 not done and we did not exit BDAT mode. Followon from the previous item
499 though a different problem.
505 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
506 than -2003 did; needs libidn2 in addition to libidn.
508 JH/02 The path option on a pipe transport is now expanded before use.
510 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
511 Patch provided by "Björn", documentation fix added too.
513 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
514 missing a wire-to-host endian conversion.
516 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
517 close after a BDAT command line could be taken as a following command,
518 giving a synch failure. Fix by only checking for synch immediately
519 before acknowledging the chunk.
521 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
522 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
523 macro. Patches provided by Josh Soref.
525 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
526 Previously we did not; the RFC seems ambiguous and VRFY is not listed
527 by IANA as a service extension. However, John Klensin suggests that we
530 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
531 the dkim code may be unix-mode line endings rather than smtp wire-format
532 CRLF, so prepend a CR to any bare LF.
534 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
535 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
537 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
538 the first were themselves being wrongly included in the feed into dkim
539 processing; with most chunk sizes in use this resulted in an incorrect
540 body hash calculated value.
542 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
543 DKIM signature block, for verification. Although advised against by
544 standards it is specifically not ruled illegal.
546 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
548 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
549 missing a body hash (the bh= tag).
551 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
552 It seems that HAProxy sends the Proxy Protocol information in clear and
553 only then does a TLS startup, so do the same.
555 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
556 TCP connections (such as for Spamd) unless the daemon successfully set
557 Fast Open mode on its listening sockets. This fixes breakage seen on
558 too-old kernels or those not configured for Fast Open, at the cost of
559 requiring both directions being enabled for TFO, and TFO never being used
560 by non-daemon-related Exim processes.
562 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
563 endings, at least on the first header line. Try to canonify any that get
564 past that check, despite the cost.
566 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
567 now limited to an arbitrary five deep, while parsing addresses with the
568 strip_excess_angle_brackets option enabled.
570 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
571 instead leave the unprompted TLS handshake in socket buffer for the
572 TLS library to consume.
574 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
576 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
578 JH/16 Drop variables when they go out of scope. Memory management drops a whole
579 region in one operation, for speed, and this leaves assigned pointers
580 dangling. Add checks run only under the testsuite which checks all
581 variables at a store-reset and panics on a dangling pointer; add code
582 explicitly nulling out all the variables discovered. Fixes one known
583 bug: a transport crash, where a dangling pointer for $sending_ip_address
584 originally assigned in a verify callout, is re-used.
586 PP/06 Drop '.' from @INC in various Perl scripts.
588 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
590 PP/08 Reduce a number of compilation warnings under clang; building with
591 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
592 should be warning-free.
594 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
596 HS/01 Fix portability problems introduced by PP/08 for platforms where
597 realloc(NULL) is not equivalent to malloc() [SunOS et al].
599 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
600 chunk. This allows us to accept broken chunked messages. We need a more
601 general solution here.
603 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
604 already-broken messages in the queue.
606 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
608 JH/19 Fix reference counting bug in routing-generated-address tracking.
614 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
615 supports it and a size is available (ie. the sending peer gave us one).
617 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
618 years of being deprecated). The replacements are the ACLs
619 acl_smtp_mime and acl_not_smtp_mime.
621 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
622 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
623 or even an in-clear connection were permitted. Now, if the host lookup
624 was dnssec and dane was requested then the host is only used if the
625 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
626 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
627 if one fails this test.
628 This means that a poorly-configured remote DNS will make it incommunicado;
629 but it protects against a DNS-interception attack on it.
631 JH/04 Bug 1810: make continued-use of an open smtp transport connection
632 non-noisy when a race steals the message being considered.
634 JH/05 If main configuration option tls_certificate is unset, generate a
635 self-signed certificate for inbound TLS connections.
637 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
638 in rewrites and routers.
640 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
641 and logged a warning sing 4.83; now they are a configuration file error.
643 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
644 (lacking @domain). Apply the same qualification processing as RCPT.
646 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
648 JH/10 Support ${sha256:} applied to a string (as well as the previous
651 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
652 a cutthrough deliver is pending, as we always want to make a connection.
653 This also avoids re-routing the message when later placing the cutthrough
654 connection after a verify cache hit.
655 Do not update it with the verify result either.
657 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
658 when routing results in more than one destination address.
660 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
661 signing (which inhibits the cutthrough capability). Previously only
662 the presence of an option was tested; now an expansion evaluating as
663 empty is permissible (obviously it should depend only on data available
664 when the cutthrough connection is made).
666 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
667 the relevant preceding SMTP command did not note the pipelining mode.
669 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
670 Previously they were not counted.
672 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
673 as one having no matching records. Previously we deferred the message
674 that needed the lookup.
676 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
677 distinguished as "(=".
679 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
680 for missing MX records. Previously it only worked for missing A records.
682 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
684 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
685 after the data-go-ahead and data-ack. Patch from Jason Betts.
687 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
688 even for a "none" policy. Patch from Tony Meyer.
690 JH/22 Fix continued use of a connection for further deliveries. If a port was
691 specified by a router, it must also match for the delivery to be
694 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
695 When one of the recipients of a message was unsuitable for the connection
696 (has no matching addresses), we lost track of needing to mark it
697 deferred. As a result mail would be lost.
699 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
701 JH/25 Decoding ACL controls is now done using a binary search; the source code
702 takes up less space and should be simpler to maintain. Merge the ACL
703 condition decode tables also, with similar effect.
705 JH/26 Fix problem with one_time used on a redirect router which returned the
706 parent address unchanged. A retry would see the parent address marked as
707 delivered, so not attempt the (identical) child. As a result mail would
710 JH/27 Fix a possible security hole, wherein a process operating with the Exim
711 UID can gain a root shell. Credit to http://www.halfdog.net/ for
712 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
715 JH/28 Enable {spool,log} filesystem space and inode checks as default.
716 Main config options check_{log,spool}_{inodes,space} are now
717 100 inodes, 10MB unless set otherwise in the configuration.
719 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
720 Previously it only applied to the main-section connection policy
723 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
725 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
726 by me. Added RFC7919 DH primes as an alternative.
728 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
729 are not in the system include path.
731 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
732 GnuTLS, when a session startup failed (eg because the client disconnected)
733 Exim did stdio operations after fclose. This was exposed by a recent
734 change which nulled out the file handle after the fclose.
736 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
737 signed directly by the cert-signing cert, rather than an intermediate
738 OCSP-signing cert. This is the model used by LetsEncrypt.
740 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
742 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
743 an incoming connection.
745 HS/02 Bug 1802: Do not half-close the connection after sending a request
748 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
749 fallback to "prime256v1".
751 JH/34 SECURITY: Use proper copy of DATA command in error message.
752 Could leak key material. Remotely exploitable. CVE-2016-9963.
758 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
759 and 3.4.4 - once the server is enabled to respond to an OCSP request
760 it does even when not requested, resulting in a stapling non-aware
761 client dropping the TLS connection.
763 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
764 support variable-length bit vectors. No functional change.
766 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
767 The I= interface field on outgoing lines is now after the H= remote
768 host field, same as incoming lines. There is a separate
769 outgoing_interface log selector which allows you to disable the
772 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
773 If not running log_selector +smtp_connection the mainlog would be held
774 open indefinitely after a "too many connections" event, including to a
775 deleted file after a log rotate. Leave the per net connection logging
776 leaving it open for efficiency as that will be quickly detected by the
777 check on the next write.
779 HS/01 Bug 1671: Fix post transport crash.
780 Processing the wait-<transport> messages could crash the delivery
781 process if the message IDs didn't exist for some reason. When
782 using 'split_spool_directory=yes' the construction of the spool
783 file name failed already, exposing the same netto behaviour.
785 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
786 mime_regex ACL conditions.
788 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
789 to DSN fail messages (bounces): remote IP, remote greeting, remote response
790 to HELO, local diagnostic string.
792 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
793 log line to debug. Even when configured with a tls authenticator many
794 client connections are expected to not authenticate in this way, so
795 an authenticate fail is not an error.
797 HS/02 Add the Exim version string to the process info. This way exiwhat
798 gives some more detail about the running daemon.
800 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
801 matter for fast-change records such as DNSBLs.
803 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
804 retry record, even if constant. There may be multiple transports with
805 different interface settings and the retry behaviour needs to be kept
808 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
810 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
812 JH/10 Bug 840: fix log_defer_output option of pipe transport
814 JH/11 Bug 830: use same host for all RCPTS of a message, even under
815 hosts_randomize. This matters a lot when combined with mua_wrapper.
817 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
818 ${quote_pgsql:<string>} operator.
820 JH/13 Bug 1708: avoid misaligned access in cached lookup.
822 JH/14 Change header file name for freeradius-client. Relevant if compiling
823 with Radius support; from the Gentoo tree and checked under Fedora.
825 JH/15 Bug 1712: Introduce $prdr_requested flag variable
827 JH/16 Bug 1714: Permit an empty string as expansion result for transport
828 option transport_filter, meaning no filtering.
830 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
832 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
833 defaults to "*" (all hosts). The variable is now available when not built
834 with TLS, default unset, mainly to enable keeping the testsuite sane.
835 If a server certificate is not supplied (via tls_certificate) an error is
836 logged, and clients will find TLS connections fail on startup. Presumably
837 they will retry in-clear.
838 Packagers of Exim are strongly encouraged to create a server certificate
839 at installation time.
841 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
842 with the $config_file variable.
844 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
845 in transport context, after the attempt, and per-recipient. The latter type
846 is per host attempted. The event data is the error message, and the errno
847 information encodes the lookup type (A vs. MX) used for the (first) host,
848 and the trailing two digits of the smtp 4xx response.
850 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
851 to write to mainlog (or rejectlog, paniclog) in the window between file
852 creation and permissions/ownership being changed. Particularly affects
853 installations where exicyclog is run as root, rather than exim user;
854 result is that the running daemon panics and dies.
856 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
858 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
859 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
860 "pri" and "weight". Note that the previous implicit priority given by the
861 list order is no longer honoured.
863 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
866 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
867 by defining SUPPORT_SOCKS.
869 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
870 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
871 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
872 variables are renamed to proxy_{local,external}_{address,port}.
874 JH/27 Move Internationalisation support from Experimental to mainline, enabled
875 for a build by defining SUPPORT_I18N
877 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
878 of the query string, and make ${quote_redis:} do that quoting.
880 JH/29 Move Events support from Experimental to mainline, enabled by default
881 and removable for a build by defining DISABLE_EVENT.
883 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
885 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
886 cached by the daemon.
888 JH/32 Move Redis support from Experimental to mainline, enabled for a build
889 by defining LOOKUP_REDIS. The libhiredis library is required.
891 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
892 keys are given for lookup.
894 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
895 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
896 only supported when built with TLS support. The PolarSSL SHA routines
897 are still used when the TLS library is too old for convenient support.
899 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
900 openssl_options), for security. OpenSSL forces this from version 1.1.0
901 server-side so match that on older versions.
903 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
904 allocation for $value could be released as the expansion processing
905 concluded, but leaving the global pointer active for it.
907 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
908 and to use the domains and local_parts ACL conditions.
910 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
911 incorrectly not doubled on cutthrough transmission, hence seen as a
912 body-termination at the receiving system - resulting in truncated mails.
913 Commonly the sender saw a TCP-level error, and retransmitted the message
914 via the normal store-and-forward channel. This could result in duplicates
915 received - but deduplicating mailstores were liable to retain only the
916 initial truncated version.
918 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
920 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
922 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
923 we're in there, support oversigning also; bug 1309.
925 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
927 HS/04 Add support for keep_environment and add_environment options.
929 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
930 either intentional arithmetic overflow during PRNG, or testing config-
933 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
934 delivery resulted in actual delivery. Cancel cutthrough before DATA
937 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
938 rejects a recipient: pass the reject to the originator.
940 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
941 Many were false-positives and ignorable, but it's worth fixing the
944 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
945 for the new environment-manipulation done at startup. Move the routines
946 from being local to tls.c to being global via the os.c file.
948 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
949 an extract embedded as result-arg for a map, the first arg for extract
950 is unavailable so we cannot tell if this is a numbered or keyed
951 extraction. Accept either.
957 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
960 JH/02 The smtp transport option "multi_domain" is now expanded.
962 JH/03 The smtp transport now requests PRDR by default, if the server offers
965 JH/04 Certificate name checking on server certificates, when exim is a client,
966 is now done by default. The transport option tls_verify_cert_hostnames
967 can be used to disable this per-host. The build option
968 EXPERIMENTAL_CERTNAMES is withdrawn.
970 JH/05 The value of the tls_verify_certificates smtp transport and main options
971 default to the word "system" to access the system default CA bundle.
972 For GnuTLS, only version 3.0.20 or later.
974 JH/06 Verification of the server certificate for a TLS connection is now tried
975 (but not required) by default. The verification status is now logged by
976 default, for both outbound TLS and client-certificate supplying inbound
979 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
982 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
983 Status Notification (bounce) messages are now MIME format per RFC 3464.
984 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
985 under the control of the dsn_advertise_hosts option, and routers may
986 have a dsn_lasthop option.
988 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
989 default, modifiable by a malware= option. The list separator for
990 the options can now be changed in the usual way. Bug 68.
992 JH/10 The smtp_receive_timeout main option is now expanded before use.
994 JH/11 The incoming_interface log option now also enables logging of the
995 local interface on delivery outgoing connections.
997 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
998 if the interface and destination host and port all match.
1000 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1003 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1004 Patch from Andrew Lewis.
1006 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1007 now supports optional time-restrictions, weighting, and priority
1008 modifiers per server. Patch originally by <rommer@active.by>.
1010 JH/16 The spamd_address main option now supports a mixed list of local
1011 and remote servers. Remote servers can be IPv6 addresses, and
1012 specify a port-range.
1014 JH/17 Bug 68: The spamd_address main option now supports an optional
1015 timeout value per server.
1017 JH/18 Bug 1581: Router and transport options headers_add/remove can
1018 now have the list separator specified.
1020 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1023 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1026 JH/21 Support for the A6 type of dns record is withdrawn.
1028 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1029 rather than the verbs used.
1031 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1032 from 255 to 1024 chars.
1034 JH/24 Verification callouts now attempt to use TLS by default.
1036 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1037 are generic router options now. The defaults didn't change.
1039 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1040 Original patch from Alexander Shikoff, worked over by JH.
1042 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1043 files if ZCAT_COMMAND is not executable.
1045 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1047 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1049 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1050 Normally benign, it bites when the pair was led to by a CNAME;
1051 modern usage is to not canonicalize the domain to a CNAME target
1052 (and we were inconsistent anyway for A-only vs AAAA+A).
1054 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1056 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1057 when evaluating $sender_host_dnssec.
1059 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1060 $sender_helo_dnssec variable.
1062 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1064 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1066 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1068 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1069 documented as working, but never had. Support all but $spam_report.
1071 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1072 added for tls authenticator.
1074 HS/03 Add perl_taintmode main config option
1080 TL/01 When running the test suite, the README says that variables such as
1081 no_msglog_check are global and can be placed anywhere in a specific
1082 test's script, however it was observed that placement needed to be near
1083 the beginning for it to behave that way. Changed the runtest perl
1084 script to read through the entire script once to detect and set these
1085 variables, reset to the beginning of the script, and then run through
1086 the script parsing/test process like normal.
1088 TL/02 The BSD's have an arc4random API. One of the functions to induce
1089 adding randomness was arc4random_stir(), but it has been removed in
1090 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1091 function when detected.
1093 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1094 cause callback expansion.
1096 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1097 syntax errors in an expansion can be treated as a string instead of
1098 logging or causing an error, due to the internal use of bool_lax
1099 instead of bool when processing it.
1101 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1102 server certificates when making smtp deliveries.
1104 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1106 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1108 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1110 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1111 Merged patch from Sebastian Wiedenroth.
1113 JH/05 Fix results-pipe from transport process. Several recipients, combined
1114 with certificate use, exposed issues where response data items split
1115 over buffer boundaries were not parsed properly. This eventually
1116 resulted in duplicates being sent. This issue only became common enough
1117 to notice due to the introduction of connection certificate information,
1118 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1120 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1121 size buffer was used, resulting in syntax errors when an expansion
1124 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1125 version 3.3.6 or later.
1127 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1128 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1129 both become "event_action", the variables become $event_name, $event_data
1130 and $event_defer_errno. There is a new variable $verify_mode, usable in
1131 routers, transports and related events. The tls:cert event is now also
1132 raised for inbound connections, if the main configuration event_action
1135 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1136 early OCSP support, but no stapling (appears to be less than 1.0.0).
1138 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1139 server certificate names available under the smtp transport option
1140 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1143 JH/10 Time-related extraction expansions from certificates now use the main
1144 option "timezone" setting for output formatting, and are consistent
1145 between OpenSSL and GnuTLS compilations. Bug 1541.
1147 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1148 encoded parameter in the incoming message. Bug 1558.
1150 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1151 include certificate info, eximon was claiming there were spoolfile
1154 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1156 JH/14 Log delivery-related information more consistently, using the sequence
1157 "H=<name> [<ip>]" wherever possible.
1159 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1160 are problematic for Debian distribution, omit them from the release
1163 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1165 JH/16 Fix string representation of time values on 64bit time_t architectures.
1168 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1169 output list separator was used.
1174 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1175 checkers that were complaining about end of non-void function with no
1178 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1179 This was a regression introduced in 4.83 by another bugfix.
1181 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1183 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1184 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1190 TF/01 Correctly close the server side of TLS when forking for delivery.
1192 When a message was received over SMTP with TLS, Exim failed to clear up
1193 the incoming connection properly after forking off the child process to
1194 deliver the message. In some situations the subsequent outgoing
1195 delivery connection happened to have the same fd number as the incoming
1196 connection previously had. Exim would try to use TLS and fail, logging
1197 a "Bad file descriptor" error.
1199 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1200 utilities have not been installed.
1202 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1203 temporary space as the ACL may create new global variables.
1205 TL/01 LDAP support uses per connection or global context settings, depending
1206 upon the detected version of the libraries at build time.
1208 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1209 to extract and use the src ip:port in logging and expansions as if it
1210 were a direct connection from the outside internet. PPv2 support was
1211 updated based on HAProxy spec change in May 2014.
1213 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1215 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1216 Properly escape header and check for NULL return.
1218 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1221 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1223 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1224 characters in header names, implemented as a verify condition.
1225 Contributed by Michael Fischer v. Mollard.
1227 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1228 results permerror and temperror. Previous values are deprecated but
1229 still accepted. In a future release, err_perm and err_temp will be
1230 completely removed, which will be a backward incompatibility if the
1231 ACL tests for either of these two old results. Patch contributed by
1232 user bes-internal on the mailing list.
1234 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1236 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1237 selectors, in both main and reject logs.
1239 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1242 JH/07 Add malware type "sock" for talking to simple daemon.
1244 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1246 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1247 routers/transports under cutthrough routing.
1249 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1250 numbers. Touch up "bool" conditional to keep the same definition.
1252 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1254 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1256 JH/12 Expand items in router/transport headers_add or headers_remove lists
1257 individually rather than the list as a whole. Bug 1452.
1259 Required for reasonable handling of multiple headers_ options when
1260 they may be empty; requires that headers_remove items with embedded
1261 colons must have them doubled (or the list-separator changed).
1263 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1264 view the policy declared in the DMARC record. Currently, $dmarc_status
1265 is a combined value of both the record presence and the result of the
1268 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1270 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1271 dnslookup router and the smtp transport (applying to the forward
1274 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1275 of ldap servers used for a specific lookup. Patch provided by Heiko
1278 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1279 New variable $lookup_dnssec_authenticated for observability.
1281 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1282 Patch submitted by Lars Timman.
1284 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1286 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1287 Requires trusted mode and valid format message id, aborts otherwise.
1288 Patch contributed by Heiko Schlichting.
1290 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1291 certextract with support for various fields. Bug 1358.
1293 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1294 is requested by default, modifiable by smtp transport option
1297 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1298 operate on certificate variables to give certificate fingerprints
1299 Also new ${sha256:cert_variable}.
1301 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1303 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1304 Christian Aistleitner.
1306 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1308 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1309 file. Patch from Wolfgang Breyha.
1311 JH/25 Expand the coverage of the delivery $host and $host_address to
1312 client authenticators run in verify callout. Bug 1476.
1314 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1315 align with daemon_smtp_ports. Bug 72.
1317 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1318 support and error reporting did not work properly.
1320 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1321 and is readable. Patch from Andrew Colin Kissa.
1323 TL/14 Enhance documentation of ${run expansion and how it parses the
1324 commandline after expansion, particularly in the case when an
1325 unquoted variable expansion results in an empty value.
1327 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1329 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1330 by renaming away from it. Addresses GH issue 15, reported by
1333 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1335 TL/15 SECURITY: prevent double expansion in math comparison functions
1336 (can expand unsanitized data). Not remotely exploitable.
1343 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1345 PP/02 Make -n do something, by making it not do something.
1346 When combined with -bP, the name of an option is not output.
1348 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1351 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1352 $sender_host_name and config options to manage this, and basic check
1355 PP/05 DSCP support for outbound connections and control modifier for inbound.
1357 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1358 (Only plugin which currently uses this is kerberos4, which nobody should
1359 be using, but we should make it available and other future plugins might
1360 conceivably use it, even though it would break NAT; stuff *should* be
1361 using channel bindings instead).
1363 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1364 name; added for Sendmail compatibility; requires admin caller.
1365 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1366 just ignore it); requires trusted caller.
1367 Also parse but ignore: -Ac -Am -X<logfile>
1370 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1372 TL/02 Add +smtp_confirmation as a default logging option.
1374 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1375 Patch by Magnus Holmgren from 2007-02-20.
1377 TL/04 Bugzilla 1281 - Spec typo.
1378 Bugzilla 1283 - Spec typo.
1379 Bugzilla 1290 - Spec grammar fixes.
1381 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1383 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1385 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1386 mailing list by by Dmitry Isaikin.
1388 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1390 JH/02 Support "G" suffix to numbers in ${if comparisons.
1392 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1394 NM/01 Bugzilla 1197 - Spec typo
1395 Bugzilla 1196 - Spec examples corrections
1397 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1399 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1400 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1403 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1404 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1406 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1407 "acl {{name}{arg}...}", and optional args on acl condition
1410 JH/05 Permit multiple router/transport headers_add/remove lines.
1412 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1414 JH/07 Avoid using a waiting database for a single-message-only transport.
1415 Performance patch from Paul Fisher. Bugzilla 1262.
1417 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1420 JH/09 Add $headers_added variable, with content from use of ACL modifier
1421 add_header (but not yet added to the message). Bugzilla 199.
1423 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1424 Pulled from Bugzilla 817 by Wolfgang Breyha.
1426 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1428 (nb: this is the same fix as in Exim 4.80.1)
1430 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1433 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1436 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1438 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1439 advertise SMTP AUTH mechanism to us, instead of a generic
1440 protocol violation error. Also, make Exim more robust to bad
1441 data from the Dovecot auth socket.
1443 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1445 When a queue runner is handling a message, Exim first routes the
1446 recipient addresses, during which it prunes them based on the retry
1447 hints database. After that it attempts to deliver the message to
1448 any remaining recipients. It then updates the hints database using
1451 So if a recipient address works intermittently, it can get repeatedly
1452 deferred at routing time. The retry hints record remains fresh so the
1453 address never reaches the final cutoff time.
1455 This is a fairly common occurrence when a user is bumping up against
1456 their storage quota. Exim had some logic in its local delivery code
1457 to deal with this. However it did not apply to per-recipient defers
1458 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1460 This change adds a proper retry rule check during routing so that the
1461 final cutoff time is checked against the message's age. We only do
1462 this check if there is an address retry record and there is not a
1463 domain retry record; this implies that previous attempts to handle
1464 the address had the retry_use_local_parts option turned on. We use
1465 this as an approximation for the destination being like a local
1466 delivery, as in LMTP.
1468 I suspect this new check makes the old local delivery cutoff check
1469 redundant, but I have not verified this so I left the code in place.
1471 TF/02 Correct gecos expansion when From: is a prefix of the username.
1473 Test 0254 submits a message to Exim with the header
1477 When I ran the test suite under the user fanf2, Exim expanded
1478 the header to contain my full name, whereas it should have added
1479 a Resent-Sender: header. It erroneously treats any prefix of the
1480 username as equal to the username.
1482 This change corrects that bug.
1484 GF/01 DCC debug and logging tidyup
1485 Error conditions log to paniclog rather than rejectlog.
1486 Debug lines prefixed by "DCC: " to remove any ambiguity.
1488 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1490 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1491 Bug spotted by Jeremy Harris; was flawed since initial commit.
1492 Would have resulted in OCSP responses post-SNI triggering an Exim
1493 NULL dereference and crash.
1495 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1497 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1498 Bug detection, analysis and fix by Samuel Thibault.
1499 Bugzilla 1331, Debian bug #698092.
1501 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1503 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1504 Server implementation by Todd Lyons, client by JH.
1505 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1506 config variable "prdr_enable" controls whether the server
1507 advertises the facility. If the client requests PRDR a new
1508 acl_data_smtp_prdr ACL is called once for each recipient, after
1509 the body content is received and before the acl_smtp_data ACL.
1510 The client is controlled by both of: a hosts_try_prdr option
1511 on the smtp transport, and the server advertisement.
1512 Default client logging of deliveries and rejections involving
1513 PRDR are flagged with the string "PRDR".
1515 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1516 fclose(). Diagnosis by Todd Lyons.
1518 PP/17 Update configure.default to handle IPv6 localhost better.
1519 Patch by Alain Williams (plus minor tweaks).
1522 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1523 This is now consistent with GnuTLS, and is now documented: the
1524 previous undocumented portable approach to treating the option as
1525 unset was to force an expansion failure. That still works, and
1526 an empty string is now equivalent.
1528 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1529 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1530 not performing validation itself.
1532 PP/20 Added force_command boolean option to pipe transport.
1533 Patch from Nick Koston, of cPanel Inc.
1535 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1538 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1540 PP/21 Fix eximon continuous updating with timestamped log-files.
1541 Broken in a format-string cleanup in 4.80, missed when I repaired the
1542 other false fix of the same issue.
1543 Report and fix from Heiko Schlichting.
1546 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1547 Report from Prashanth Katuri.
1549 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1550 It's SecureTransport, so affects any MacOS clients which use the
1551 system-integrated TLS libraries, including email clients.
1553 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1554 using a MIME ACL for non-SMTP local injection.
1555 Report and assistance in diagnosis by Warren Baker.
1557 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1559 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1561 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1562 last id that failed so it may be referenced in subsequent ACL's.
1564 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1567 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1568 ldap library initialization, allowing self-signed CA's to be
1569 used. Also properly sets require_cert option later in code by
1570 using NULL (global ldap config) instead of ldap handle (per
1571 session). Bug diagnosis and testing by alxgomz.
1573 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1574 the src/util/ subdirectory.
1576 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1577 renamed to Transport Post Delivery Action by Jeremy Harris, as
1580 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1581 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1582 redis_servers = needs to be configured which will be used by the redis
1583 lookup. Patch from Warren Baker, of The Packet Hub.
1585 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1587 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1588 hostname or reverse DNS when processing a host list. Used suggestions
1589 from multiple comments on this bug.
1591 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1593 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1594 Missed a few lines, added it to make the runtest require no keyboard
1597 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1598 contains upper case chars. Make router use caseful_local_part.
1600 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1601 support when GnuTLS has been built with p11-kit.
1607 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1609 This, or similar/improved, will also be change PP/11 of 4.82.
1615 PP/01 Handle short writes when writing local log-files.
1616 In practice, only affects FreeBSD (8 onwards).
1617 Bugzilla 1053, with thanks to Dmitry Isaikin.
1619 NM/01 Bugzilla 949 - Documentation tweak
1621 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1624 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1626 PP/02 Implemented gsasl authenticator.
1628 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1630 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1631 `pkg-config foo` for cflags/libs.
1633 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1634 with rest of GSASL and with heimdal_gssapi.
1636 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1637 `pkg-config foo` for cflags/libs for the TLS implementation.
1639 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1640 properties get this fed in as external SSF. A number of robustness
1641 and debugging improvements to the cyrus_sasl authenticator.
1643 PP/08 cyrus_sasl server now expands the server_realm option.
1645 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1646 Patch by Jeremy Harris.
1648 PP/10 Added dbmjz lookup type.
1650 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1652 PP/12 MAIL args handles TAB as well as SP, for better interop with
1653 non-compliant senders.
1654 Analysis and variant patch by Todd Lyons.
1656 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1657 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1658 Patch from Dirk Mueller <dmueller@suse.com>
1660 PP/13 tls_peerdn now print-escaped for spool files.
1661 Observed some $tls_peerdn in wild which contained \n, which resulted
1662 in spool file corruption.
1664 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1665 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1666 or write after TLS renegotiation, which otherwise led to messages
1669 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1670 as a tracking header (ie: a signed header comes before the signature).
1671 Patch from Wolfgang Breyha.
1673 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1674 comma-sep list; embedded commas doubled.
1676 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1678 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1680 Report and patch from Dmitry Banschikov.
1682 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1683 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1684 failures. We appear to now support TLS1.1+ with Exim.
1686 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1687 lets Exim select keys and certificates based upon TLS SNI from client.
1688 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1689 before an outbound SMTP session. New log_selector, +tls_sni.
1691 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1692 NULL dereference. Report and patch from Alun Jones.
1694 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1695 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1696 Not seeing resolver debug output on NetBSD, but suspect this is a
1697 resolver implementation change.
1699 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1700 Left warnings. Added "eximon gdb" invocation mode.
1702 PP/21 Defaulting "accept_8bitmime" to true, not false.
1704 PP/22 Added -bw for inetd wait mode support.
1706 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1707 locate the relevant includes and libraries. Made this the default.
1709 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1710 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1712 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1713 This may cause build issues on older platforms.
1715 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1716 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1717 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1718 Added SNI support via GnuTLS too.
1719 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1721 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1723 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1724 Applied second patch from Janne, implementing suggestion to default
1725 multiple-strings-in-record handling to match SPF spec.
1727 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1729 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1730 read-only, out of scope).
1731 Patch from Wolfgang Breyha, report from Stuart Northfield.
1733 PP/29 Fix three issues highlighted by clang analyser static analysis.
1734 Only crash-plausible issue would require the Cambridge-specific
1735 iplookup router and a misconfiguration.
1736 Report from Marcin Mirosław.
1738 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1740 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1741 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1742 As part of this, removing so much warning spew let me fix some minor
1743 real issues in debug logging.
1745 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1746 assignment on my part. Fixed.
1748 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1749 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1750 Janne Snabb (who went above and beyond: thank you).
1752 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1753 string otherwise requires a connection and a bunch more work and it's
1754 relatively easy to get wrong. Should also expose TLS library linkage
1757 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1758 64-bit ${eval} (JH/03).
1760 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1761 GNU libc to support some of the 64-bit stuff, should not lead to
1762 conflicts. Defined before os.h is pulled in, so if a given platform
1763 needs to override this, it can.
1765 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1766 protection layer was required, which is not implemented.
1767 Bugzilla 1254, patch from Wolfgang Breyha.
1769 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1770 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1771 tls_dhparam take prime identifiers. Also unbreak combination of
1772 OpenSSL+DH_params+TLSSNI.
1774 PP/39 Disable SSLv2 by default in OpenSSL support.
1780 PP/01 Solaris build fix for Oracle's LDAP libraries.
1781 Bugzilla 1109, patch from Stephen Usher.
1783 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1785 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1788 TF/02 Fix a couple more cases where we did not log the error message
1789 when unlink() failed. See also change 4.74-TF/03.
1791 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1792 lock up or crash if it happened to be inside a call to libc when it
1793 got a SIGUSR1 from exiwhat.
1795 The SIGUSR1 handler appends the current process status to the process
1796 log which is later printed by exiwhat. It used to use the general
1797 purpose logging code to do this, but several functions it calls are
1798 not safe for signals.
1800 The new output code in the SIGUSR1 handler is specific to the process
1801 log, and simple enough that it's easy to inspect for signal safety.
1802 Removing some special cases also simplifies the general logging code.
1803 Removing the spurious timestamps from the process log simplifies
1806 TF/04 Improved ratelimit ACL condition.
1808 The /noupdate option has been deprecated in favour of /readonly which
1809 has clearer semantics. The /leaky, /strict, and /readonly update modes
1810 are mutually exclusive. The update mode is no longer included in the
1811 database key; it just determines when the database is updated. (This
1812 means that when you upgrade Exim will forget old rate measurements.)
1814 Exim now checks that the per_* options are used with an update mode that
1815 makes sense for the current ACL. For example, when Exim is processing a
1816 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1817 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1818 must specify per_mail/readonly. If you omit the update mode it defaults to
1819 /leaky where that makes sense (as before) or /readonly where required.
1821 The /noupdate option is now undocumented but still supported for
1822 backwards compatibility. It is equivalent to /readonly except that in
1823 ACLs where /readonly is required you may specify /leaky/noupdate or
1824 /strict/noupdate which are treated the same as /readonly.
1826 A useful new feature is the /count= option. This is a generalization
1827 of the per_byte option, so that you can measure the throughput of other
1828 aggregate values. For example, the per_byte option is now equivalent
1829 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1831 The per_rcpt option has been generalized using the /count= mechanism
1832 (though it's more complicated than the per_byte equivalence). When it is
1833 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1834 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1835 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1836 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1837 Note that using per_rcpt with a non-readonly update mode in more than
1838 one ACL will cause the recipients to be double-counted. (The per_mail
1839 and per_byte options don't have this problem.)
1841 The handling of very low rates has changed slightly. If the computed rate
1842 is less than the event's count (usually one) then this event is the first
1843 after a long gap. In this case the rate is set to the same as this event's
1844 count, so that the first message of a spam run is counted properly.
1846 The major new feature is a mechanism for counting the rate of unique
1847 events. The new per_addr option counts the number of different
1848 recipients that someone has sent messages to in the last time period. It
1849 behaves like per_rcpt if all the recipient addresses are different, but
1850 duplicate recipient addresses do not increase the measured rate. Like
1851 the /count= option this is a general mechanism, so the per_addr option
1852 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1853 example, measure the rate that a client uses different sender addresses
1854 with the options per_mail/unique=$sender_address. There are further
1855 details in the main documentation.
1857 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1859 TF/06 Removed a few PCRE remnants.
1861 TF/07 Automatically extract Exim's version number from tags in the git
1862 repository when doing development or release builds.
1864 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1865 Bugzilla 879. Patch from Paul Fisher.
1867 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1868 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1871 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1873 PP/05 Variable $av_failed, true if the AV scanner deferred.
1874 Bugzilla 1078. Patch from John Horne.
1876 PP/06 Stop make process more reliably on build failure.
1877 Bugzilla 1087. Patch from Heiko Schlittermann.
1879 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1880 Bugzilla 1089. Patch from Heiko Schlittermann.
1882 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1883 Bugzilla 1131. Patch from Holger Weiß.
1885 PP/09 Handle IPv6 addresses with SPF.
1886 Bugzilla 860. Patch from Wolfgang Breyha.
1888 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1890 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1893 PP/11 match_* no longer expand right-hand-side by default.
1894 New compile-time build option, EXPAND_LISTMATCH_RHS.
1895 New expansion conditions, "inlist", "inlisti".
1897 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1899 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1901 PP/14 fix log_write() format string regression from TF/03.
1902 Bugzilla 1152. Patch from Dmitry Isaikin.
1908 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1910 PP/02 Harmonised TLS library version reporting; only show if debugging.
1911 Layout now matches that introduced for other libraries in 4.74 PP/03.
1913 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1915 PP/04 New "dns_use_edns0" global option.
1917 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1920 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1921 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1923 TK/01 Updated PolarSSL code to 0.14.2.
1924 Bugzilla 1097. Patch from Andreas Metzler.
1926 PP/07 Catch divide-by-zero in ${eval:...}.
1927 Fixes bugzilla 1102.
1929 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1932 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1933 format-string attack -- SECURITY: remote arbitrary code execution.
1935 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1936 time unintentionally subject to list matching rules, letting the header
1937 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1938 arbitrary string expansion). This allowed for information disclosure.
1940 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1941 INT_MIN/-1 -- value coerced to INT_MAX.
1947 NM/01 Workaround for PCRE version dependency in version reporting
1950 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1951 This fixes portability to compilers other than gcc, notably
1952 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1954 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1955 makefiles for portability to HP-UX and POSIX correctness.
1957 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1958 Also via indented variable definition in the Makefile.
1959 (Debugging by Oliver Heesakkers).
1961 PP/02 Restore caching of spamd results with expanded spamd_address.
1962 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1964 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1965 Improves build reliability. Fix from: Frank Elsner
1967 NM/02 Fix wide character breakage in the rfc2047 coding
1968 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1970 NM/03 Allow underscore in dnslist lookups
1971 Fixes bug 1026. Patch from Graeme Fowler
1973 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1974 Code patches from Adam Ciarcinski of NetBSD.
1976 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1979 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1980 is logged, to avoid truncation. Patch from John Horne.
1982 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1983 Patch from Jakob Hirsch.
1985 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1986 SQL string expansion failure details.
1987 Patch from Andrey Oktyabrski.
1989 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1990 Patch from Simon Arlott.
1992 PP/09 New lookups functionality failed to compile on old gcc which rejects
1993 extern declarations in function scope.
1994 Patch from Oliver Fleischmann
1996 PP/10 Use sig_atomic_t for flags set from signal handlers.
1997 Check getgroups() return and improve debugging.
1998 Fixed developed for diagnosis in bug 927 (which turned out to be
2001 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2002 Patch from Mark Zealey.
2004 PP/12 Bugzilla 1056: Improved spamd server selection.
2005 Patch from Mark Zealey.
2007 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2008 Based on patch from Heiko Schlittermann.
2010 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2011 Patch from Uwe Doering, sign-off by Michael Haardt.
2013 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2014 variable declaration deep within a block. Bug and patch from
2017 PP/15 lookups-Makefile IRIX compatibility coercion.
2019 PP/16 Make DISABLE_DKIM build knob functional.
2021 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2022 Patch by Simon Arlott
2024 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2025 variable argument macros. Our copy now differs from upstream.
2031 TF/01 Failure to get a lock on a hints database can have serious
2032 consequences so log it to the panic log.
2034 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2035 controlled using the smtp_confirmation log selector.
2037 TF/03 Include the error message when we fail to unlink a spool file.
2039 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2040 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2041 for maintaining out-of-tree patches for some time.
2043 PP/01 Bugzilla 139: Documentation and portability issues.
2044 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2045 Handle per-OS dynamic-module compilation flags.
2047 PP/02 Let /dev/null have normal permissions.
2048 The 4.73 fixes were a little too stringent and complained about the
2049 permissions on /dev/null. Exempt it from some checks.
2050 Reported by Andreas M. Kirchwitz.
2052 PP/03 Report version information for many libraries, including
2053 Exim version information for dynamically loaded libraries. Created
2054 version.h, now support a version extension string for distributors
2055 who patch heavily. Dynamic module ABI change.
2057 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2058 privilege escalation vulnerability whereby the Exim run-time user
2059 can cause root to append content of the attacker's choosing to
2062 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2065 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2066 If dropping privileges for untrusted macros, we disabled normal logging
2067 on the basis that it would fail; for the Exim run-time user, this is not
2068 the case, and it resulted in successful deliveries going unlogged.
2069 Fixed. Reported by Andreas Metzler.
2075 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2076 only prepend for the Resent-* case. Fixes regression introduced in
2077 Exim 4.70 by NM/22 for Bugzilla 607.
2079 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2080 increasing numbers of administrators be bitten by this.
2082 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2084 PP/03 Bugzilla 994: added openssl_options main configuration option.
2086 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2088 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2090 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2092 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2093 without a peer certificate, leading to a segfault because of an
2094 assumption that peers always have certificates. Be a little more
2095 paranoid. Problem reported by Martin Tscholak.
2097 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2098 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2099 NB: ClamAV planning to remove STREAM in "middle of 2010".
2100 CL also introduces -bmalware, various -d+acl logging additions and
2101 more caution in buffer sizes.
2103 PP/09 Implemented reverse_ip expansion operator.
2105 PP/10 Bugzilla 937: provide a "debug" ACL control.
2107 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2109 PP/12 Bugzilla 973: Implement --version.
2111 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2113 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2115 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2117 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2118 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2119 ignore trailing whitespace.
2121 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2123 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2126 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2127 Notification from Dr Andrew Aitchison.
2129 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2130 ExtendedDetectionInfo response format.
2131 Notification from John Horne.
2133 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2136 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2137 XSL and documented dependency on system catalogs, with examples of how
2140 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2143 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2144 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2145 configuration file which is writeable by the Exim user or group.
2147 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2148 of configuration files to cover files specified with the -C option if
2149 they are going to be used with root privileges, not just the default
2152 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2153 option (effectively making it always true).
2155 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2156 files to be used while preserving root privileges.
2158 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2159 that rogue child processes cannot use them.
2161 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2162 run-time user, instead of root.
2164 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2165 Exim run-time user without dropping privileges.
2167 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2168 result string, instead of calling string_vformat() twice with the same
2171 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2172 for other users. Others should always drop root privileges if they use
2173 -C on the command line, even for a whitelisted configure file.
2175 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2177 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2183 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2184 $data_path, and $header_path variables; fixed documentation bugs and
2187 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2188 exipick to access non-standard spools, including the "frozen" queue
2191 NM/01 Bugzilla 965: Support mysql stored procedures.
2192 Patch from Alain Williams
2194 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2196 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2197 Patch from Andreas Metzler
2199 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2200 Patch from Kirill Miazine
2202 NM/05 Bugzilla 671: Added umask to procmail example.
2204 JJ/03 installed exipick 20100323.0, fixing doc bug
2206 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2207 directory. Notification and patch from Dan Rosenberg.
2209 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2211 TK/02 Improve log output when DKIM signing operation fails.
2213 MH/01 Treat the transport option dkim_domain as a colon separated
2214 list, not as a single string, and sign the message with each element,
2215 omitting multiple occurences of the same signer.
2217 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2218 Bugzilla 985, 986. Patch by Simon Arlott
2220 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2221 Patch by Simon Arlott
2223 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2224 MBX locking. Notification from Dan Rosenberg.
2230 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2232 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2234 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2236 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2238 NM/04 Bugzilla 915: Flag broken perl installation during build.
2244 TK/01 Added patch by Johannes Berg that expands the main option
2245 "spamd_address" if it starts with a dollar sign.
2247 TK/02 Write list of recipients to X-Envelope-Sender header when building
2248 the mbox-format spool file for content scanning (suggested by Jakob
2251 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2252 (http://www.dcc-servers.net/) support via dccifd. Activated by
2253 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2255 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2256 by Mark Daniel Reidel <mr@df.eu>.
2258 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2259 When building exim an external PCRE library is now needed -
2260 PCRE is a system library on the majority of modern systems.
2261 See entry on PCRE_LIBS in EDITME file.
2263 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2264 conversation. Added nologin parameter to request.
2265 Patch contributed by Kirill Miazine.
2267 TF/01 Do not log submission mode rewrites if they do not change the address.
2269 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2271 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2272 log files in place. Contributed by Roberto Lima.
2274 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2276 TF/03 Bugzilla 615: When checking the local_parts router precondition
2277 after a local_part_suffix or local_part_prefix option, Exim now
2278 does not use the address's named list lookup cache, since this
2279 contains cached lookups for the whole local part.
2281 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2282 Robert Millan. Documentation is in experimental-spec.txt.
2284 TF/04 Bugzilla 668: Fix parallel build (make -j).
2286 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2288 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2289 Patch provided by Jan Srzednicki.
2291 TF/05 Leading white space used to be stripped from $spam_report which
2292 wrecked the formatting. Now it is preserved.
2294 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2295 that they are available at delivery time.
2297 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2299 TF/08 TLS error reporting now respects the incoming_interface and
2300 incoming_port log selectors.
2302 TF/09 Produce a more useful error message if an SMTP transport's hosts
2303 setting expands to an empty string.
2305 NM/06 Bugzilla 744: EXPN did not work under TLS.
2306 Patch provided by Phil Pennock.
2308 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2309 Patch provided by Richard Godbee.
2311 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2312 acl_smtp_notquit, added index entry.
2314 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2315 Patch provided by Eugene Bujak.
2317 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2318 accept(). Patch provided by Maxim Dounin.
2320 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2321 Patch provided by Phil Pennock.
2323 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2325 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2326 Patch provided by Brad "anomie" Jorsch.
2328 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2329 Patch provided by Dean Brooks.
2331 TK/05 Add native DKIM support (does not depend on external libraries).
2333 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2334 Patch provided by Graeme Fowler.
2336 NM/16 Bugzilla 851: Documentation example syntax fix.
2338 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2340 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2343 NM/19 Bugzilla 745: TLS version reporting.
2344 Patch provided by Phil Pennock.
2346 NM/20 Bugzilla 167: bool: condition support.
2347 Patch provided by Phil Pennock.
2349 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2350 clients. Patch provided by Phil Pennock.
2352 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2353 Patch provided by Brad "anomie" Jorsch.
2355 NM/23 Bugzilla 687: Fix misparses in eximstats.
2356 Patch provided by Heiko Schlittermann.
2358 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2359 Patch provided by Heiko Schlittermann.
2361 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2362 plus update to original patch.
2364 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2366 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2367 Patch provided by David Brownlee.
2369 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2371 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2373 NM/30 Bugzilla 888: TLS documentation bugfixes.
2375 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2377 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2378 Unlike the original bugzilla I have changed all shell scripts in src tree.
2380 NM/33 Bugzilla 898: Transport filter timeout fix.
2381 Patch by Todd Rinaldo.
2383 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2384 Patch by Serge Demonchaux.
2386 NM/35 Bugzilla 39: Base64 decode bug fixes.
2387 Patch by Jakob Hirsch.
2389 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2391 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2393 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2395 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2401 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2402 ALT-N's libdkim that I have put here:
2403 http://duncanthrax.net/exim-experimental/
2405 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2406 were called 'true' and it seems that C99 defines that as a reserved
2407 keyword to be used with 'bool' variable types. That means you could
2408 not include C99-style headers which use bools without triggering
2409 build errors in sieve.c.
2411 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2412 as mailq or other aliases. Changed the --help handling significantly
2413 to do whats expected. exim_usage() emits usage/help information.
2415 SC/01 Added the -bylocaldomain option to eximstats.
2417 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2419 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2421 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2427 PH/01 Another patch from the Sieve maintainer.
2429 PH/02 When an IPv6 address is converted to a string for single-key lookup
2430 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2431 dots are used instead of colons so that keys in lsearch files need not
2432 contain colons. This was done some time before quoting was made available
2433 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2434 (notated using the quote facility) so as to distinguish them from IPv4
2435 keys. This meant that lookups for IP addresses in host lists did not work
2436 for iplsearch lookups.
2438 This has been fixed by arranging for IPv6 addresses to be expressed with
2439 colons if the lookup type is iplsearch. This is not incompatible, because
2440 previously such lookups could never work.
2442 The situation is now rather anomalous, since one *can* have colons in
2443 ordinary lsearch keys. However, making the change in all cases is
2444 incompatible and would probably break a number of configurations.
2446 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2449 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2450 conversion specification without a maximum field width, thereby enabling
2451 a rogue spamd server to cause a buffer overflow. While nobody in their
2452 right mind would setup Exim to query an untrusted spamd server, an
2453 attacker that gains access to a server running spamd could potentially
2454 exploit this vulnerability to run arbitrary code as the Exim user.
2456 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2457 $primary_hostname instead of what libspf2 thinks the hosts name is.
2459 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2460 a directory entry by the name of the lookup key. Previously, if a
2461 symlink pointed to a non-existing file or a file in a directory that
2462 Exim lacked permissions to read, a lookup for a key matching that
2463 symlink would fail. Now it is enough that a matching directory entry
2464 exists, symlink or not. (Bugzilla 503.)
2466 PH/03 The body_linecount and body_zerocount variables are now exported in the
2469 PH/04 Added the $dnslist_matched variable.
2471 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2472 This means they are set thereafter only if the connection becomes
2475 PH/06 Added the client_condition to authenticators so that some can be skipped
2476 by clients under certain conditions.
2478 PH/07 The error message for a badly-placed control=no_multiline_responses left
2479 "_responses" off the end of the name.
2481 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2483 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2484 (without spaces) instead of just copying the configuration text.
2486 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2488 PH/11 Added $max_received_linelength.
2490 PH/12 Added +ignore_defer and +include_defer to host lists.
2492 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2493 way in which PCRE > 7.0 is built.
2495 PH/14 Implemented queue_only_load_latch.
2497 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2498 MAIL command. The effect was to mangle the value on 64-bit systems.
2500 PH/16 Another patch from the Sieve maintainer.
2502 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2504 PH/18 If a system quota error occurred while trying to create the file for
2505 a maildir delivery, the message "Mailbox is full" was not appended to the
2506 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2507 only to a quota excession during the actual writing of the file.
2509 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2510 characters?) which causes problems in log lines. The DN values are now
2511 passed through string_printing() before being added to log lines.
2513 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2514 and InterBase are left for another time.)
2516 PH/21 Added message_body_newlines option.
2518 PH/22 Guard against possible overflow in moan_check_errorcopy().
2520 PH/23 POSIX allows open() to be a macro; guard against that.
2522 PH/24 If the recipient of an error message contained an @ in the local part
2523 (suitably quoted, of course), incorrect values were put in $domain and
2524 $local_part during the evaluation of errors_copy.
2530 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2531 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2534 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2535 issue a MAIL command.
2537 PH/02 In an ACL statement such as
2539 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2541 if a client was not listed at all, or was listed with a value other than
2542 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2543 the condition was not true (as it should be), so access was not denied.
2544 The bug was that the ! inversion was incorrectly passed on to the second
2545 item. This has been fixed.
2547 PH/03 Added additional dnslists conditions == and =& which are different from
2548 = and & when the dns lookup returns more than one IP address.
2550 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2551 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2553 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2554 FSYNC, which compiles an option called disable_fsync that allows for
2555 bypassing fsync(). The documentation is heavily laced with warnings.
2557 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2559 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2560 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2561 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2562 including adding "make clean"; (3) Added -fPIC when compiling the test
2563 dynamically loaded module, to get rid of a warning.
2565 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2566 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2567 = 0s. The bug is otherwise harmless.
2569 PH/07 There was a bug in the dovecot authenticator such that the value of
2570 $auth1 could be overwritten, and so not correctly preserved, after a
2571 successful authentication. This usually meant that the value preserved by
2572 the server_setid option was incorrect.
2574 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2576 PH/09 Installed PCRE release 7.0.
2578 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2579 run for batched SMTP input. It is now run at the start of every message
2580 in the batch. While fixing this I discovered that the process information
2581 (output by running exiwhat) was not always getting set for -bs and -bS
2582 input. This is fixed, and it now also says "batched" for BSMTP.
2584 PH/11 Added control=no_pipelining.
2586 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2587 patch, slightly modified), and move the expansion of helo_data till after
2588 the connection is made in the smtp transport (so it can use these
2591 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2593 PH/14 Added log_selector = +pid.
2595 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2597 PH/16 Add ${if forany and ${if forall.
2599 PH/17 Added dsn_from option to vary the From: line in DSNs.
2601 PH/18 Flush SMTP output before performing a callout, unless control =
2602 no_callout_flush is set.
2604 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2605 was true (the default) a successful delivery failed to delete the retry
2606 item, thus causing premature timeout of the address. The bug is now
2609 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2611 PH/21 Long custom messages for fakedefer and fakereject are now split up
2612 into multiline responses in the same way that messages for "deny" and
2613 other ACL rejections are.
2615 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2616 with slight modification.
2618 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2619 draft, changing the syntax and factoring some duplicate code".
2621 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2622 for deliveries of the second and subsequent messages over the same SMTP
2625 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2626 ${reduce, with only minor "tidies".
2628 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2630 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2631 expansion side effects.
2633 PH/27 When a message times out after an over-quota error from an Exim-imposed
2634 quota, the bounce message says "mailbox is full". This message was not
2635 being given when it was a system quota that was exceeded. It now should
2638 MH/03 Made $recipients available in local_scan(). local_scan() already has
2639 better access to the recipient list through recipients_list[], but
2640 $recipients can be useful in postmaster-provided expansion strings.
2642 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2643 in the case of a MAIL command with additional options following the
2644 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2645 were accidentally chopped off.
2647 PH/29 SMTP synchronization checks are implemented when a command is read -
2648 there is a check that no more input is waiting when there shouldn't be
2649 any. However, for some commands, a delay in an ACL can mean that it is
2650 some time before the response is written. In this time, more input might
2651 arrive, invalidly. So now there are extra checks after an ACL has run for
2652 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2653 pipelining has not been advertised.
2655 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2657 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2658 returned in the SMTP response when smtp_return_error_details was false.
2659 This has been fixed.
2661 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2662 instead of the C I/O that was originally supplied, because problems were
2663 reported on Solaris.
2665 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2666 Exim which did not show up earlier: it was assuming that a call to
2667 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2668 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2669 was a macro that became an assignment, so it seemed to work. This has
2670 changed to a proper function call with a void return, hence the compile
2671 error. Exim's code has been fixed.
2673 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2676 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2678 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2680 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2681 because it turns out to be too big or there is a timeout) there is no
2682 "Completed" line in the log. When some messages of this type were
2683 selected by exigrep, they were listed as "not completed". Others were
2684 picked up by some special patterns. I have improved the selection
2685 criteria to be more general.
2687 PH/38 The host_find_failed option in the manualroute router can now be set
2688 to "ignore", to completely ignore a host whose IP address cannot be
2689 found. If all hosts are ignored, the behaviour is controlled by the new
2690 host_all_ignored option.
2692 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2693 homing or because of multiple MX records with /mx) generated more than
2694 one IP address, and the following item turned out to be the local host,
2695 all the secondary addresses of the first item were incorrectly removed
2696 from the list, along with the local host and any following hosts (which
2697 is what is supposed to happen).
2699 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2700 whoever called Exim into the -H file. In the case of the daemon it was
2701 behaving confusingly. When first started, it used values for whoever
2702 started the daemon, but after a SIGHUP it used the Exim user (because it
2703 calls itself on a restart). I have changed the code so that it now always
2706 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2707 message are rejected with the same error (e.g. no authentication or bad
2708 sender address), and a DATA command is nevertheless sent (as can happen
2709 with PIPELINING or a stupid MUA), the error message that was given to the
2710 RCPT commands is included in the rejection of the DATA command. This is
2711 intended to be helpful for MUAs that show only the final error to their
2714 PH/42 Another patch from the Sieve maintainer.
2716 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2717 Eximstats - Fixed some broken HTML links and added missing column headers
2719 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2720 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2722 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2724 PH/43 Yet another patch from the Sieve maintainer.
2726 PH/44 I found a way to check for a TCP/IP connection going away before sending
2727 the response to the final '.' that terminates a message, but only in the
2728 case where the client has not sent further data following the '.'
2729 (unfortunately, this is allowed). However, in many cases there won't be
2730 any further data because there won't be any more messages to send. A call
2731 to select() can be used: if it shows that the input is "ready", there is
2732 either input waiting, or the socket has been closed. An attempt to read
2733 the next input character can distinguish the two cases. Previously, Exim
2734 would have sent an OK response which the client would never have see.
2735 This could lead to message repetition. This fix should cure that, at
2736 least in a lot of common cases.
2738 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2739 advertised in response to EHLO.
2745 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2746 fixed by 4.65/MH/01 (is this a record?) are fixed:
2748 (i) An empty string was always treated as zero by the numeric comparison
2749 operators. This behaviour has been restored.
2751 (ii) It is documented that the numeric comparison operators always treat
2752 their arguments as decimal numbers. This was broken in that numbers
2753 starting with 0 were being interpreted as octal.
2755 While fixing these problems I realized that there was another issue that
2756 hadn't been noticed. Values of message_size_limit (both the global option
2757 and the transport option) were treated as octal if they started with 0.
2758 The documentation was vague. These values are now always treated as
2759 decimal, and I will make that clear in the documentation.
2765 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2766 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2769 MH/01 Don't check that the operands of numeric comparison operators are
2770 integers when their expansion is in "skipping" mode (fixes bug
2771 introduced by 4.64-PH/07).
2773 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2774 child addresses, Exim now panics and dies. Previously, because the count
2775 is held in a short int, deliveries were likely to be lost. As such a
2776 large number of recipients for a single message is ridiculous
2777 (performance will be very, very poor), I have chosen to impose a limit
2778 rather than extend the field.
2784 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2785 leftover -K file (the existence of which was triggered by #402).
2786 While we were at it, introduced process PID as part of the -K
2787 filename. This should rule out race conditions when creating
2790 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2791 processing considerably. Previous code took too long for large mails,
2792 triggering a timeout which in turn triggers #401.
2794 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2795 in the DK code in transports.c. sendfile() is not really portable,
2796 hence the _LINUX specificness.
2798 TF/01 In the add_headers option to the mail command in an Exim filter,
2799 there was a bug that Exim would claim a syntax error in any
2800 header after the first one which had an odd number of characters
2803 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2804 callout verification, Exim cached a "reject" for the entire domain. This
2805 is correct for most verifications, but it is not correct for a recipient
2806 verification with use_sender or use_postmaster set, because in that case
2807 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2808 case of MAIL FROM:<> rejection from other early rejections (e.g.
2809 rejection of HELO). When verifying a recipient using a non-null MAIL
2810 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2811 Whatever the result of the callout, the value of the domain cache is
2812 left unchanged (for any other kind of callout, getting as far as trying
2813 RCPT means that the domain itself is ok).
2815 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2818 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2819 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2820 session, the daemon ignores ECONNECT errors and logs others; it now
2821 ignores EPIPE as well.
2823 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2824 (quoted-printable decoding).
2826 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2827 later the small subsequent patch to fix an introduced bug.
2829 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2831 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2833 PH/08 An error is now given if message_size_limit is specified negative.
2835 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2836 to be given (somewhat) arbitrary names.
2838 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2841 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2842 miscellaneous code fixes
2844 PH/10 Added the log_reject_target ACL modifier to specify where to log
2847 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2848 hostname. This is wrong, because it relates to the incoming message (and
2849 probably the interface on which it is arriving) and not to the outgoing
2850 callout (which could be using a different interface). This has been
2851 changed to use the value of the helo_data option from the smtp transport
2852 instead - this is what is used when a message is actually being sent. If
2853 there is no remote transport (possible with a router that sets up host
2854 addresses), $smtp_active_hostname is used.
2856 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2857 tweaks were necessary in order to get it to work (see also 21 below):
2858 (a) The code assumed that strncpy() returns a negative number on buffer
2859 overflow, which isn't the case. Replaced with Exim's string_format()
2861 (b) There were several signed/unsigned issues. I just did the minimum
2862 hacking in of casts. There is scope for a larger refactoring.
2863 (c) The code used strcasecmp() which is not a standard C function.
2864 Replaced with Exim's strcmpic() function.
2865 (d) The code set only $1; it now sets $auth1 as well.
2866 (e) A simple test gave the error "authentication client didn't specify
2867 service in request". It would seem that Dovecot has changed its
2868 interface. Fortunately there's a specification; I followed it and
2869 changed what the client sends and it appears to be working now.
2871 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2874 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2875 address A is aliased to B and C, where B exists and C does not. Without
2876 -v the output is "A verified" because verification stops after a
2877 successful redirection if more than one address is generated. However,
2878 with -v the child addresses are also verified. Exim was outputting "A
2879 failed to verify" and then showing the successful verification for C,
2880 with its parentage. It now outputs "B failed to verify", showing B's
2881 parentage before showing the successful verification of C.
2883 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2884 look up a TXT record in a specific list after matching in a combined
2887 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2888 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2889 they consult the DNS. I had assumed they would set it the way they
2890 wanted; and indeed my experiments on Linux seem to show that in some
2891 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2892 To be on the safe side, however, I have now made the interface to
2893 host_find_byname() similar to host_find_bydns(), with an argument
2894 containing the DNS resolver options. The host_find_byname() function now
2895 sets these options at its start, just as host_find_bydns() does. The smtp
2896 transport options dns_qualify_single and dns_search_parents are passed to
2897 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2898 of host_find_byname() use the default settings of RES_DEFNAMES
2899 (qualify_single) but not RES_DNSRCH (search_parents).
2901 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2902 spool_read_header() do less string testing, by means of a preliminary
2903 switch on the second character of optional "-foo" lines. (This is
2904 overdue, caused by the large number of possibilities that now exist.
2905 Originally there were few.) While I was there, I also converted the
2906 str(n)cmp tests so they don't re-test the leading "-" and the first
2907 character, in the hope this might squeeze out yet more improvement.
2909 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2910 flag allowing group syntax was set by the header_syntax check but not
2911 turned off, possible causing trouble later; (2) The flag was not being
2912 set at all for the header_verify test, causing "group"-style headers to
2913 be rejected. I have now set it in this case, and also caused header_
2914 verify to ignore an empty address taken from a group. While doing this, I
2915 came across some other cases where the code for allowing group syntax
2916 while scanning a header line wasn't quite right (mostly, not resetting
2917 the flag correctly in the right place). These bugs could have caused
2918 trouble for malformed header lines. I hope it is now all correct.
2920 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2921 with the "reply" argument non-NULL. The code, however (which originally
2922 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2923 but it didn't always do it. This confused somebody who was copying the
2924 code for some other use. I have removed all the tests.
2926 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2927 feature that was used to support insecure browsers during the U.S. crypto
2928 embargo. It requires special client support, and Exim is probably the
2929 only MTA that supported it -- and would never use it because real RSA is
2930 always available. This code has been removed, because it had the bad
2931 effect of slowing Exim down by computing (never used) parameters for the
2932 RSA_EXPORT functionality.
2934 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2935 authenticator to fail if there's a tab character in the incoming data
2936 (there should never be unless someone is messing about, as it's supposed
2937 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2938 if the connection is using TLS or if the remote IP is the same as the
2939 local IP, and the "valid-client-cert option" if a client certificate has
2942 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2943 authenticators. This can be used for authorization after authentication
2944 succeeds. (In the case of plaintext, it servers for both authentication
2947 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2948 if any retry times were supplied.
2950 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2951 connection, where there is no client IP address to check. In this
2952 situation, the verify now always succeeds.
2954 PH/25 Applied John Jetmore's -Mset patch.
2956 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2958 PH/27 In a string expansion for a processed (not raw) header when multiple
2959 headers of the same name were present, leading whitespace was being
2960 removed from all of them, but trailing whitespace was being removed only
2961 from the last one. Now trailing whitespace is removed from each header
2962 before concatenation. Completely empty headers in a concatenation (as
2963 before) are ignored.
2965 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2966 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2968 PH/29 [Removed. This was a change that I later backed out, and forgot to
2969 correct the ChangeLog entry (that I had efficiently created) before
2970 committing the later change.]
2972 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2973 address errors (4xx response to RCPT) over the same connection as other
2974 messages routed to the same hosts. Such deliveries are always "forced",
2975 so retry times are not inspected. This resulted in far too many retries
2976 for the affected addresses. The effect occurred only when there were more
2977 hosts than the hosts_max_try setting in the smtp transport when it had
2978 the 4xx errors. Those hosts that it had tried were not added to the list
2979 of hosts for which the message was waiting, so if all were tried, there
2980 was no problem. Two fixes have been applied:
2982 (i) If there are any address or message errors in an SMTP delivery, none
2983 of the hosts (tried or untried) are now added to the list of hosts
2984 for which the message is waiting, so the message should not be a
2985 candidate for sending over the same connection that was used for a
2986 successful delivery of some other message. This seems entirely
2987 reasonable: after all the message is NOT "waiting for some host".
2988 This is so "obvious" that I'm not sure why it wasn't done
2989 previously. Hope I haven't missed anything, but it can't do any
2990 harm, as the worst effect is to miss an optimization.
2992 (ii) If, despite (i), such a delivery is accidentally attempted, the
2993 routing retry time is respected, so at least it doesn't keep
2994 hammering the server.
2996 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2997 in ${readsocket because some servers need this prod.
2999 PH/32 Added some extra debug output when updating a wait-xxx database.
3001 PH/33 The hint "could be header name not terminated by colon", which has been
3002 given for certain expansion errors for a long time, was not being given
3003 for the ${if def:h_colon_omitted{... case.
3005 PH/34 The spec says: "With one important exception, whenever a domain list is
3006 being scanned, $domain contains the subject domain." There was at least
3007 one case where this was not true.
3009 PH/35 The error "getsockname() failed: connection reset by peer" was being
3010 written to the panic log as well as the main log, but it isn't really
3011 panic-worthy as it just means the connection died rather early on. I have
3012 removed the panic log writing for the ECONNRESET error when getsockname()
3015 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3016 runs only) independently of the message's sender address. This meant
3017 that, if the 4xx error was in fact related to the sender, a different
3018 message to the same recipient with a different sender could confuse
3019 things. In particular, this can happen when sending to a greylisting
3020 server, but other circumstances could also provoke similar problems.
3021 I have changed the default so that the retry time for these errors is now
3022 based a combination of the sender and recipient addresses. This change
3023 can be overridden by setting address_retry_include_sender=false in the
3026 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3027 remote server are returned as part of bounce messages. This was not
3028 happening for LMTP over a pipe (the lmtp transport), but now it is the
3029 same for both kinds of LMTP.
3031 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3032 in header lines that were in fact CNAMEs. This is no longer the case.
3034 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3035 and queue runs started by the daemon processed all messages. This has
3036 been fixed so that -R and -S can now usefully be given with -q<time>.
3038 PH/40 Import PCRE release 6.7 (fixes some bugs).
3040 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3042 PH/42 Give an error if -q is specified more than once.
3044 PH/43 Renamed the variables $interface_address and $interface_port as
3045 $received_ip_address and $received_port, to make it clear that these
3046 values apply to message reception, and not to the outgoing interface when
3047 a message is delivered. (The old names remain recognized, of course.)
3049 PH/44 There was no timeout on the connect() call when using a Unix domain
3050 socket in the ${readsocket expansion. There now is.
3052 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3053 be meaningful with "accept".
3055 SC/01 Eximstats V1.43
3056 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3058 SC/02 Eximstats V1.44
3059 Use a glob alias rather than an array ref in the generated
3060 parser. This improves both readability and performance.
3062 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3063 Collect SpamAssassin and rejection statistics.
3064 Don't display local sender or destination tables unless
3065 there is data to show.
3066 Added average volumes into the top table text output.
3068 SC/04 Eximstats V1.46
3069 Collect data on the number of addresses (recipients)
3070 as well as the number of messages.
3072 SC/05 Eximstats V1.47
3073 Added 'Message too big' to the list of mail rejection
3074 reasons (thanks to Marco Gaiarin).
3076 SC/06 Eximstats V1.48
3077 Mainlog lines which have GMT offsets and are too short to
3078 have a flag are now skipped.
3080 SC/07 Eximstats V1.49 (Alain Williams)
3081 Added the -emptyok flag.
3083 SC/08 Eximstats V1.50
3084 Fixes for obtaining the IP address from reject messages.
3086 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3087 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3088 whitespace changes from 4.64-PH/27
3090 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3093 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3094 are found, allow negative numbers in numeric criteria)
3096 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3098 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3099 to match changes made in 4.64-PH/43
3101 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3103 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3104 "domain" is actually an IP address, and force a failure. This locks out
3105 those revolvers/nameservers that support "A-for-A" lookups, in
3106 contravention of the specifications.
3108 PH/48 When a host name was looked up from an IP address, and the subsequent
3109 forward lookup of the name timed out, the host name was left in
3110 $sender_host_name, contrary to the specification.
3112 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3113 restricted to single-key lookups, Exim was not diagnosing an error if
3114 * or *@ was used with a query-style lookup.
3116 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3118 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3119 long ago, but noone interested enough thought of it. Let's just say that
3120 the "1.1" means that there are some new functions that weren't there at
3121 some point in the past.
3123 PH/51 Error processing for expansion failure of helo_data from an smtp
3124 transport during callout processing was broken.
3126 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3127 tested/used via the -bh/-bhc/-bs options.
3129 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3130 bug, fixed in subsequent PCRE releases).
3132 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3133 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3135 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3141 SC/01 Use a glob alias rather than an array ref in eximstats generated
3142 parser. This improves both readability and performance.
3144 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3145 Don't display local sender or destination tables in eximstats unless
3146 there is data to show.
3147 Added average volumes into the eximstats top table text output.
3149 SC/03 Collect data on the number of addresses (recipients) as well
3150 as the number of messages in eximstats.
3152 TF/01 Correct an error in the documentation for the redirect router. Exim
3153 does (usually) call initgroups() when daemonizing.
3155 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3156 with consistent privilege compared to when running as a daemon.
3158 TF/03 Note in the spec that $authenticated_id is not set for local
3159 submissions from trusted users.
3161 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3162 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3164 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3165 by adding some example configuration directives to the default
3166 configuration file. A little bit of work is required to uncomment the
3167 directives and define how usernames and passwords are checked, but
3168 there is now a framework to start from.
3170 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3171 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3172 without this. I don't know how relevant this is to other LDAP libraries.
3174 PH/02 Add the verb name to the "unknown ACL verb" error.
3176 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3178 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3180 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3181 directory not expanded when it should be if an expanded home directory
3182 was set for the address (which is overridden by the transport).
3184 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3187 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3188 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3189 because it is too late at that time, and has no effect.
3191 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3192 security issue with \' (bugzilla #107). I could not use the
3193 PQescapeStringConn() function, because it needs a PGconn value as one of
3196 PH/08 When testing addresses using -bt, indicate those final addresses that
3197 are duplicates that would not cause an additional delivery. At least one
3198 person was confused, thinking that -bt output corresponded to deliveries.
3199 (Suppressing duplicates isn't a good idea as you lose the information
3200 about possibly different redirections that led to the duplicates.)
3202 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3203 systems where poll() doesn't work, in particular OS X.
3205 PH/10 Added more information to debugging output for retry time not reached.
3207 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3208 operations in malware.c.
3210 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3213 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3214 syslog_duplication set false, log lines that would normally be written
3215 both the the main log and to the reject log were not written to syslog at
3218 PH/14 In the default configuration, change the use of "message" in ACL warn
3219 statements to "add_header".
3221 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3222 not followed by a command (e.g. "seen endif").
3224 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3225 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3228 PH/17 Added extra conditions to the default value of delay_warning_condition
3232 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3233 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3234 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3237 The Auto-Submitted: and various List- headers are standardised, whereas I
3238 don't think Precedence: ever was.
3240 PH/18 Refactored debugging code in route_finduser() to show more information,
3241 in particular, the error code if getpwnam() issues one.
3243 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3244 This is apparently needed in addition to the PH/07 change above to avoid
3245 any possible encoding problems.
3247 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3248 but not after initializing Perl.
3250 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3251 output them only if debugging. By default they are written stderr,
3252 apparently, which is not desirable.
3254 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3257 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3260 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3262 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3263 authenticated or an ident call has been made. Suppress the default
3264 values for $authenticated_id and $authenticated_sender (but permit -oMai
3265 and -oMas) when testing with -bh.
3267 PH/24 Re-jigged the order of the tests in the default configuration so that the
3268 tests for valid domains and recipients precede the DNS black list and CSA
3269 tests, on the grounds that those ones are more expensive.
3271 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3272 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3273 This bug exists in every version of Exim that I still have, right back to
3276 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3277 However, an attempt to turn on case-sensitivity in a regex key by
3278 including (?-i) didn't work because the subject string was already
3279 lowercased, and the effects were non-intuitive. It turns out that a
3280 one-line patch can be used to allow (?-i) to work as expected.
3286 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3287 other effects) broke the use of negated acl sub-conditions.
3289 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3292 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3293 "Deny" causes Exim to reject the incoming connection with a 554 error.
3294 Unfortunately, if there is a major crisis, such as a disk failure,
3295 tcp-wrappers gives "deny", whereas what one would like would be some
3296 kind of temporary error. A kludge has been added to help with this.
3297 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3298 554 error is used if errno is still zero or contains ENOENT (which occurs
3299 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3302 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3304 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3305 errors. Otherwise a message that provokes a temporary error (when other
3306 messages do not) can cause a whole host to time out.
3308 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3309 addresses were routed directly to files or pipes from a redirect router.
3310 File deliveries just didn't batch; pipe deliveries might have suffered
3313 PH/06 A failure to get a lock for a hints database would erroneously always say
3314 "Failed to get write lock", even when it was really a read lock.
3316 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3317 of 0600. This has been changed to use the value of the lockfile_mode
3318 option (which defaults to 0600).
3320 PH/08 Applied small patch from the Sieve maintainer.
3322 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3323 folder from quota calculations, a direct delivery into this folder messed
3324 up the contents of the maildirsize file. This was because the regex was
3325 used only to exclude .Trash (or whatever) when the size of the mailbox
3326 was calculated. There was no check that a delivery was happening into an
3327 excluded directory. This bug has been fixed by ignoring all quota
3328 processing for deliveries into excluded directories.
3330 PH/10 Added the maildirfolder_create_regex option to appendfile.
3336 PH/01 The code for finding all the local interface addresses on a FreeBSD
3337 system running IPv6 was broken. This may well have applied to all BSD
3338 systems, as well as to others that have similar system calls. The broken
3339 code found IPv4 interfaces correctly, but gave incorrect values for the
3340 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3341 that it would not match correctly against @[] and not recognize the IPv6
3344 PH/02 The ipliteral router was not recognizing addresses of the form user@
3345 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3347 PH/03 Added disable_ipv6.
3349 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3350 decoded form, because it is most often used to construct To: headers
3351 lines in autoreplies, and the decoded form may well be syntactically
3352 invalid. However, $reply_address has leading white space removed, and all
3353 newlines turned into spaces so that the autoreply transport does not
3356 PH/05 If group was specified without a user on a router, and no group or user
3357 was specified on a transport, the group from the router was ignored.
3359 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3360 for visible compile-time settings that can be used to change these
3361 numbers, for those that want even more. Backwards compatibility with old
3362 spool files has been maintained. However, going back to a previous Exim
3363 release will lost any variables that are in spool files.
3365 PH/07 Two small changes when running in the test harness: increase delay when
3366 passing a TCP/IP connection to a new process, in case the original
3367 process has to generate a bounce, and remove special handling of
3368 127.0.0.2 (sic), which is no longer necessary.
3370 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3371 be the same on different OS.
3373 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3376 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3377 whether --show-vars was specified or not
3379 JJ/02 exipick: Added support for new ACL variable spool format introduced
3382 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3383 syntactically invalid From: or Reply-to: line, and a filter used this to
3384 generate an autoreply, and therefore failed to obtain an address for the
3385 autoreply, Exim could try to deliver to a non-existent relative file
3386 name, causing unrelated and misleading errors. What now happens is that
3387 it logs this as a hard delivery error, but does not attempt to create a
3390 PH/11 The exinext utility has a -C option for testing purposes, but although
3391 the given file was scanned by exinext itself; it wasn't being passed on
3392 when Exim was called.
3394 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3395 an end-of-file indication when reading a command response.
3397 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3398 compiled. In many other places in Exim, IPv6 addresses are always
3399 recognized, so I have changed this. It also means that IPv4 domain
3400 literals of the form [IPV4:n.n.n.n] are now always recognized.
3402 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3403 used if the router is not running as root, for example, when verifying at
3404 ACL time, or when using -bh. The debugging output from this situation was
3405 non-existent - all you got was a failure to exec. I have made two
3408 (a) Failures to set uid/gid, the current directory, or a process leader
3409 in a subprocess such as that created by queryprogram now generate
3410 suitable debugging output when -d is set.
3412 (b) The queryprogram router detects when it is not running as root,
3413 outputs suitable debugging information if -d is set, and then runs
3414 the subprocess without attempting to change uid/gid.
3416 PH/15 Minor change to Makefile for building test_host (undocumented testing
3419 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3420 additional section of a DNS packet that returns MX or SRV records.
3421 Instead, it always explicitly searches for A/AAAA records. This avoids
3422 major problems that occur when a DNS server includes only records of one
3423 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3424 fixed another bug: if SRV records were looked up and the corresponding
3425 address records were *not* found in the additional section, the port
3426 values from the SRV records were lost.
3428 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3429 using the correct key (the original address) when searching the retry
3430 rules in order to find which one to use for generating the retry hint.
3432 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3433 adding the default one. Similarly, if it contains a Reply-To: header, the
3434 errors_reply_to option, if set, is not used.
3436 PH/19 When calculating a retry time, Exim used to measure the "time since
3437 failure" by looking at the "first failed" field in the retry record. Now
3438 it does not use this if it is later than than the arrival time of the
3439 message. Instead it uses the arrival time. This makes for better
3440 behaviour in cases where some deliveries succeed, thus re-setting the
3441 "first failed" field. An example is a quota failure for a huge message
3442 when small messages continue to be delivered. Without this change, the
3443 "time since failure" will always be short, possible causing more frequent
3444 delivery attempts for the huge message than are intended.
3445 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3447 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3448 $1, $2, $3) because the numerical variables can be reset during some
3449 expansion items (e.g. "match"), thereby losing the authentication data.
3451 PH/21 Make -bV show the size of off_t variables so that the test suite can
3452 decide whether to run tests for quotas > 2G.
3454 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3455 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3456 filecount value is greater than 2G or if a quota value is greater than 2G
3457 on a system where the size of off_t is not greater than 4, a panic error
3460 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3461 never match. The debug and -bh output now contains an explicit error
3462 message indicating a malformed IPv4 address or mask.
3464 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3465 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3466 PH/23 above applies.
3468 PH/25 Do not write to syslog when running in the test harness. The only
3469 occasion when this arises is a failure to open the main or panic logs
3470 (for which there is an explicit test).
3472 PH/26 Added the /no_tell option to "control=freeze".
3474 PH/27 If a host name lookup failed very early in a connection, for example, if
3475 the IP address matched host_lookup and the reverse lookup yielded a name
3476 that did not have a forward lookup, an error message of the form "no IP
3477 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3478 could be logged. Now it outputs the IP address instead of "NULL".
3480 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3481 allows the sender and the authenticated sender to be set when
3482 submitting a message from within Exim. Since child_open_exim() is
3483 documented for local_scan(), the new function should be too.
3485 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3486 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3487 results in an empty string is now treated as unset.
3489 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3491 PH/31 Added sender_verify_fail logging option.
3493 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3494 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3495 tidied the source and removed it altogether.
3497 PH/33 When a queue run was abandoned because the load average was too high, a
3498 log line was always written; now it is written only if the queue_run log
3499 selector is set. In addition, the log line for abandonment now contains
3500 information about the queue run such as the pid. This is always present
3501 in "start" and "stop" lines but was omitted from the "abandon" line.
3503 PH/34 Omit spaces between a header name and the colon in the error message that
3504 is given when verify = headers_syntax fails (if there are lots of them,
3505 the message gets confusing).
3507 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3508 names, as there are now some PTR records that contain slashes. This check
3509 is only to protect against broken name servers that fall over on strange
3510 characters, so the fact that it applies to all lookups doesn't matter.
3512 PH/36 Now that the new test suite is complete, we can remove some of the
3513 special code in Exim that was needed for the old test suite. For example,
3514 sorting DNS records because real resolvers return them in an arbitrary
3515 order. The new test suite's fake resolver always returns records in the
3518 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3519 bounces) except when queue_only is set, to avoid logging races between
3520 the different processes.
3522 PH/38 Panic-die if .include specifies a non-absolute path.
3524 PH/39 A tweak to the "H" retry rule from its user.
3526 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3527 a label. They prevented compilation on older perls.
3529 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3530 a warning to be raised on newish perls.
3532 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3533 on queue. Changes to match documented behaviour of showing count of
3534 messages matching specified criteria.
3536 PH/40 Changed the default ident timeout from 30s to 5s.
3538 PH/41 Added support for the use of login_cap features, on those BSD systems
3539 that have them, for controlling the resources used by pipe deliveries.
3541 PH/42 The content-scanning code uses fopen() to create files in which to put
3542 message data. Previously it was not paying any attention to the mode of
3543 the files. Exim runs with umask(0) because the rest of the code creates
3544 files with open(), and sets the required mode explicitly. Thus, these
3545 files were ending up world-writeable. This was not a big issue, because,
3546 being within the spool directory, they were not world-accessible. I have
3547 created a function called modefopen, which takes an additional mode
3548 argument. It sets umask(777), creates the file, chmods it to the required
3549 mode, then resets the umask. All the relevant calls to fopen() in the
3550 content scanning code have been changed to use this function.
3552 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3553 to 24 hours. This avoids potential overflow problems when processing G
3554 and H retry rules. I suspect nobody ever tinkers with this value.
3556 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3558 PH/45 When the plaintext authenticator is running as a client, the server's
3559 challenges are checked to ensure they are valid base64 strings. By
3560 default, the authentication attempt is cancelled if an invalid string is
3561 received. Setting client_ignore_invalid_base64 true ignores these errors.
3562 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3563 they are received. Thus, the responses can be made to depend on the
3564 challenges. If an invalid string is ignored, an empty string is placed in
3567 PH/46 Messages that are created by the autoreply transport now contains a
3568 References: header, in accordance with RFCs 2822 and 3834.
3570 PH/47 Added authenticated_sender_force to the smtp transport.
3572 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3574 PH/49 Installed latest patch from the Sieve maintainer.
3576 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3577 was also set, the appendfile transport was unnecessarily scanning a
3578 directory of message files (e.g. for maildir delivery) to find the count
3579 of files (along with the size), even though it did not need this
3580 information. It now does the scan only if it needs to find either the
3581 size of the count of files.
3583 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3585 PH/52 Two bugs concerned with error handling when the smtp transport is
3588 (i) Exim was not creating retry information for temporary errors given
3589 for individual recipients after the DATA command when the smtp transport
3590 was used in LMTP mode. This meant that they could be retried too
3591 frequently, and not timed out correctly.
3593 (ii) Exim was setting the flag that allows error details to be returned
3594 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3595 recipients that were returned after the DATA command.
3597 PH/53 This is related to PH/52, but is more general: for any failing address,
3598 when detailed error information was permitted to be returned to the
3599 sender, but the error was temporary, then after the final timeout, only
3600 "retry timeout exceeded" was returned. Now it returns the full error as
3601 well as "retry timeout exceeded".
3603 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3604 do this, and (what is worse) MTAs that accept it.
3606 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3607 will now be deprecated.
3609 PH/56 New os.c-cygwin from the Cygwin maintainer.
3611 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3612 formats (previously only available in exim formats via -bpr, -bpru,
3613 and -bpra. Now also available in native and exiqgrep formats)
3615 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3616 with very large, slow to parse queues
3618 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3620 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3622 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3623 responses to authentication challenges, though it was showing the
3624 challenges; (ii) I've removed the CR characters from the debug output for
3627 PH/58 Allow for the insertion of a newline as well as a space when a string
3628 is turned into more than one encoded-word during RFC 2047 encoding. The
3629 Sieve code now uses this.
3631 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3632 data_4xx, lost_connection, tls_required.
3634 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3635 message was being sent as an SMTP response.
3637 PH/61 Add -l and -k options to exicyclog.
3639 PH/62 When verifying, if an address was redirected to one new address, so that
3640 verification continued, and the new address failed or deferred after
3641 having set something in $address_data, the value of $address_data was not
3642 passed back to the ACL. This was different to the case when no
3643 redirection occurred. The value is now passed back in both cases.
3645 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3646 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3647 use login_cap.h, so on its own it isn't the distinguishing feature. The
3648 new name refers directly to the setclassresources() function.
3650 PH/65 Added configuration files for NetBSD3.
3652 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3654 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3655 is preferred over IPv4.
3657 PH/68 The bounce_return_message and bounce_return_body options were not being
3658 honoured for bounces generated during the reception of non-SMTP messages.
3659 In particular, this applied to messages rejected by the ACL. This bug has
3660 been fixed. However, if bounce_return_message is true and bounce_return_
3661 body is false, the headers that are returned for a non-SMTP message
3662 include only those that have been read before the error was detected.
3663 (In the case of an ACL rejection, they have all been read.)
3665 PH/69 The HTML version of the specification is now built in a directory called
3666 spec_html instead of spec.html, because the latter looks like a path with
3667 a MIME-type, and this confuses some software.
3669 PH/70 Catch two compiler warnings in sieve.c.
3671 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3672 function verify_get_ident() calls ip_connect() to connect a socket, but
3673 if the "connect()" function timed out, ip_connect() used to close the
3674 socket. However, verify_get_ident() also closes the socket later, and in
3675 between Exim writes to the log, which may get opened at this point. When
3676 the socket was closed in ip_connect(), the log could get the same file
3677 descriptor number as the socket. This naturally causes chaos. The fix is
3678 not to close the socket in ip_connect(); the socket should be closed by
3679 the function that creates it. There was only one place in the code where
3680 this was missing, in the iplookup router, which I don't think anybody now
3681 uses, but I've fixed it anyway.
3683 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3684 well as to direct DNS lookups. Otherwise the handling of names in host
3685 lists is inconsistent and therefore confusing.
3691 PH/01 Two changes to the default runtime configuration:
3693 (1) Move the checks for relay_from_hosts and authenticated clients from
3694 after to before the (commented out) DNS black list checks.
3696 (2) Add control=submission to the relay_from_hosts and authenticated
3697 clients checks, on the grounds that messages accepted by these
3698 statements are most likely to be submissions.
3700 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3702 (1) Generate an error if the third argument for the ${prvs expansion is
3705 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3708 (3) Reset the variables that are obtained from the first argument of
3709 ${prvscheck and used in the second argument before leaving the code,
3710 because their memory is reclaimed, so using them afterwards may do
3713 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3714 one (it's much easier than Tom thought :-).
3716 (5) Because of (4), we can now allow for the use of $prvscheck_result
3717 inside the third argument.
3719 PH/03 For some reason, the default setting of PATH when running a command from
3720 a pipe transport was just "/usr/bin". I have changed it to
3723 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3724 anything to be listed in the output from -bV.
3726 PH/05 When a filter generated an autoreply, the entire To: header line was
3727 quoted in the delivery log line, like this:
3729 => >A.N.Other <ano@some.domain> <original@ddress> ...
3731 This has been changed so that it extracts the operative address. There
3732 may be more than one such address. If so, they are comma-separated, like
3735 => >ano@some.domain,ona@other.domain <original@ddress> ...
3737 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3738 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3739 not being looked up in the rDNS to get a host name, Exim was showing the
3740 IP address twice in Received: lines, even though the IP addresses were
3741 identical. For example:
3743 Received: from [1.2.3.4] (helo=[1.2.3.4])
3745 However, if the real host name was known, it was omitting the HELO data
3746 if it matched the actual IP address. This has been tidied up so that it
3747 doesn't show the same IP address twice.
3749 PH/07 When both +timestamp and +memory debugging was on, the value given by
3750 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3751 called by the debug printing, thereby overwriting the timestamp buffer.
3752 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3754 PH/08 When the original message was included in an autoreply transport, it
3755 always said "this is a copy of the message, including all the headers",
3756 even if body_only or headers_only was set. It now gives an appropriate
3759 PH/09 Applied a patch from the Sieve maintainer which:
3761 o fixes some comments
3762 o adds the (disabled) notify extension core
3763 o adds some debug output for the result of if/elsif tests
3764 o points to the current vacation draft in the documentation
3765 and documents the missing references header update
3769 o fixes a bug in processing the envelope test (when testing
3770 multiple envelope elements, the last element determined the
3773 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3774 Electronic Mail") by including:
3776 Auto-submitted: auto-generated
3778 in the messages that it generates (bounce messages and others, such as
3779 warnings). In the case of bounce messages for non-SMTP messages, there was
3780 also a typo: it was using "Auto_submitted" (underscore instead of
3781 hyphen). Since every message generated by Exim is necessarily in response
3782 to another message, thes have all been changed to:
3784 Auto-Submitted: auto-replied
3786 in accordance with these statements in the RFC:
3788 The auto-replied keyword:
3790 - SHOULD be used on messages sent in direct response to another
3791 message by an automatic process,
3793 - MUST NOT be used on manually-generated messages,
3795 - MAY be used on Delivery Status Notifications (DSNs) and Message
3796 Disposition Notifications (MDNs),
3798 - MUST NOT be used on messages generated by automatic or periodic
3799 processes, except for messages which are automatic responses to
3802 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3803 to the default Received: header definition.
3805 PH/12 Added log selector acl_warn_skipped (default on).
3807 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3808 variables because (a) they are in the wrong storage pool and (b) even if
3809 they were copied, it wouldn't work properly because of the caching.
3811 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3812 checking when decoding. Apparently there are clients that generate
3813 overlong encoded strings. Why am I not surprised?
3815 PH/15 If the first argument of "${if match_address" was not empty, but did not
3816 contain an "@" character, Exim crashed. Now it writes a panic log message
3817 and treats the condition as false.
3819 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3821 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3822 extension "envelope-auth". The code is finished and in agreement with
3823 other implementations, but there is no documentation so far and in fact,
3824 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3825 not changing the active code.
3827 Print executed "if" and "elsif" statements when debugging is used. This
3828 helps a great deal to understand what a filter does.
3830 Document more things not specified clearly in RFC3028. I had all this
3831 sorted out, when out of a sudden new issues came to my mind. Oops."
3833 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3836 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3837 canonical form (as documented). However, after a host name lookup from
3838 the IP address, check_host() was doing a simple string comparison with
3839 addresses acquired from the DNS when checking that the found name did
3840 have the original IP as one of its addresses. Since any found IPv6
3841 addresses are likely to be in abbreviated form, the comparison could
3842 fail. Luckily, there already exists a function for doing the comparison
3843 by converting both addresses to binary, so now that is used instead of
3844 the text comparison.
3846 PH/20 There was another similar case to PH/19, when a complete host name was
3847 given in a host list; looking up its IP address could give an abbreviated
3848 form, whereas the current host's name might or might not be abbreviated.
3849 The same fix has been applied.
3855 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3856 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3859 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3860 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3862 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3864 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3865 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3866 Also, for "Auto-Submitted", treat anything other than "no" as
3867 non-personal, in accordance with RFC 3834. (Previously it treated
3868 anything starting "auto-" as non-personal.)
3870 TF/01 The control=submission/name=... option had a problem with syntax
3871 errors if the name included a slash character. The /name= option
3872 now slurps the rest of the string, so it can include any characters
3873 but it must come last in the list of options (after /sender_retain
3876 PH/05 Some modifications to the interface to the fake nameserver for the new
3884 TK/01 Added the "success_on_redirect" address verification option. See
3885 NewStuff for rationale and an example.
3887 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3889 PH/02 Patch to exigrep to allow it to work on syslog lines.
3891 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3892 fread() to skip over the body file's header line, because in Cygwin the
3893 header line is locked and is inaccessible.
3895 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3896 co-exist for some time) to make it clear that it is the Exim ID that is
3897 referenced, not the Message-ID: header line.
3899 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3900 string_format() function, because snprintf() does not exist on all
3903 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3904 ${stat: expansion item.
3906 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3907 protocol synchronization error", to keep the pedants happy.
3909 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3910 well as for IRIX systems, when gcc is being used. See the host.c source
3913 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3915 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3918 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3919 transport and to the smtp transport in LMTP mode.
3921 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3923 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3924 run a filter in a subprocess. This could lead to confusion in subsequent
3925 lookups in the parent process. There should also be a search_tidyup() at
3926 the end of the subprocess.
3928 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3929 only if the host matched helo_try_verify_hosts, which caused the
3930 verification to occur when the EHLO/HELO command was issued. The ACL just
3931 tested the remembered result. Now, if a previous verification attempt has
3932 not happened, "verify = helo" does it there and then.
3934 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3936 TK/03 Fix log output including CR from clamd.
3938 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3939 exist provoked a memory error which could cause a segfault.
3941 PH/15 Installed PCRE 6.2
3943 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3945 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3946 of the problem. Specifically, suggested +O2 rather than +O1 for the
3949 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3951 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3952 record, the port was not correctly logged when the outgoing_port log
3953 selector was set (it logged the transort's default port).
3955 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3956 fallback_hosts, and "hosts" in the smtp transport.
3958 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3959 host errors such as "Connection refused".
3961 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3962 authentication with radiusclient 0.4.9:
3964 - Error returned from rc_read_config was caught wrongly
3965 - Username/password not passed on to radius server due to wrong length.
3967 The presumption is that some radiusclient API changes for 4.51/PH/17
3968 were not taken care of correctly. The code is still untested by me (my
3969 Linux distribution still has 0.3.2 of radiusclient), but it was
3970 contributed by a Radius user.
3972 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3973 expanding the "port" option of the smtp transport.
3975 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3976 while reading a MIME header. Thanks to Tom Hughes for a patch.
3978 PH/24 Include config.h inside local_scan.h so that configuration settings are
3981 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3982 that in an ACL for RCPT (for example), you can examine exactly what was
3985 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3986 commands, but it was not correctly comparing the address with the actual
3987 client host address. Thus, it would show the EHLO address in Received:
3988 header lines when this was not necessary.
3990 PH/27 Added the % operator to ${eval:}.
3992 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3993 it should be ignoring failures (because with -C, for example, it has lost
3994 privilege). It wasn't ignoring creation failures other than "already
3997 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4000 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4001 because an input line was too long, either on its own, or by virtue of
4002 too many continuations, the temporary file was not being removed, and the
4003 return code was incorrect.
4005 PH/31 Missing "BOOL" in function definition in filtertest.c.
4007 PH/32 Applied Sieve patches from the maintainer.
4009 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4011 PH/33 Added "verify = not_blind".
4013 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4014 Local/Makefile (with some defaults set). These are used in built scripts
4015 such as exicyclog, but they have never been used in the exim_install
4016 script (though there are many overriding facilities there). I have
4017 arranged that the exim_install script now takes note of these two
4020 PH/35 Installed configuration files for Dragonfly.
4022 PH/36 When a locally submitted message by a trusted user did not contain a
4023 From: header, and the sender address was obtained from -f or from an SMTP
4024 MAIL command, and the trusted user did not use -F to supply a sender
4025 name, $originator_name was incorrectly used when constructing a From:
4026 header. Furthermore, $originator_name was used for submission mode
4027 messages from external hosts without From: headers in a similar way,
4028 which is clearly wrong.
4030 PH/37 Added control=suppress_local_fixups.
4032 PH/38 When log_selector = +received_sender was set, and the addition of the
4033 sender made the log line's construction buffer exactly full, or one byte
4034 less than full, an overflow happened when the terminating "\n" was
4037 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4038 when the result of a list match is failure because a DNS lookup failed.
4040 PH/40 RM_COMMAND is now used in the building process.
4042 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4043 the "build-* directories that it finds.
4045 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4046 address in a domain literal was a prefix of an interface address.
4048 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4049 when verifying a sender address, unless rewrite_headers is false.
4051 PH/44 Wrote a long comment about why errors_to addresses are verified as
4052 recipients, not senders.
4054 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4055 the ratelimit ACL was added.
4057 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4059 PH/46 Added extra information about PostgreSQL errors to the error string.
4061 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4062 avoiding the need to install special zones in a real server. This is
4063 backwards compatible; if it can't find the fake resolver, it drops back.
4064 Thus, both old and new test suites can be run.
4066 TF/02 Added util/ratelimit.pl
4068 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4069 clock is set back in time.
4071 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4072 Candler <B.Candler@pobox.com>.
4074 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4075 OK for addresses that are the result of redirections.
4077 PH/48 A number of further additions for the benefit of the new test suite,
4078 including a fake gethostbyname() that interfaces to the fake DNS resolver
4081 TF/06 The fix for widen_domains has also been applied to qualify_single and
4082 search_parents which are the other dnslookup options that can cause
4085 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4088 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4090 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4091 service" if it encountered an MX record with an empty target hostname.
4092 The message is now "an MX or SRV record indicated no SMTP service".
4094 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4095 if the DNS of the sending site is misconfigured. This is quite a
4096 common situation. This change restores the behaviour of treating a
4097 helo verification defer as a failure.
4099 PH/51 If self=fail was set on a router, the bounce message did not include the
4100 actual error message.
4106 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4108 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4109 command itself ended in error, the underlying message about the transport
4110 filter timeout was being overwritten with the pipe command error. Now the
4111 underlying error message should be appended to the second error message.
4113 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4115 PH/02 Reduce the amount of output that "make" produces by default. Full output
4116 can still be requested.
4118 PH/03 The warning log line about a condition test deferring for a "warn" verb
4119 was being output only once per connection, rather than after each
4120 occurrence (because it was using the same function as for successful
4121 "warn" verbs). This seems wrong, so I have changed it.
4123 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4124 it should not have, which might have caused a crash in the right
4125 circumstances, but probably never did.
4127 PH/04 Installed a modified version of Tony Finch's patch to make submission
4128 mode fix the return path as well as the Sender: header line, and to
4129 add a /name= option so that you can make the user's friendly name appear
4132 TF/03 Added the control = fakedefer ACL modifier.
4134 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4135 Mark Lowes for thorough testing.
4137 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4139 TK/03 Merged latest SRS patch from Miles Wilton.
4141 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4142 with the definition in sysexits.h (which is #included earlier).
4143 Fortunately, Exim does not actually use EX_OK. The code used to try to
4144 preserve the sysexits.h value, by assuming that macro definitions were
4145 scanned for macro replacements. I have been disabused of this notion,
4146 so now the code just undefines EX_OK before #including unistd.h.
4148 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4149 in the smtp transport. When a block could not be written in a single
4150 write() function, the timeout was being re-applied to each part-write.
4151 This seems wrong - if the receiver was accepting one byte at a time it
4152 would take for ever. The timeout is now adjusted when this happens. It
4153 doesn't have to be particularly precise.
4155 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4156 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4158 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4159 without a domain if the check to <postmaster@domain> fails.
4161 SC/01 Eximstats: added -xls and the ability to specify output files
4162 (patch written by Frank Heydlauf).
4164 SC/02 Eximstats: use FileHandles for outputting results.
4166 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4168 SC/04 Eximstats: fixed display of large numbers with -nvr option
4170 SC/05 Eximstats: fixed merging of reports with empty tables.
4172 SC/06 Eximstats: added the -include_original_destination flag
4174 SC/07 Eximstats: removed tabs and trailing whitespace.
4176 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4178 TK/06 MBOX spool code: Add real "From " MBOX separator line
4179 so the .eml file is really in mbox format (even though
4180 most programs do not really care). Patch from Alex Miller.
4182 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4183 The latter is generated from $received_to and is only set if the
4184 message has one envelope recipient. SA can use these headers,
4185 obviously out-of-the-box. Patch from Alex Miller.
4187 PH/08 The ${def test on a variable was returning false if the variable's
4188 value was "0", contrary to what the specification has always said!
4189 The result should be true unless the variable is empty.
4191 PH/09 The syntax error of a character other than { following "${if
4192 def:variable_name" (after optional whitespace) was not being diagnosed.
4193 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4194 accidental colon was present, for example, could give incorrect results.
4196 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4197 result is used (not including appendfile, where other changes are about
4200 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4201 This involved changing a lot of size variables from int to off_t. It
4202 should work with maildirs and everything.
4204 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4205 spamd dying while we are connected to it.
4207 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4210 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4211 function for BDB 4.3.
4213 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4215 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4216 the sources for PCRE have been re-arranged and more files are now
4219 PH/15 The code I had for printing potentially long long variables in PH/11
4220 above was not the best (it lost precision). The length of off_t variables
4221 is now inspected at build time, and an appropriate printing format (%ld
4222 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4223 to be "long long int" or "long int". This is needed for the internal
4224 formatting function string_vformat().
4226 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4227 the configuration file to be ":syslog", then the script "guesses" where
4228 the logs files are, rather than using the compiled in default. In our
4229 case the guess is not the same as the compiled default, so the script
4230 suddenly stopped working when I started to use syslog. The patch checks
4231 to see if log_file_path is "". If so, it attempts to read it from exim
4232 with no configuration file to get the compiled in version, before it
4233 falls back to the previous guessing code."
4235 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4236 implementing BATV in an Exim configuration. See NewStuff for the gory
4239 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4240 Makefile that are specific to HP-UX.
4242 PH/18 If the "use_postmaster" option was set for a recipient callout together
4243 with the "random" option, the postmaster address was used as the MAIL
4244 FROM address for the random test, but not for the subsequent recipient
4245 test. It is now used for both.
4247 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4248 patch removes a few documentation additions to RFC 3028, because the
4249 latest draft now contains them. It adds the new en;ascii-case comparator
4250 and a new error check for 8bit text in MIME parts. Comparator and
4251 require names are now matched exactly. I enabled the subaddress
4252 extension, but it is not well tested yet (read: it works for me)."
4254 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4255 rework some of the code of TK/09 above to avoid the hardwired use of
4256 "%lld" and "long long". Replaced the call to snprintf() with a call to
4259 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4260 records point to non-existent hosts", "retry timeout exceeded", and
4261 "retry time not reached for any host after a long failure period".
4263 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4264 experimental DomainKeys support:
4266 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4267 (2) On an error such as an illegally used "control", the wrong name for
4268 the control was given.
4270 These problems did NOT occur unless DomainKeys support was compiled.
4272 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4274 PH/24 Added ${if match_ip condition.
4276 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4277 databases so that it will be absolutely obvious if a crash occurs in the
4278 DB library. This is a regular occurrence (often caused by mis-matched
4281 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4282 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4283 picked up on a user's system that detects such things. There doesn't seem
4284 to be a gcc warning option for this - only an attribute that has to be
4285 put on the function's prototype. It seems that in Fedora Core 4 they have
4286 set this on a number of new functions. No doubt there will be more in due
4289 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4290 specify a transport. However, if an address that was verified by such a
4291 router was the subject of a callout, Exim crashed because it tried to
4292 read the rcpt_include_affixes from the non-existent transport. Now it
4293 just assumes that the setting of that option is false. This bug was
4294 introduced by 4.51/PH/31.
4296 PH/28 Changed -d+all to exclude +memory, because that information is very
4297 rarely of interest, but it makes the output a lot bigger. People tend to
4298 do -d+all out of habit.
4300 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4301 code in os-type was giving problems when libc.so lives in lib64, like on
4304 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4305 aren't the modern standard, and it seems that some systems' include files
4306 don't always have them. Exim was already checking for some of the newer
4307 ones like T_AAAA, and defining it itself. I've added checks for all the
4308 record types that Exim uses.
4310 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4311 not automatically generating a new one, as it is supposed to. This
4312 prevented TLS from working. If the file did exist, but contained invalid
4313 data, a new version was generated, as expected. It was only the case of a
4314 non-existent file that was broken.
4316 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4317 with a change in libdomainkeys > 0.64.
4319 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4320 from DNS. If the selector record carries the flag, it now has
4321 precedence over the domain-wide flag.
4323 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4325 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4326 the use of a port name that isn't defined in /etc/services) occurred, the
4327 message was deferred as in a normal delivery, and thus remained on the
4328 spool, instead of being failed because of the mua_wrapper setting. This
4329 is now fixed, and I tidied up some of the mua_wrapper messages at the
4332 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4333 the messages in a hash of arrays rather than using individual hashes.
4334 This is a bit cleaner and results in dramatic memory savings, albeit
4335 at a slight CPU cost.
4337 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4338 as requested by Marc Sherman.
4340 SC/10 Eximstats: added histograms for user specified patterns as requested
4343 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4345 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4346 fopen() in the content-scanning modules that did not already have it.
4352 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4353 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4355 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4357 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4359 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4360 to test Sieve filters that use "vacation".
4362 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4363 that changes the way the GnuTLS parameters are stored in the cache file.
4364 The new format can be generated externally. For backward compatibility,
4365 if the data in the cache doesn't make sense, Exim assumes it has read an
4366 old-format file, and it generates new data and writes a new file. This
4367 means that you can't go back to an older release without removing the
4370 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4371 work if there are any delivery delays because "one_time" forces the
4372 parent to be marked "delivered", so its unseen clone is never tried
4373 again. For this reason, Exim now forbids the simultaneous setting of
4376 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4377 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4378 entry if you want to know the details. The fix, however, neglected to
4379 consider the case when local delivery batching is involved. The test for
4380 "previously delivered" was not happening when checking to see if an
4381 address could be batched with a previous (undelivered) one; under
4382 certain circumstances this could lead to multiple deliveries to the same
4385 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4386 in its include files, and this causes problems building Exim.
4388 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4389 header_syntax) but Exim was just ignoring anything given after a slash.
4390 In particular, this caused confusion with an attempt to use "verify =
4391 reverse_host_lookup/defer_ok". An error is now given when options are
4392 supplied for verify items that do not have them. (Maybe reverse_host_
4393 lookup should have a defer_ok option, but that's a different point.)
4395 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4396 defined by RFC 821) to 2048, because there were problems with some AUTH
4397 commands, and RFC 1869 says the size should be increased for extended
4398 SMTP commands that take arguments.
4400 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4403 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4404 "unknown" error; now it says that the functionality isn't in the binary.
4406 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4407 an address' error message when a string expansion fails (syntax or
4408 whatever). Otherwise the password may appear in the log. Following change
4409 PH/42 below, there is no longer a chance of it appearing in a bounce
4412 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4414 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4415 of its addresses was ever tried. (Bugzilla bug #2.)
4417 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4418 the result incorrectly in the debug output. (It correctly added a newline
4419 to what was transported.)
4421 TF/01 Added $received_time.
4423 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4424 commented out examples of how to interface to a virus scanner and to
4425 SpamAssassin. Also added commented examples of av_scanner and
4426 spamd_address settings.
4428 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4429 and controls are allowed in which ACLs. There were a couple of minor
4430 errors. Some of the entries in the conditions table (which is a table of
4431 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4432 negation of where the condition IS allowed.
4434 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4436 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4437 header file does not have a version number, so I've had to invent a new
4438 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4439 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4440 radiusclient), but it was contributed by a Radius user.
4442 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4443 files or directories, for OpenSSL.
4445 PH/19 When an Exim process that is running as root has to create an Exim log
4446 file, it does so in a subprocess that runs as exim:exim so as to get the
4447 ownership right at creation (otherwise, other Exim processes might see
4448 the file with the wrong ownership). There was no test for failure of this
4449 fork() call, which would lead to the process getting stuck as it waited
4450 for a non-existent subprocess. Forks do occasionally fail when resources
4451 run out. I reviewed all the other calls to fork(); they all seem to check
4454 PH/20 When checking for unexpected SMTP input at connect time (before writing
4455 the banner), Exim was not dealing correctly with a non-positive return
4456 from the read() function. If the client had disconnected by this time,
4457 the result was a log entry for a synchronization error with an empty
4458 string after "input=" when read() returned zero. If read() returned -1
4459 (an event I could not check), uninitialized data bytes were printed.
4460 There were reports of junk text (parts of files, etc) appearing after
4463 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4465 PH/22 Added support for macro redefinition, and (re)definition in between
4466 driver and ACL definitions.
4468 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4469 forgetting to use the resulting value; it was using the unexpanded value.
4471 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4472 hadn't been configured. The fix is from Juergen Kreileder, who
4473 understands it better than I do:
4475 "Here's what I see happening with three configured cyrus_sasl
4476 authenticators configured (plain, login, cram-md5):
4478 On startup auth_cyrus_sasl_init() gets called for each of these.
4479 This means three calls to sasl_listmech() without a specified mech_list.
4480 => SASL tests which mechs of all available mechs actually work
4481 => three warnings about OTP not working
4482 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4484 With the patch, sasl_listmech() also gets called three times. But now
4485 SASL's mech_list option is set to the server_mech specified in the the
4486 authenticator. Or in other words, the answer from sasl_listmech()
4487 gets limited to just the mech you're testing for (which is different
4489 => the return list contains just 'plain' or 'login', 'cram-md5' or
4490 nothing depending on the value of ob->server_mech.
4492 I've just tested the patch: Authentication still works fine,
4493 unavailable mechs specified in the exim configuration are still
4494 caught, and the auth.log warnings about OTP are gone."
4496 PH/25 When debugging is enabled, the contents of the command line are added
4497 to the debugging output, even when log_selector=+arguments is not
4500 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4501 answer is "GNU", and only if the return is "GNU/something" is the answer
4504 PH/27 $acl_verify_message is now set immediately after the failure of a
4505 verification in an ACL, and so is available in subsequent modifiers. In
4506 particular, the message can be preserved by coding like this:
4508 warn !verify = sender
4509 set acl_m0 = $acl_verify_message
4511 Previously, $acl_verify_message was set only while expanding "message"
4512 and "log_message" when a very denied access.
4514 PH/28 Modified OS/os.c-Linux with
4516 -#ifndef OS_LOAD_AVERAGE
4517 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4519 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4520 nomenclature these days.)
4522 PH/29 Installed patch from the Sieve maintainer that adds the options
4523 sieve_useraddress and sieve_subaddress to the redirect router.
4525 PH/30 In these circumstances:
4526 . Two addresses routed to the same list of hosts;
4527 . First host does not offer TLS;
4528 . First host accepts first address;
4529 . First host gives temporary error to second address;
4530 . Second host offers TLS and a TLS session is established;
4531 . Second host accepts second address.
4532 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4533 and peerdn, if requested) that were in fact used only for the second
4536 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4537 attention to any local part prefix or suffix that was matched by the
4538 router that accepted the address. It now behaves in the same way as it
4539 does for delivery: the affixes are removed from the local part unless
4540 rcpt_include_affixes is set on the transport.
4542 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4543 timeout during the DATA phase of an incoming message.
4545 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4546 applied a patch sanctioned by the Sieve maintainer.
4548 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4549 the uid or gid is negative. A case of a negative gid caused this to be
4550 noticed. The fix allows for either to be negative.
4552 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4553 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4556 PH/36 The change PH/12 above was broken. Fixed it.
4558 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4559 the grounds that routing the same address twice would always produce the
4560 same answer. This might have been true once, but it is certainly no
4561 longer true now. Routing a child address may depend on the previous
4562 routing that produced that child. Some complicated redirection strategies
4563 went wrong when messages had multiple recipients, and made Exim's
4564 behaviour dependent on the order in which the addresses were given.
4566 I have moved the duplicate checking until after the routing is complete.
4567 Exim scans the addresses that are assigned to local and remote
4568 transports, and removes any duplicates. This means that more work will be
4569 done, as duplicates will always all be routed, but duplicates are
4570 presumably rare, so I don't expect this is of any significance.
4572 For deliveries to pipes, files, and autoreplies, the duplicate checking
4573 still happens during the routing process, since they are not going to be
4576 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4577 It corrects a timeout issue with spamd. This is Ian's comment: "The
4578 background is that sometimes spamd either never reads data from a
4579 connection it has accepted, or it never writes response data. The exiscan
4580 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4581 blindly assumes that writes won't block so it may never time out."
4583 PH/39 Allow G after quota size as well as K and M.
4585 PH/40 The value set for $authenticated_id in an authenticator may not contain
4586 binary zeroes or newlines because the value is written to log lines and
4587 to spool files. There was no check on this. Now the value is run through
4588 the string_printing() function so that such characters are converted to
4589 printable escape sequences.
4591 PH/41 $message_linecount is a new variable that contains the total number of
4592 lines in the message. Compare $body_linecount, which is the count for the
4595 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4596 bounce and delay warning messages, except in certain special cases, which
4599 (a) An SMTP error message from a remote host;
4600 (b) A message specified in a :fail: redirection;
4601 (c) A message specified in a "fail" command in a system filter;
4602 (d) A message specified in a FAIL return from the queryprogram router;
4603 (e) A message specified by the cannot_route_message router option.
4605 In these cases only, Exim does include the error details in bounce and
4606 warning messages. There are also a few cases where bland messages such
4607 as "unrouteable address" or "local delivery error" are given.
4609 PH/43 $value is now also set for the "else" part of a ${run expansion.
4611 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4612 being worked on, but at least Exim now implements the latest version to
4615 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4616 process to complete was treated as a delivery failure, a timeout while
4617 writing the message to the pipe was logged, but erroneously treated as a
4618 successful delivery. Such timeouts include transport filter timeouts. For
4619 consistency with the overall process timeout, these timeouts are now
4620 treated as errors, giving rise to delivery failures by default. However,
4621 there is now a new Boolean option for the pipe transport called
4622 timeout_defer, which, if set TRUE, converts the failures into defers for
4623 both kinds of timeout. A transport filter timeout is now identified in
4626 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4627 systems where "make" and "gmake" are different, calling "gmake" at top
4628 level broke things. I've arranged for the value of $(MAKE) to be passed
4629 from the Makefile to this script so that it can call the same version of
4633 A note about Exim versions 4.44 and 4.50
4634 ----------------------------------------
4636 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4637 changes of various kinds. As a consequence, a big documentation update was
4638 needed. This delayed the release for rather longer than seemed good, especially
4639 in the light of a couple of (minor) security issues. Therefore, the changes
4640 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4641 release. So 4.44 and 4.50 are in effect two different branches that both start
4644 I have left the 4.50 change log unchanged; it contains all the changes since
4645 4.43. The change log for 4.44 is below; many of its items are identical to
4646 those for 4.50. This seems to be the most sensible way to preserve the
4647 historical information.
4653 1. Minor wording change to the doc/README.SIEVE file.
4655 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4656 computation of the current number of files was incorrect.
4658 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4659 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4662 4. Give more explanation in the error message when the command for a transport
4663 filter fails to execute.
4665 5. There are several places where Exim runs a non-Exim command in a
4666 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4667 was being done only for the command run by the queryprogram router. It is
4668 now done for all such subprocesses. The other cases are: ${run, transport
4669 filters, and the commands run by the lmtp and pipe transports.
4671 6. Added CONFIGURE_GROUP build-time option.
4673 7. Some older OS have a limit of 256 on the maximum number of file
4674 descriptors. Exim was using setrlimit() to set 1000 as a large value
4675 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4676 systems. I've change it so that if it can't get 1000, it tries for 256.
4678 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4679 was an oversight, and furthermore, ever since the addition of extra
4680 controls (e.g. 4.43/32), the checks on when to allow different forms of
4681 "control" were broken. There should now be diagnostics for all cases when a
4682 control that does not make sense is encountered.
4684 9. Added the /retain_sender option to "control=submission".
4686 10. $recipients is now available in the predata ACL (oversight).
4688 11. Tidy the search cache before the fork to do a delivery from a message
4689 received from the command line. Otherwise the child will trigger a lookup
4690 failure and thereby defer the delivery if it tries to use (for example) a
4691 cached ldap connection that the parent has called unbind on.
4693 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4694 of $address_data from the recipient verification was clobbered by the
4695 sender verification.
4697 13. The value of address_data from a sender verification is now available in
4698 $sender_address_data in subsequent conditions in the ACL statement.
4700 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4702 15. Added a new option "connect=<time>" to callout options, to set a different
4705 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4706 was its contents. (It was OK if the option was not defined at all.)
4708 17. A "Completed" log line is now written for messages that are removed from
4709 the spool by the -Mrm option.
4711 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4712 information about exactly what failed.
4714 19. Added -dd to debug only the daemon process.
4716 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4717 handles timeouts, both on the server side and network timeouts. Renamed the
4718 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4720 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4721 It is now set to "smtps".
4723 22. $host_address is now set to the target address during the checking of
4724 ignore_target_hosts.
4726 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4727 being passed; this would have caused $sender_host_name to have been used if
4728 matching the list had actually called for a host name (not very likely,
4729 since this list is usually IP addresses). A host name is now passed as
4732 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4733 code that specifies a non-restarting handler (typically sigaction() in
4734 modern systems) in an attempt to fix a rare and obscure crash bug.
4736 25. Narrowed the window for a race in the daemon that could cause it to ignore
4737 SIGCHLD signals. This is not a major problem, because they are used only to
4738 wake it up if nothing else does.
4740 26. A malformed maildirsize file could cause Exim to calculate negative values
4741 for the mailbox size or file count. Odd effects could occur as a result.
4742 The maildirsize information is now recalculated if the size or filecount
4745 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4746 support for a long time. Removed HAVE_SYS_VFS_H.
4748 28. Installed the latest version of exipick from John Jetmore.
4750 29. In an address list, if the pattern was not a regular expression, an empty
4751 subject address (from a bounce message) matched only if the pattern was an
4752 empty string. Non-empty patterns were not even tested. This was the wrong
4753 because it is perfectly reasonable to use an empty address as part of a
4754 database query. An empty address is now tested by patterns that are
4755 lookups. However, all the other forms of pattern expect the subject to
4756 contain a local part and a domain, and therefore, for them, an empty
4757 address still always fails if the pattern is not itself empty.
4759 30. Exim went into a mad DNS loop when attempting to do a callout where the
4760 host was specified on an smtp transport, and looking it up yielded more
4761 than one IP address.
4763 31. Re-factored the code for checking spool and log partition space into a
4764 function that finds that data and another that does the check. The former
4765 is then used to implement four new variables: $spool_space, $log_space,
4766 $spool_inodes, and $log_inodes.
4768 32. The RFC2047 encoding function was originally intended for short strings
4769 such as real names; it was not keeping to the 75-character limit for
4770 encoded words that the RFC imposes. It now respects the limit, and
4771 generates multiple encoded words if necessary. To be on the safe side, I
4772 have increased the buffer size for the ${rfc2047: expansion operator from
4775 33. It is now permitted to omit both strings after an "if" condition; if the
4776 condition is true, the result is "true". As before, when the second string
4777 is omitted, a false condition yields an empty string. This makes it less
4778 cumbersome to write custom ACL and router conditions.
4780 34. Failure to deliver a bounce message always caused it to be frozen, even if
4781 there was an errors_to setting on the router. The errors_to setting is now
4784 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4785 canonical form (fully expanded) before being placed in
4786 $sender_host_address.
4788 36. The table in the code that translates DNS record types into text (T_A to
4789 "A" for instance) was missing entries for NS and CNAME. It is just possible
4790 that this could have caused confusion if both these types were looked up
4791 for the same domain, because the text type is used as part of Exim's
4792 per-process caching. But the chance of anyone hitting this buglet seems
4795 37. The dnsdb lookup has been extended in a number of ways.
4797 (1) There is a new type, "zns", which walks up the domain tree until it
4798 finds some nameserver records. It should be used with care.
4800 (2) There is a new type, "mxh", which is like "mx" except that it returns
4801 just the host names, not the priorities.
4803 (3) It is now possible to give a list of domains (or IP addresses) to be
4804 looked up. The behaviour when one of the lookups defers can be
4805 controlled by a keyword.
4807 (4) It is now possible to specify the separator character for use when
4808 multiple records are returned.
4810 38. The dnslists ACL condition has been extended: it is now possible to supply
4811 a list of IP addresses and/or domains to be looked up in a particular DNS
4814 39. Added log_selector=+queue_time_overall.
4816 40. When running the queue in the test harness, wait just a tad after forking a
4817 delivery process, to get repeatability of debugging output.
4819 41. Include certificate and key file names in error message when GnuTLS fails
4820 to set them up, because the GnuTLS error message doesn't include the name
4821 of the failing file when there is a problem reading it.
4823 42. Allow both -bf and -bF in the same test run.
4825 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4827 44. The "Exiscan patch" is now merged into the mainline Exim source.
4829 45. Sometimes the final signoff response after QUIT could fail to get
4830 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4831 < 0 before doing a fflush(). This bug looks as though it goes back to the
4832 introduction of TLS in release 3.20, but "sometimes" must have been rare
4833 because the tests only now provoked it.
4835 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4836 (this can affect the format of dates).
4838 47. exim_tidydb, when checking for the continued existence of a message for
4839 which it has found a message-specific retry record, was not finding
4840 messages that were in split spool directories. Consequently, it was
4841 deleting retry records that should have stayed in existence.
4843 48. Steve fixed some bugs in eximstats.
4845 49. The SPA authentication driver was not abandoning authentication and moving
4846 on to the next authenticator when an expansion was forced to fail,
4847 contradicting the general specification for all authenticators. Instead it
4848 was generating a temporary error. It now behaves as specified.
4850 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4851 (the order specifies the preference for clients). The order is now AES256,
4852 AES128, 3DES, ARCFOUR128.
4854 51. Small patch to Sieve code - explicitly set From: when generating an
4857 52. Exim crashed if a remote delivery caused a very long error message to be
4858 recorded - for instance if somebody sent an entire SpamAssassin report back
4859 as a large number of 550 error lines. This bug was coincidentally fixed by
4860 increasing the size of one of Exim's internal buffers (big_buffer) that
4861 happened as part of the Exiscan merge. However, to be on the safe side, I
4862 have made the code more robust (and fixed the comments that describe what
4865 53. Now that there can be additional text after "Completed" in log lines (if
4866 the queue_time_overall log selector is set), a one-byte patch to exigrep
4867 was needed to allow it to recognize "Completed" as not the last thing in
4870 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4871 patch that reportedly fixes this has been added. I am not expert enough to
4872 create a test for it. This is what the patch creator wrote:
4874 "I found a little strange behaviour of ldap code when working with
4875 Windows 2003 AD Domain, where users was placed in more than one
4876 Organization Units. When I tried to give exim partial DN, the exit code
4877 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4878 But simultaneously result of request was absolutely normal ldap result,
4879 so I produce this patch..."
4881 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4882 REFERENCE, so I have modified the code to exclude the patch when that macro
4885 55. Some experimental protocols are using DNS PTR records for new purposes. The
4886 keys for these records are domain names, not reversed IP addresses. The
4887 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4888 leaves it alone. Component reversal etc. now happens only for IP addresses.
4891 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4893 57. Double the size of the debug message buffer (to 2048) so that more of very
4894 long debug lines gets shown.
4896 58. The exicyclog utility now does better if the number of log files to keep
4897 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4899 59. Two changes related to the smtp_active_hostname option:
4901 (1) $smtp_active_hostname is now available as a variable.
4902 (2) The default for smtp_banner uses $smtp_active_hostname instead
4903 of $primary_hostname.
4905 60. The host_aton() function is supposed to be passed a string that is known
4906 to be a valid IP address. However, in the case of IPv6 addresses, it was
4907 not checking this. This is a hostage to fortune. Exim now panics and dies
4908 if the condition is not met. A case was found where this could be provoked
4909 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4910 components; fortuitously, this particular loophole had already been fixed
4911 by change 4.50/55 above.
4913 If there are any other similar loopholes, the new check in host_aton()
4914 itself should stop them being exploited. The report I received stated that
4915 data on the command line could provoke the exploit when Exim was running as
4916 exim, but did not say which command line option was involved. All I could
4917 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4918 running as the user.
4921 61. There was a buffer overflow vulnerability in the SPA authentication code
4922 (which came originally from the Samba project). I have added a test to the
4923 spa_base64_to_bits() function which I hope fixes it.
4926 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4927 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4929 63. The daemon start-up calls getloadavg() while still root for those OS that
4930 need the first call to be done as root, but it missed one case: when
4931 deliver_queue_load_max is set with deliver_drop_privilege. This is
4932 necessary for the benefit of the queue runner, because there is no re-exec
4933 when deliver_drop_privilege is set.
4935 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4936 This has been fixed.
4938 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4939 was in use, was not putting the data itself into the right store pool;
4940 consequently, it could be overwritten for a subsequent message in the same
4941 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4944 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4946 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4947 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4948 the function were treating the return as a boolean value, which happened to
4949 work because 0=false and not-0=true, but is not correct code.
4951 68. The host_aton() function was not handling scoped IPv6 addresses (those
4952 with, for example, "%eth0" on the end) correctly.
4954 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4955 negated items (that is, ~something) in unsigned ints. Some compilers
4956 apparently mutter when there is no cast.
4958 70. If an address verification called from an ACL failed, and did not produce a
4959 user-specific message (i.e. there was only a "system" message), nothing was
4960 put in $acl_verify_message. In this situation, it now puts the system
4963 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4964 session; change 4.31/43 added the unwanted input to the log line - except
4965 that it did not do this in the start of session case. It now does.
4967 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4968 This is wrong and can cause the other end to generate a synchronization
4969 error if it is another Exim or anything else that does the synchronization
4970 check. A QUIT command is no longer sent after a timeout.
4972 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4973 during host lookups.
4975 74. The defer_ok option of callout verification was not working if it was used
4976 when verifying addresses in header lines, that is, for this case:
4978 verify = header_sender/callout=defer_ok
4980 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4981 those file descriptors could be used for SMTP connections. If anything
4982 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4983 could be sent to the SMTP client, causing chaos. The daemon now opens
4984 stdin, stdout, and stderr to /dev/null when it puts itself into the
4987 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4988 log by default. The user can override this with suitable Perl magic.
4990 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4991 the log message when discard triggers, was not working for the DATA ACL or
4992 for the non-SMTP ACL.
4994 78. Error message wording change in sieve.c.
4996 79. If smtp_accept_max_per_host was set, the number of connections could be
4997 restricted to fewer than expected, because the daemon was trying to set up
4998 a new connection before checking whether the processes handling previous
4999 connections had finished. The check for completed processes is now done
5000 earlier. On busy systems, this bug wouldn't be noticed because something
5001 else would have woken the daemon, and it would have reaped the completed
5004 80. If a message was submitted locally by a user whose login name contained one
5005 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5006 It caused a spool format error. I have fixed the spool reading code. A
5007 related problem was that the "from" clause in the Received: line became
5008 illegal because of the space(s). It is now covered by ${quote_local_part.
5010 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5013 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5015 83. Patch from Sieve maintainer to fix unterminated string problem in
5016 "vacation" handling.
5018 84. Some minor changes to the Linux configuration files to help with other
5019 OS variants using glibc.
5021 85. One more patch for Sieve to update vacation handling to latest spec.
5024 ----------------------------------------------------
5025 See the note above about the 4.44 and 4.50 releases.
5026 ----------------------------------------------------
5032 1. Change 4.43/35 introduced a bug that caused file counts to be
5033 incorrectly computed when quota_filecount was set in an appendfile
5036 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5037 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5040 3. Give more explanation in the error message when the command for a transport
5041 filter fails to execute.
5043 4. There are several places where Exim runs a non-Exim command in a
5044 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5045 was being done only for the command run by the queryprogram router. It is
5046 now done for all such subprocesses. The other cases are: ${run, transport
5047 filters, and the commands run by the lmtp and pipe transports.
5049 5. Some older OS have a limit of 256 on the maximum number of file
5050 descriptors. Exim was using setrlimit() to set 1000 as a large value
5051 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5052 systems. I've change it so that if it can't get 1000, it tries for 256.
5054 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5055 was an oversight, and furthermore, ever since the addition of extra
5056 controls (e.g. 4.43/32), the checks on when to allow different forms of
5057 "control" were broken. There should now be diagnostics for all cases when a
5058 control that does not make sense is encountered.
5060 7. $recipients is now available in the predata ACL (oversight).
5062 8. Tidy the search cache before the fork to do a delivery from a message
5063 received from the command line. Otherwise the child will trigger a lookup
5064 failure and thereby defer the delivery if it tries to use (for example) a
5065 cached ldap connection that the parent has called unbind on.
5067 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5068 of $address_data from the recipient verification was clobbered by the
5069 sender verification.
5071 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5072 was its contents. (It was OK if the option was not defined at all.)
5074 11. A "Completed" log line is now written for messages that are removed from
5075 the spool by the -Mrm option.
5077 12. $host_address is now set to the target address during the checking of
5078 ignore_target_hosts.
5080 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5081 being passed; this would have caused $sender_host_name to have been used if
5082 matching the list had actually called for a host name (not very likely,
5083 since this list is usually IP addresses). A host name is now passed as
5086 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5087 code that specifies a non-restarting handler (typically sigaction() in
5088 modern systems) in an attempt to fix a rare and obscure crash bug.
5090 15. Narrowed the window for a race in the daemon that could cause it to ignore
5091 SIGCHLD signals. This is not a major problem, because they are used only to
5092 wake it up if nothing else does.
5094 16. A malformed maildirsize file could cause Exim to calculate negative values
5095 for the mailbox size or file count. Odd effects could occur as a result.
5096 The maildirsize information is now recalculated if the size or filecount
5099 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5100 support for a long time. Removed HAVE_SYS_VFS_H.
5102 18. Updated exipick to current release from John Jetmore.
5104 19. Allow an empty sender to be matched against a lookup in an address list.
5105 Previously the only cases considered were a regular expression, or an
5108 20. Exim went into a mad DNS lookup loop when doing a callout where the
5109 host was specified on the transport, if the DNS lookup yielded more than
5112 21. The RFC2047 encoding function was originally intended for short strings
5113 such as real names; it was not keeping to the 75-character limit for
5114 encoded words that the RFC imposes. It now respects the limit, and
5115 generates multiple encoded words if necessary. To be on the safe side, I
5116 have increased the buffer size for the ${rfc2047: expansion operator from
5119 22. Failure to deliver a bounce message always caused it to be frozen, even if
5120 there was an errors_to setting on the router. The errors_to setting is now
5123 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5124 canonical form (fully expanded) before being placed in
5125 $sender_host_address.
5127 24. Updated eximstats to version 1.33
5129 25. Include certificate and key file names in error message when GnuTLS fails
5130 to set them up, because the GnuTLS error message doesn't include the name
5131 of the failing file when there is a problem reading it.
5133 26. Expand error message when OpenSSL has problems setting up cert/key files.
5136 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5137 (this can affect the format of dates).
5139 28. exim_tidydb, when checking for the continued existence of a message for
5140 which it has found a message-specific retry record, was not finding
5141 messages that were in split spool directories. Consequently, it was
5142 deleting retry records that should have stayed in existence.
5144 29. eximstats updated to version 1.35
5145 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5146 1.35 - bugfix such that pie charts by volume are generated correctly
5148 30. The SPA authentication driver was not abandoning authentication and moving
5149 on to the next authenticator when an expansion was forced to fail,
5150 contradicting the general specification for all authenticators. Instead it
5151 was generating a temporary error. It now behaves as specified.
5153 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5154 (the order specifies the preference for clients). The order is now AES256,
5155 AES128, 3DES, ARCFOUR128.
5157 31. Small patch to Sieve code - explicitly set From: when generating an
5160 32. Exim crashed if a remote delivery caused a very long error message to be
5161 recorded - for instance if somebody sent an entire SpamAssassin report back
5162 as a large number of 550 error lines. This bug was coincidentally fixed by
5163 increasing the size of one of Exim's internal buffers (big_buffer) that
5164 happened as part of the Exiscan merge. However, to be on the safe side, I
5165 have made the code more robust (and fixed the comments that describe what
5168 33. Some experimental protocols are using DNS PTR records for new purposes. The
5169 keys for these records are domain names, not reversed IP addresses. The
5170 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5171 leaves it alone. Component reversal etc. now happens only for IP addresses.
5174 34. The host_aton() function is supposed to be passed a string that is known
5175 to be a valid IP address. However, in the case of IPv6 addresses, it was
5176 not checking this. This is a hostage to fortune. Exim now panics and dies
5177 if the condition is not met. A case was found where this could be provoked
5178 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5179 components; fortuitously, this particular loophole had already been fixed
5180 by change 4.50/55 or 4.44/33 above.
5182 If there are any other similar loopholes, the new check in host_aton()
5183 itself should stop them being exploited. The report I received stated that
5184 data on the command line could provoke the exploit when Exim was running as
5185 exim, but did not say which command line option was involved. All I could
5186 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5187 running as the user.
5190 35. There was a buffer overflow vulnerability in the SPA authentication code
5191 (which came originally from the Samba project). I have added a test to the
5192 spa_base64_to_bits() function which I hope fixes it.
5195 36. The daemon start-up calls getloadavg() while still root for those OS that
5196 need the first call to be done as root, but it missed one case: when
5197 deliver_queue_load_max is set with deliver_drop_privilege. This is
5198 necessary for the benefit of the queue runner, because there is no re-exec
5199 when deliver_drop_privilege is set.
5201 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5202 was in use, was not putting the data itself into the right store pool;
5203 consequently, it could be overwritten for a subsequent message in the same
5204 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5207 38. Sometimes the final signoff response after QUIT could fail to get
5208 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5209 < 0 before doing a fflush(). This bug looks as though it goes back to the
5210 introduction of TLS in release 3.20, but "sometimes" must have been rare
5211 because the tests only now provoked it.
5217 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5218 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5219 Now it returns an integer. A number of calls were still expecting a T/F
5220 return. Fortuitously, in all cases, the tests worked in OK situations,
5221 which is the norm. However, things would have gone wrong on any write
5222 failures on the smtp file descriptor. This function is used when sending
5223 messages over SMTP and also when doing verify callouts.
5225 2. When Exim is called to do synchronous delivery of a locally submitted
5226 message (the -odf or -odi options), it no longer closes stderr before doing
5229 3. Implemented the mua_wrapper option.
5231 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5233 5. Implemented the functions header_remove(), header_testname(),
5234 header_add_at_position(), and receive_remove_recipient(), and exported them
5237 6. If an ACL "warn" statement specified the addition of headers, Exim already
5238 inserted X-ACL-Warn: at the start if there was no header name. However, it
5239 was not making this test for the second and subsequent header lines if
5240 there were newlines in the string. This meant that an invalid header could
5241 be inserted if Exim was badly configured.
5243 7. Allow an ACL "warn" statement to add header lines at the start or after all
5244 the Received: headers, as well as at the end.
5246 8. Added the rcpt_4xx retry error code.
5248 9. Added postmaster_mailfrom=xxx to callout verification option.
5250 10. Added mailfrom=xxxx to the callout verification option, for verify=
5253 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5254 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5256 12. Inserted some casts to stop certain compilers warning when using pointer
5257 differences as field lengths or precisions in printf-type calls (mostly
5258 affecting debugging statements).
5260 13. Added optional readline() support for -be (dynamically loaded).
5262 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5263 same clock tick as a message's arrival, so that its received time was the
5264 same as the "first fail" time on the retry record, and that message
5265 remained on the queue past the ultimate address timeout, every queue runner
5266 would try a delivery (because it was past the ultimate address timeout) but
5267 after another failure, the ultimate address timeout, which should have then
5268 bounced the address, did not kick in. This was a "< instead of <=" error;
5269 in most cases the first failure would have been in the next clock tick
5270 after the received time, and all would be well.
5272 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5273 being recognized when the domain list was tested by the match_domain
5274 condition in an expansion string.
5276 16. Added the ${str2b64: operator.
5278 17. Exim was always calling setrlimit() to set a large limit for the number of
5279 processes, without checking whether the existing limit was already
5280 adequate. (It did check for the limit on file descriptors.) Furthermore,
5281 errors from getrlimit() and setrlimit() were being ignored. Now they are
5282 logged to the main and panic logs, but Exim does carry on, to try to do its
5283 job under whatever limits there are.
5285 18. Imported PCRE 5.0.
5287 19. Trivial typo in log message " temporarily refused connection" (the leading
5290 20. If the log selector return_path_on_delivery was set and an address was
5291 redirected to /dev/null, the delivery process crashed because it assumed
5292 that a return path would always be set for a "successful" delivery. In this
5293 case, the whole delivery is bypassed as an optimization, and therefore no
5296 21. Internal re-arrangement: the function for sending a challenge and reading
5297 a response while authentication was assuming a zero-terminated challenge
5298 string. It's now changed to take a pointer and a length, to allow for
5299 binary data in such strings.
5301 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5303 23. Exim was not respecting finduser_retries when seeking the login of the
5304 uid under which it was called; it was always trying 10 times. (The default
5305 setting of finduser_retries is zero.) Also, it was sleeping after the final
5306 failure, which is pointless.
5308 24. Implemented tls_on_connect_ports.
5310 25. Implemented acl_smtp_predata.
5312 26. If the domain in control=submission is set empty, Exim assumes that the
5313 authenticated id is a complete email address when it generates From: or
5314 Sender: header lines.
5316 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5317 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5318 chown and chgrp in /bin and hostname in /usr/bin.
5320 28. Exim was keeping the "process log" file open after each use, just as it
5321 does for the main log. This opens the possibility of it remaining open for
5322 long periods when the USR1 signal hits a daemon. Occasional processlog
5323 errors were reported, that could have been caused by this. Anyway, it seems
5324 much more sensible not to leave this file open at all, so that is what now
5327 29. The long-running daemon process does not normally write to the log once it
5328 has entered its main loop, and it closes the log before doing so. This is
5329 so that log files can straightforwardly be renamed and moved. However,
5330 there are a couple of unusual error situations where the daemon does write
5331 log entries, and I had neglected to close the log afterwards.
5333 30. The text of an SMTP error response that was received during a remote
5334 delivery was being truncated at 512 bytes. This is too short for some of
5335 the long messages that one sometimes sees. I've increased the limit to
5338 31. It is now possible to make retry rules that apply only when a message has a
5339 specific sender, in particular, an empty sender.
5341 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5342 it possible to be selective about when SMTP synchronization is enforced.
5344 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5346 32. Implemented hosts_connection_nolog.
5348 33. Added an ACL for QUIT.
5350 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5353 35. Added mailbox_size and mailbox_filecount to appendfile.
5355 36. Added control = no_multiline_responses to ACLs.
5357 37. There was a bug in the logic of the code that waits for the clock to tick
5358 in the case where the clock went backwards by a substantial amount such
5359 that the microsecond fraction of "now" was more than the microsecond
5360 fraction of "then" (but the whole seconds number was less).
5362 38. Added support for the libradius Radius client library this is found on
5363 FreeBSD (previously only the radiusclient library was supported).
5369 1. When certain lookups returned multiple values in the form name=value, the
5370 quoting of the values was not always being done properly. Specifically:
5371 (a) If the value started with a double quote, but contained no whitespace,
5373 (b) If the value contained whitespace other than a space character (i.e.
5374 tabs or newlines or carriage returns) it was not quoted.
5375 This fix has been applied to the mysql and pgsql lookups by writing a
5376 separate quoting function and calling it from the lookup code. The fix
5377 should probably also be applied to nisplus, ibase and oracle lookups, but
5378 since I cannot test any of those, I have not disturbed their existing code.
5380 2. A hit in the callout cache for a specific address caused a log line with no
5381 reason for rejecting RCPT. Now it says "Previous (cached) callout
5382 verification failure".
5384 3. There was an off-by-one bug in the queryprogram router. An over-long
5385 return line was truncated at 256 instead of 255 characters, thereby
5386 overflowing its buffer with the terminating zero. As well as fixing this, I
5387 have increased the buffer size to 1024 (and made a note to document this).
5389 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5390 when Exim is waiting for an SMTP response from a remote server, Exim
5391 restarts its select() call on the socket, thereby resetting its timeout.
5392 This is not a problem when such interrupts are rare. Somebody set up a cron
5393 job to run exiwhat every 2 minutes, which is less than the normal select()
5394 timeout (5 or 10 minutes). This meant that the select() timeout never
5395 kicked in because it was always reset. I have fixed this by comparing the
5396 time when an interrupt arrives with the time at the start of the first call
5397 to select(). If more time than the timeout has elapsed, the interrupt is
5398 treated as a timeout.
5400 5. Some internal re-factoring in preparation for the addition of Sieve
5401 extensions (by MH). In particular, the "personal" test is moved to a
5402 separate function, and given an option for scanning Cc: and Bcc: (which is
5403 not set for Exim filters).
5405 6. When Exim created an email address using the login of the caller as the
5406 local part (e.g. when creating a From: or Sender: header line), it was not
5407 quoting the local part when it contained special characters such as @.
5409 7. Installed new OpenBSD configuration files.
5411 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5412 try to make them clearer.
5414 9. Callout options, other than the timeout value, were being ignored when
5415 verifying sender addresses in header lines. For example, when using
5417 verify = header_sender/callout=no_cache
5419 the cache was (incorrectly) being used.
5421 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5422 only the Cygwin environment.
5424 11. When return_path_on_delivery was set as a log selector, if different remote
5425 addresses in the same message used different return paths and parallel
5426 remote delivery occurred, the wrong values would sometimes be logged.
5427 (Whenever a remote delivery process finished, the return path value from
5428 the most recently started remote delivery process was logged.)
5430 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5431 lines when AUTH and/or TLS are in use. This is the "received protocol"
5432 field. Exim used to use "asmtp" for authenticated SMTP, without any
5433 indication (in the protocol name) for TLS use. Now it follows the RFC and
5434 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5435 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5436 names appear in log lines as well as in Received: header lines.
5438 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5439 extensions, and comparison tests, and to fix some bugs.
5441 14. Changes to the "personal" filter test:
5443 (1) The test was buggy in that it was just doing the equivalent of
5444 "contains" tests on header lines. For example, if a user's address was
5445 anne@some.where, the "personal" test would incorrectly be true for
5447 To: susanne@some.where
5449 This test is now done by extracting each address from the header in turn,
5450 and checking the entire address. Other tests that are part of "personal"
5451 are now done using regular expressions (for example, to check local parts
5452 of addresses in From: header lines).
5454 (2) The list of non-personal local parts in From: addresses has been
5455 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5456 taken from the Sieve specification recommendations.
5458 (3) If the message contains any header line starting with "List-" it is
5459 treated as non-personal.
5461 (4) The test for "circular" in the Subject: header line has been removed
5462 because it now seems ill-conceived.
5464 15. Minor typos in src/EDITME comments corrected.
5466 16. Installed latest exipick from John Jetmore.
5468 17. If headers_add on a router specified a text string that was too long for
5469 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5470 of string_sprintf() is now avoided.
5472 18. $message_body_size was not set (it was always zero) when running the DATA
5473 ACL and the local_scan() function.
5475 19. For the "mail" command in an Exim filter, no default was being set for
5476 the once_repeat time, causing a random time value to be used if "once" was
5477 specified. (If the value happened to be <= 0, no repeat happened.) The
5478 default is now 0s, meaning "never repeat". The "vacation" command was OK
5479 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5480 (I found it when inspecting the code).
5482 20. There is now an overall timeout for performing a callout verification. It
5483 defaults to 4 times the callout timeout, which applies to individual SMTP
5484 commands during the callout. The overall timeout applies when there is more
5485 than one host that can be tried. The timeout is checked before trying the
5486 next host. This prevents very long delays if there are a large number of
5487 hosts and all are timing out (e.g. when the network connections are timing
5488 out). The value of the overall timeout can be changed by specifying an
5489 additional sub-option for "callout", called "maxwait". For example:
5491 verify = sender/callout=5s,maxwait=20s
5493 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5494 to the end before writing, but this should make it even safer).
5496 22. Exim was forgetting that it had advertised PIPELINING for the second and
5497 subsequent messages on an SMTP connection. It was also not resetting its
5498 memory on STARTTLS and an internal HELO.
5500 23. When Exim logs an SMTP synchronization error within a session, it now
5501 records whether PIPELINING has been advertised or not.
5503 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5504 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5505 rather than long int.
5507 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5509 26. Added the never_mail option to autoreply.
5515 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5516 crash if the getsockname() call failed; this can happen if a connection is
5517 closed very soon after it is established. The problem was simply in the
5518 order in which certain operations were done, causing Exim to try to write
5519 to the SMTP stream before it had set up the file descriptor. The bug has
5520 been fixed by making things happen in the correct order.
5526 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5527 before the connection was closed, thus losing the rejection response.
5529 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5530 some early Solaris releases, but causes trouble in current releases where
5531 socklen_t is defined.
5533 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5536 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5539 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5540 that releasing the top of it at the end releases what was used for sub-
5541 expansions (unless the block got too big). However, discard this block if
5542 the first thing is a variable or header, so that we can use its block when
5543 it is dynamic (useful for very large $message_headers, for example).
5545 6. Lookups now cache *every* query, not just the most recent. A new, separate
5546 store pool is used for this. It can be recovered when all lookup caches are
5547 flushed. Lookups now release memory at the end of their result strings.
5548 This has involved some general refactoring of the lookup sources.
5550 7. Some code has been added to the store_xxx() functions to reduce the amount
5551 of flapping under certain conditions.
5553 8. log_incoming_interface used to affect only the <= reception log lines. Now
5554 it causes the local interface and port to be added to several more SMTP log
5555 lines, for example "SMTP connection from", and rejection lines.
5557 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5559 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5561 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5562 could be overwritten at the end of the current message (or the start of a
5563 new message if it was set in a HELO ACL). The value is now preserved for
5564 the duration of the SMTP connection.
5566 12. If a transport had a headers_rewrite setting, and a matching header line
5567 contained an unqualified address, that address was qualified, even if it
5568 did not match any rewriting rules. The underlying bug was that the values
5569 of the flags that permit the existence of unqualified sender and recipient
5570 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5571 non-local messages, and by -bnq for local messages) were not being
5572 preserved with the message after it was received.
5574 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5575 "next input=" as part of the text comprising the host identity instead of
5576 the correct text. The code was using the same buffer for two different
5577 strings. However, depending on which order the printing function evaluated
5578 its arguments, the bug did not always show up. Under Linux, for example, my
5579 test suite worked just fine.
5581 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5582 doesn't work with some older versions of Perl. It has been changed to "my",
5583 which in any case is probably the better facility to use.
5585 15. A really picky compiler found some instances of statements for creating
5586 error messages that either had too many or two few arguments for the format
5589 16. The size of the buffer for calls to the DNS resolver has been increased
5590 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5591 for addresses that have a lot of PTR records. This alleviates a problem; it
5592 does not fully solve it.
5594 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5595 buffer now truncates the list and logs the incident, which is the same
5596 action as happens when Exim is looking up a host name and its aliases.
5597 Previously in this situation something unpredictable would happen;
5598 sometimes it was "internal error: store_reset failed".
5600 18. If a server dropped the connection unexpectedly when an Exim client was
5601 using GnuTLS and trying to read a response, the client delivery process
5602 crashed while trying to generate an error log message.
5604 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5605 string, for example:
5607 warn message = H1: something\nH2: something
5609 the text was added as a single header line from Exim's point of view
5610 though it ended up OK in the delivered message. However, searching for the
5611 second and subsequent header lines using $h_h2: did not work. This has been
5612 fixed. Similarly, if a system filter added multiple headers in this way,
5613 the routers could not see them.
5615 20. Expanded the error message when iplsearch is called with an invalid key to
5616 suggest using net-iplsearch in a host list.
5618 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5619 ACLs are no longer actually imposed (and a message to that effect is
5622 22. If a "gecos" field in a passwd entry contained escaped characters, in
5623 particular, if it contained a \" sequence, Exim got it wrong when building
5624 a From: or a Sender: header from that name. A second bug also caused
5625 incorrect handling when an unquoted " was present following a character
5626 that needed quoting.
5628 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5629 was not being matched caselessly.
5631 24. Arranged for all hyphens in the exim.8 source to be escaped with
5634 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5635 properly. Recipient callout cache records were still being keyed to include
5636 the sender, even when use_sender was set false. This led to far more
5637 callouts that were necessary. The sender is no longer included in the key
5638 when use_sender is false.
5640 26. Added "control = submission" modifier to ACLs.
5642 27. Added the ${base62d: operator to decode base 62 numbers.
5644 28. dnsdb lookups can now access SRV records.
5646 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5647 the configuration file.
5649 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5650 (-v) mode. This makes the output for a verbose queue run more intelligible.
5652 31. Added a use_postmaster feature to recipient callouts.
5654 32. Added the $body_zerocount variable, containing the number of binary zero
5655 bytes in the message body.
5657 33. The time of last modification of the "new" subdirectory is now used as the
5658 "mailbox time last read" when there is a quota error for a maildir
5661 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5663 35. Added +ignore_unknown as a special item in host lists.
5665 36. Code for decoding IPv6 addresses in host lists is now included, even if
5666 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5667 address was recognized as an IP address, but was then not correctly decoded
5668 into binary, causing unexpected and incorrect effects when compared with
5675 1. Very minor rewording of debugging text in manualroute to say "list of
5676 hosts" instead of "hostlist".
5678 2. If verify=header_syntax was set, and a header line with an unqualified
5679 address (no domain) and a large number of spaces between the end of the
5680 name and the colon was received, the reception process suffered a buffer
5681 overflow, and (when I tested it) crashed. This was caused by some obsolete
5682 code that should have been removed. The fix is to remove it!
5684 3. When running in the test harness, delay a bit after writing a bounce
5685 message to get a bit more predictability in the log output.
5687 4. Added a call to search_tidyup() just before forking a reception process. In
5688 theory, someone could use a lookup in the expansion of smtp_accept_max_
5689 per_host which, without the tidyup, could leave open a database connection.
5691 5. Added the variables $recipient_data and $sender_data which get set from a
5692 lookup success in an ACL "recipients" or "senders" condition, or a router
5693 "senders" option, similar to $domain_data and $local_part_data.
5695 6. Moved the writing of debug_print from before to after the "senders" test
5698 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5699 problems for message scanning, either using a data ACL, or using
5700 local_scan() because the Received: header was not generated till after they
5701 were called (in order to set the time as the time of reception completion).
5702 I have revised the way this works. The header is now generated after the
5703 body is received, but before the ACL or local_scan() are called. After they
5704 are run, the timestamp in the header is updated.
5710 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5711 before starting a queue runner without re-exec. This happened only when
5712 deliver_drop_privilege was set or when the Exim user was set to root. The
5713 effect of the bug was that timeouts during subsequent deliveries caused
5714 crashes instead of being properly handled. The handler is now left at its
5715 default (and expected) setting.
5717 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5718 message, again when deliver_drop_privilege is set or Exim is run as root.
5719 The bug described in (1) was not present in this case, but the tidying up
5720 of the other signals was missing. I have made the two cases consistent.
5722 3. The ignore_target_hosts setting on a manualroute router was being ignored
5723 for hosts that were looked up using the /MX notation.
5725 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5728 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5729 operated on the sender address. After changing the $sender_address to <>
5730 for the sender address verify, Exim was re-instated it as the original
5731 (before rewriting) address, but remembering that it had rewritten it, so it
5732 wasn't rewriting it again. This bug also had the effect of breaking the
5733 sender address verification caching when the sender address was rewritten.
5735 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5736 This has been changed so that if the ip literal address matches
5737 ignore_target_hosts, the router declines.
5739 7. Added expansion conditions match_domain, match_address, and match_local_
5740 part (NOT match_host).
5742 8. The placeholder for the Received: header didn't have a length field set.
5744 9. Added code to Exim itself and to exim_lock to test for a specific race
5745 condition that could lead to file corruption when using MBX delivery. The
5746 issue is with the lockfile that is created in /tmp. If this file is removed
5747 after a process has opened it but before that process has acquired a lock,
5748 there is the potential for a second process to recreate the file and also
5749 acquire a lock. This could lead to two Exim processes writing to the file
5750 at the same time. The added code performs the same test as UW imapd; it
5751 checks after acquiring the lock that its file descriptor still refers to
5752 the same named file.
5754 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5755 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5758 11. Added the smtp_active_hostname option. If used, this will typically be made
5759 to depend on the incoming interface address. Because $interface_address is
5760 not set up until the daemon has forked a reception process, error responses
5761 that can happen earlier (such as "too many connections") no longer contain
5764 12. If an expansion in a condition on a "warn" statement fails because a lookup
5765 defers, the "warn" statement is abandoned, and the next ACL statement is
5766 processed. Previously this caused the whole ACL to be aborted.
5768 13. Added the iplsearch lookup type.
5770 14. Added ident_timeout as a log selector.
5772 15. Added tls_certificate_verified as a log selector.
5774 16. Added a global option tls_require_ciphers (compare the smtp transport
5775 option of the same name). This controls incoming TLS connections.
5777 17. I finally figured out how to make tls_require_ciphers do a similar thing
5778 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5779 before starting the TLS session.
5781 18. Tabs are now shown as \t in -bP output.
5783 19. If the log selector return_path_on_delivery was set, Exim crashed when
5784 bouncing a message because it had too many Received: header lines.
5786 20. If two routers both had headers_remove settings, and the first one included
5787 a superfluous trailing colon, the final name in the first list and the
5788 first name in the second list were incorrectly joined into one item (with a
5789 colon in the middle).
5795 1. Added -C and -D options to the exinext utility, mainly to make it easier
5796 to include in the automated testing, but these could be helpful when
5797 multiple configurations are in use.
5799 2. The exinext utility was not formatting the output nicely when there was
5800 an alternate port involved in the retry record key, nor when there was a
5801 message id as well (for retries that were specific to a specific message
5802 and a specific host). It was also confused by IPv6 addresses, because of
5803 the additional colons they contain. I have fixed the IPv4 problem, and
5804 patched it up to do a reasonable job for IPv6.
5806 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5807 delivery, the log line now contains "pipelined" if PIPELINING was used.
5809 4. An SMTP transport process used to panic and die if the bind() call to set
5810 an explicit outgoing interface failed. This has been changed; it is now
5811 treated in the same way as a connect() failure.
5813 5. A reference to $sender_host_name in the part of a conditional expansion
5814 that was being skipped was still causing a DNS lookup. This no longer
5817 6. The def: expansion condition was not recognizing references to header lines
5818 that used bh_ and bheader_.
5820 7. Added the _cache feature to named lists.
5822 8. The code for checking quota_filecount in the appendfile transport was
5823 allowing one more file than it should have been.
5825 9. For compatibility with Sendmail, the command line option
5833 and sets the incoming protocol and host name (for trusted callers). The
5834 host name and its colon can be omitted when only the protocol is to be set.
5835 Note the Exim already has two private options, -pd and -ps, that refer to
5836 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5837 "s", but I don't think that's a major issue.
5839 10. A number of refactoring changes to the code, none of which should affect
5842 (a) The number of logging options was getting close to filling up the
5843 32-bit word that was used as a bit map. I have split them into two classes:
5844 those that are passed in the argument to log_write(), and those that are
5845 only ever tested independently outside of that function. These are now in
5846 separate 32-bit words, so there is plenty of room for expansion again.
5847 There is no change in the user interface or the logging behaviour.
5849 (b) When building, for example, log lines, the code previously used a
5850 macro that called string_cat() twice, in order to add two strings. This is
5851 not really sufficiently general. Furthermore, there was one instance where
5852 it was actually wrong because one of the argument was used twice, and in
5853 one call a function was used. (As it happened, calling the function twice
5854 did not affect the overall behaviour.) The macro has been replaced by a
5855 function that can join an arbitrary number of extra strings onto a growing
5858 (c) The code for expansion conditions now uses a table and a binary chop
5859 instead of a serial search (which was left over from when there were very
5860 few conditions). Also, it now recognizes conditions like "pam" even when
5861 the relevant support is not compiled in: a suitably worded error message is
5862 given if an attempt is made to use such a condition.
5864 11. Added ${time_interval:xxxxx}.
5866 12. A bug was causing one of the ddress fields not to be passed back correctly
5867 from remote delivery subprocesses. The field in question was not being
5868 subsequently used, so this caused to problems in practice.
5870 13. Added new log selectors queue_time and deliver_time.
5872 14. Might have fixed a bug in maildirsizefile handling that threw up
5873 "unexpected character" debug warnings, and recalculated the data
5874 unnecessarily. In any case, I expanded the warning message to give more
5877 15. Added the message "Restricted characters in address" to the statements in
5878 the default ACL that block characters like @ and % in local parts.
5880 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5881 Three changes have been made:
5883 (a) There was a serious bug; a negative response to MAIL caused the whole
5884 recipient domain to be cached as invalid, thereby blocking all messages
5885 to all local parts at the same domain, from all senders. This bug has
5886 been fixed. The domain is no longer cached after a negative response to
5887 MAIL if the sender used is not empty.
5889 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5892 (c) A new callout option, "use_sender" has been added for people who want
5893 the modified behaviour.
5899 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5902 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5903 indeed breaks things for older releases.
5905 3. Added additional logging to the case where there is a problem reading data
5906 from a filter that is running in a subprocess using a pipe, in order to
5907 try to track down a specific problem.
5909 4. Testing facility fudge: when running in the test harness and attempting
5910 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5911 getting "No route to host". Convert this to a timeout.
5913 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5916 6. Some OS don't have socklen_t but use size_t instead. This affects the
5917 fifth argument of getsockopt() amongst other things. This is now
5918 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5919 can be set for individual OS. I have set it for SunOS5, OSF1, and
5920 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5921 some earlier ones do not.
5923 7. Change 4.30/15 was not doing the test caselessly.
5925 8. The standard form for an IPv6 address literal was being rejected by address
5926 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5927 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5928 this, as well as the form without the "IPv6" on the front (but only when
5929 address literals are enabled, of course).
5931 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5933 10. Exim crashed if a message with an empty sender address specified by -f
5934 encountered a router with an errors_to setting. This could be provoked only
5935 by a command such as
5939 where an empty string was supplied; "<>" did not hit this bug.
5941 11. Installed PCRE release 4.5.
5943 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5944 remained set. It is now erased.
5946 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5947 times from message ids (which are base 36 rather than the normal 62).
5949 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5950 were being counted as actual protocol errors, and logged if the log
5951 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5952 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5953 and DATA following a set of rejected RCPTs do not count as protocol errors.
5954 In other words, Exim assumes they were pipelined, though this may not
5955 actually be the case. Of course, in all cases the client gets an
5956 appropriate error code.
5958 15. If a lookup fails in an ACL condition, a message about the failure may
5959 be available; it is used if testing the ACL cannot continue, because most
5960 such messages specify what the cause of the deferral is. However, some
5961 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5962 that caused an old message to be retained and used if a later statement
5963 caused a defer, replacing the real cause of the deferral.
5965 16. If an IP address had so many PTR records that the DNS lookup buffer
5966 was not large enough to hold them, Exim could crash while trying to process
5967 the truncated data. It now detects and logs this case.
5969 17. Further to 4.21/58, another change has been made: if (and only if) the
5970 first line of a message (the first header line) ends with CRLF, a bare LF
5971 in a subsequent header line has a space inserted after it, so as not to
5972 terminate the header.
5974 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5975 unnecessarily, used atoi() instead of strtol(), and didn't check the
5976 termination when getting file sizes from file names by regex.
5978 19. Completely re-implemented the support for maildirsize files, in the light
5979 of a number of problems with the previous contributed implementation
5980 (4.30/29). In particular:
5982 . If the quota is zero, the maildirsize file is maintained, but no quota is
5985 . If the maildir directory does not exist, it is created before any attempt
5986 to write a maildirsize file.
5988 . The quota value in the file is just a cache; if the quota is changed in
5989 the transport, the new value overrides.
5991 . A regular expression is available for excluding directories from the
5994 20. The autoreply transport checks the characters in options that define the
5995 message's headers; it allows continued headers, but it was checking with
5996 isspace() after an embedded newline instead of explicitly looking for a
5999 21. If all the "regular" hosts to which an address was routed had passed their
6000 expiry times, and had not reached their retry times, the address was
6001 bounced, even if fallback hosts were defined. Now Exim should go on to try
6004 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6005 equivalent code in the SMTP transport. Some hosts send humungous responses
6006 to HELO/EHLO, more than 1024 it seems.
6008 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6009 gives compiler warnings in some environments. I've now done it "properly",
6012 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6013 (because of problems with the built-in one) was declared to return uschar *
6014 instead of char *, causing compiler failure.
6016 25. Fixed a file descriptor leak when processing alias/forward files.
6018 26. Fixed a minor format string issue in dbfn.c.
6020 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6022 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6023 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6026 29. When a sender address is verified, it is cached, to save repeating the test
6027 when there is more than one recipient in a message. However, when the
6028 verification involves a callout, it is possible for different callout
6029 options to be set for different recipients. It is too complicated to keep
6030 track of this in the cache, so now Exim always runs a verification when a
6031 callout is required, relying on the callout cache for the optimization.
6032 The overhead is duplication of the address routing, but this should not be
6035 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6036 to be verified with callout=postmaster, and the main callout worked but the
6037 postmaster check failed, the verification correctly failed. However, if a
6038 subsequent RCPT command asked for sender verification *without* the
6039 postmaster check, incorrect caching caused this verification also to fail,
6042 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6043 it was not caching the DNS options (qualify_single, search_parents) that
6044 were used when the lookup failed. A subsequent lookup with different
6045 options therefore always gave the same answer, though there were cases
6046 where it should not have. (Example: a "domains = !$mx_any" option on a
6047 dnslookup router: the "domains" option is always processed without any
6048 widening, but the router might have qualify_single set.) Now Exim uses the
6049 cached value only when the same options are set.
6051 32. Added John Jetmore's "exipick" utility to the distribution.
6053 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6054 than a timeout (e.g. a certificate is required, and is not provided), an
6055 Exim server now closes the connection immediately. Previously it waited for
6056 the client to close - but if the client is SSL, it seems that they each
6057 wait for each other, leading to a delay before one of them times out.
6059 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6060 maintained 0.8.x compatibility because I don't think many are using it, and
6061 it is clearly obsolete.
6063 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6066 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6067 client certificate was expired. A simple patch fixes this, though I don't
6068 understand the full logic of why the verify callback is called multiple
6071 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6072 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6073 which causes problems with some clients (such as the Certicom SSL Plus
6074 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6075 disables the coutermeasure allowing Eudora to connect."
6077 38. Exim was not checking that a write() to a log file succeeded. This could
6078 lead to Bad Things if a log got too big, in particular if it hit a file
6079 size limit. Exim now panics and dies if it cannot write to a log file, just
6080 as it does if it cannot open a log file.
6082 39. Modified OS/Makefile-Linux so that it now contains
6084 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6086 The two -D definitions ensure that Exim is compiled with large file
6087 support, which makes it possible to handle log files that are bigger than
6090 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6091 instance) a domain was checked against a named list that involved a lookup,
6092 causing $domain_data to be set, then another domain was checked against the
6093 same list, then the first domain was re-checked, the value of $domain_data
6094 after the final check could be wrong. In particular, if the second check
6095 failed, it could be set empty. This bug probably also applied to
6098 41. The strip_trailing_dot option was not being applied to the address given
6099 with the -f command-line option.
6101 42. The code for reading a message's header from the spool was incrementing
6102 $received_count, but never initializing it. This meant that the value was
6103 incorrect (doubled) while delivering a message in the same process in which
6104 it was received. In the most common configuration of Exim, this never
6105 happens - a fresh exec is done - but it can happen when
6106 deliver_drop_privilege is set.
6108 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6109 it now includes up to 150 characters of the unexpected data in the log
6112 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6113 and building data strings. The size of both of these buffers was 10 000
6114 bytes - far larger than anybody would *ever* want, thought I. Needless to
6115 say, somebody hit the limit. I have increased the maximum line length to
6116 20 000 and the maximum data length of concatenated lines to 100 000. I have
6117 also fixed two bugs, because there was no checking on these buffers. Tsk,
6118 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6119 buffer is too small.
6121 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6122 lsearch lookups. Now it does.
6124 46. When parsing a route_list item in a manualroute router, a fixed-length
6125 buffer was used for the list of hosts. I made this 1024 bytes long,
6126 thinking that nobody would ever have a list of hosts that long. Wrong.
6127 Somebody had a whole pile of complicated expansion conditions, and the
6128 string was silently truncated, leading to an expansion error. It turns out
6129 that it is easier to change to an unlimited length (owing to other changes
6130 that have happened since this code was originally written) than to build
6131 structure for giving a limitation error. The length of the item that
6132 expands into the list of hosts is now unlimited.
6134 47. The lsearch lookup could not handle data where the length of text line was
6135 more than 4095 characters. Such lines were truncated, leading to shortened
6136 data being returned. It should now handle lines of any length.
6138 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6139 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6142 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6143 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6144 no longer happens in comment lines. A list of replacements is now placed
6145 at the head of all of the source files, except those whose only change is
6146 to replace PERL_COMMAND in the very first #! line.
6148 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6149 messages on the queue, with a bottom-up merge sort, using code contributed
6150 by Michael Haardt. This should make operations like -bp somewhat faster on
6151 large queues. It won't affect queue runners, except when queue_run_in_order
6154 51. Installed eximstats 1.31 in the distribution.
6156 52. Added support for SRV lookups to the dnslookup router.
6158 53. If an ACL referred to $message_body or $message_body_end, the value was not
6159 reset for any messages that followed in the same SMTP session.
6161 54. The store-handling optimization for building very long strings was not
6162 differentiating between the different store pools. I don't think this
6163 actually made any difference in practice, but I've tidied it.
6165 55. While running the routers to verify a sender address, $sender_address
6166 was still set to the sender address. This is wrong, because when routing to
6167 send a bounce to the sender, it would be empty. Therefore, I have changed
6168 it so that, while verifying a sender address, $sender_address is set to <>.
6169 (There is no change to what happens when verifying a recipient address.)
6171 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6172 target A or AAAA records (if not already returned) without resetting the
6173 qualify_single or search_parents options of the DNS resolver. These are
6174 inappropriate in this case because the targets of MX and SRV records must
6175 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6176 target that, when qualified, matched something in the local domain. These
6177 two options are now turned off when doing these lookups.
6179 57. It seems that at least some releases of Reiserfs (which does not have the
6180 concept of a fixed number of inodes) returns zero and not -1 for the
6181 number of available inodes. This interacted badly with check_spool_inodes,
6182 which assumed that -1 was the "no such thing" setting. What I have done is
6183 to check that the total number of inodes is greater than zero before doing
6184 the test of how many are available.
6186 58. When a "warn" ACL statement has a log_message modifier, the message is
6187 remembered, and not repeated. This is to avoid a lot of repetition when a
6188 message has many recipients that cause the same warning to be written.
6189 However, Exim was preserving the list of already written lines for an
6190 entire SMTP session, which doesn't seem right. The memory is now reset if a
6191 new message is started.
6193 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6194 debugging output unless log_rewrite was also set.
6196 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6197 of (int)(handle) when we know that handle contains (void *)(-1).
6199 61. The Exim daemon panic-logs an error return when it closes the incoming
6200 connection. However "connection reset by peer" seems to be common, and
6201 isn't really an error worthy of noting specially, so that particular error
6204 62. When Exim is trying to find all the local interfaces, it used to panic and
6205 die if the ioctl to get the interface flags failed. However, it seems that
6206 on at least one OS (Solaris 9) it is possible to have an interface that is
6207 included in the list of interfaces, but for which you get a failure error
6208 for this call. This happens when the interface is not "plumbed" into a
6209 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6210 failure of the "get flags" call assumes that the interface is down.
6212 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6213 makes life easier for people who are doing arithmetic on fields extracted
6214 from dates, where you often get leading zeros that should not be
6215 interpreted as octal.
6217 64. Added qualify_domain to the redirect router, to override the global
6220 65. If a pathologically long header line contained very many addresses (the
6221 report of this problem mentioned 10 000) and each of them was rewritten,
6222 Exim could use up a very large amount of memory. (It kept on making new
6223 copies of the header line as it rewrote, and never released the old ones.)
6224 At the expense of a bit more processing, the header rewriting function has
6225 been changed so that it no longer eats memory in this way.
6227 66. The generation of the Received: header has been moved from the time that a
6228 message starts to be received, to the time that it finishes. The timestamp
6229 in the Received: header should now be very close to that of the <= log
6230 line. There are two side-effects of this change:
6232 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6233 logged header lines no longer include the local Received: line, because
6234 it has not yet been created. The same applies to a copy of the message
6235 that is returned to a non-SMTP sender when a message is rejected.
6237 (b) When a filter file is tested using -bf, no additional Received: header
6238 is added to the test message. After some thought, I decided that this
6241 This change does not affect the value of $received_for. It is still set
6242 after address rewriting, but before local_scan() is called.
6244 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6246 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6247 gave an unhelpful panic error message, and a defer error. I have managed to
6248 change this behaviour so that it now rejects any supplied certificate,
6249 which seems right, as the list of acceptable certificates is empty.
6251 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6252 gave an unhelpful defer error. I have not managed to make this reject any
6253 supplied certificates, but the error message it gives is "no certificate
6254 supplied", which is not helpful.
6256 70. exigrep's output now also includes lines that are not associated with any
6257 message, but which match the given pattern. Implemented by a patch from
6258 Martin Sluka, which also tidied up the Perl a bit.
6260 71. Recipient callout verification, like sender verification, was using <> in
6261 the MAIL FROM command. This isn't really the right thing, since the actual
6262 sender may affect whether the remote host accepts the recipient or not. I
6263 have changed it to use the actual sender in the callout; this means that
6264 the cache record is now keyed on a recipient/sender pair, not just the
6265 recipient address. There doesn't seem to be a real danger of callout loops,
6266 since a callout by the remote host to check the sender would use <>.
6267 [SEE ABOVE: changed after hitting problems.]
6269 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6270 temporary errors. However, in the case of such a code being given after
6271 the end of a data transmission (i.e. after ".") Exim was failing to write
6272 a retry record for the message. (Yes, there was some broken host that was
6273 actually sending 8xx at this point.)
6275 73. An unknown lookup type in a host list could cause Exim to panic-die when
6276 the list was checked. (An example that provoked this was putting <; in the
6277 middle of a list instead of at the start.) If this happened during a DATA
6278 ACL check, a -D file could be left lying around. This kind of configuration
6279 error no longer causes Exim to die; instead it causes a defer error. The
6280 incident is still logged to the main and panic logs.
6282 74. Buglet left over from Exim 3 conversion. The message "too many messages
6283 in one connection" was written to the rejectlog but not the mainlog, except
6284 when address rewriting (yes!) was being logged.
6286 75. Added write_rejectlog option.
6288 76. When a system filter was run not as root (that is, when system_filter_user
6289 was set), the values of the $n variables were not being returned to the
6290 main process; thus, they were not subsequently available in the $sn
6293 77. Added +return_path_on_delivery log selector.
6295 78. A connection timeout was being treated differently from recipients deferred
6296 when testing hosts_max_try with a message that was older than the host's
6297 retry timeout. (The host should not be counted, thus allowing all hosts to
6298 be tried at least once before bouncing.) This may have been the cause of an
6299 occasionally reported bug whereby a message would remain on the queue
6300 longer than the retry timeout, but would be bounced if a delivery was
6301 forced. I say "may" because I never totally pinned down the problem;
6302 setting up timeout/retry tests is difficult. See also the next item.
6304 79. The ultimate address timeout was not being applied to errors that involved
6305 a combination of host plus message (for example, a timeout on a MAIL
6306 command). When an address resolved to a number of possible hosts, and they
6307 were not all tried for each delivery (e.g. because of hosts_max_try), a
6308 message could remain on the queue longer than the retry timeout.
6310 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6313 81. Fixed an obscure SMTP outgoing bug which required at least the following
6314 conditions: (a) there was another message waiting for the same server;
6315 (b) the server returned 5xx to all RCPT commands in the first message so
6316 that the message was not completed; (c) the server dropped the connection
6317 or gave a negative response to the RSET that Exim sends to abort the
6318 transaction. The observed case was a dropped connection after DATA that had
6319 been sent in pipelining mode. That is, the server had advertised PIPELINING
6320 but was not implementing it correctly. The effect of the bug was incorrect
6321 behaviour, such as trying another host, and this could lead to a crash.
6327 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6328 and daemon.c were passed as pointers to ints; they should have been
6329 pointers to socklen_t variables (which are typically unsigned ints).
6331 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6334 3. Fixed a really odd bug that affected only the testing scheme; patching a
6335 certain fixed string in the binary changed the value of another string that
6336 happened to be identical to the end of the original first string.
6338 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6339 name", it returns that address as the IP address. On some operating
6340 systems (e.g. Solaris), it also passes back the IP address string as the
6341 "host name". However, on others (e.g. Linux), it passes back an empty
6342 string. Exim wasn't checking for this, and was changing the host name to an
6343 empty string, assuming it had been canonicalized.
6345 5. Although rare, it is permitted to have more than one PTR record for a given
6346 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6347 all the names associated with an address, because they do in Solaris.
6348 However, it seems that they do not in Linux for data that comes from the
6349 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6350 I found this out when I moved to a new Linux workstation and tried to run
6351 the Exim test suite.
6353 To get round this problem I have changed the code so that it now does its
6354 own call to the DNS to look up PTR records when searching for a host name.
6355 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6356 addresses that are only in /etc/hosts are still found.
6358 This behaviour is, however, controlled by an option called host_lookup_
6359 order, which defaults to "bydns:byaddr". If people want to use the other
6360 order, or indeed, just use one or the other means of lookup, they can
6361 specify it in this variable.
6363 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6364 some operating systems, this comes back from gethostbyaddr() as an empty
6365 string, and this is what Exim used to test for. However, it seems that in
6366 other systems, "." is yielded. Exim now tests for this case too.
6368 7. The values of check_spool_space and check_log_space are now held internally
6369 as a number of kilobytes instead of an absolute number of bytes. If a
6370 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6371 kilobyte. This means that much larger values can be stored.
6373 8. Exim monitor: an attempt to get the action menu when not actually pointing
6374 at a message produces an empty menu entitled "No message selected". This
6375 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6376 no entries in it ("Shell widget menu has zero width and/or height"). So I
6377 have added a single, blank menu entry in this case.
6379 9. Added ${quote_local_part.
6381 10. MIME decoding is now applied to the contents of Subject: header lines when
6384 11. Now that a reference to $sender_host_address automatically causes a reverse
6385 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6386 host lookup before query-style lookups in lists that might use this
6387 variable. This has therefore been abolished, and the "net-" prefix is no
6388 longer necessary for query-style lookups.
6390 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6391 have been a typo for LFLAGS, so it has been changed.
6393 13. The install script calls Exim with "-C /dev/null" in order to find the
6394 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6395 to be output. However, since Exim outputs its version number before the
6396 error, it didn't break the script. It just looked ugly. I fixed this by
6397 always allowing "-C /dev/null" if the caller is root.
6399 14. Ignore overlarge ACL variable number when reading spool file - insurance
6400 against a later release with more variables having written the file.
6402 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6403 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6404 this, as well as the form without the "IPv6" on the front.
6406 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6407 OS/Makefile-Darwin file.
6409 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6410 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6412 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6413 variable. However, some OS use ints for inodes. I've added cast to long int
6414 to get rid of the compiler warning.
6416 19. I had forgotten to lock out "/../" in configuration file names when
6417 ALT_CONFIG_PREFIX was set.
6419 20. Routers used for verification do not need to specify transports. However,
6420 if such a router generated a host list, and callout was configured, Exim
6421 crashed, because it could not find a port number from the (non-existent)
6422 transport. It now assumes port 25 in this circumstance.
6424 21. Added the -t option to exigrep.
6426 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6427 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6428 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6429 them is set, LOOKUP_LSEARCH is forced.
6431 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6433 24. Added sender and host information to the "rejected by local_scan()" log
6434 line; previously there was no indication of these.
6436 25. Added .include_if_exists.
6438 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6439 Linux. It turns out that not all file systems support this. Apparently some
6440 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6441 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6442 not supported on the file descriptor, is now ignored when Exim is trying to
6443 sync a directory. This applies only to Linux.
6445 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6447 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6448 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6451 29. Added support for maildirsize files from supplied patch (modified a bit).
6453 30. The use of :fail: followed by an empty string could lead Exim to respond to
6454 sender verification failures with (e.g.):
6456 550 Verification failed for <xxx>
6457 550 Sender verify failed
6459 where the first response line was missing the '-' that indicates it is not
6460 the final line of the response.
6462 31. The loop for finding the name of the user that called Exim had a hardwired
6463 limit of 10; it now uses the value of finduser_retries, which is used for
6464 all other user lookups.
6466 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6469 33. Exim was neglecting to zero errno before one call of strtol() when
6470 expanding a string and expecting an integer value. On some systems this
6471 resulted in spurious "integer overflow" errors. Also, it was casting the
6472 result into an int without checking.
6474 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6475 did not work. The code looks as if it has *never* worked, though it appears
6476 to have been documented since at least release 1.62. I have made it work.
6478 35. The "timeout_DNS" error in retry rules, also documented since at least
6479 1.62, also never worked. As it isn't clear exactly what this means, and
6480 clearly it isn't a major issue, I have abolished the feature by treating it
6481 as "timeout", and writing a warning to the main and panic logs.
6483 36. The display of retry rules for -brt wasn't always showing the error code
6486 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6487 timeout_connect_A, timeout_connect_MX.
6489 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6490 to the empty sender.
6492 39. The daemon was not analysing the content of -oX till after it had closed
6493 stderr and disconnected from the controlling terminal. This meant that any
6494 syntax errors were only noted on the panic log, and the return code from
6495 the command was 0. By re-arranging the code a little, I've made the
6496 decoding happen first, so such errors now appear on stderr, and the return
6497 code is 1. However, the actual setting up of the sockets still happens in
6498 the disconnected process, so errors there are still only recorded on the
6501 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6502 connections (as happens on some IP stacks) was logged at start up time as
6503 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6504 it from "IPv6 and IPv4", which means that two separate sockets are being
6507 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6508 says whether AF_INET or AF_INET6 was passed as an argument.
6510 42. Exiwhat output was messed up when time zones were included in log
6513 43. Exiwhat now gives more information about the daemon's listening ports,
6514 and whether -tls-on-connect was used.
6516 44. The "port" option of the smtp transport is now expanded.
6518 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6519 silently ignored. Now an error message is written to the main and panic
6522 46. There's a new ACL modifier called "logwrite" which writes to a log file
6523 as soon as it is encountered.
6525 47. Added $local_user_uid and $local_user_gid at routing time.
6527 48. Exim crashed when trying to verify a sender address that was being
6530 49. Exim was recognizing only a space character after ".include". It now also
6531 recognizes a tab character.
6533 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6534 extracting the relevant information from the specification. The man page no
6535 longer contains scrambled data for the -d option, and I've added a section
6536 at the front about calling Exim under different names.
6538 51. Added "extra_headers" argument to the "mail" command in filter files.
6540 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6543 53. Installed eximstats 1.29.
6545 54. Added transport_filter_timeout as a generic transport option.
6547 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6548 Cc: header lines. This was required by RFC 822, but it not required by RFC
6551 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6552 incoming messages that did not have them. Now it does so only if the
6553 message originates locally, that is, if there is no associated remote host
6554 address. When Resent- header lines are present, this applies to the Resent-
6555 lines rather than the non-Resent- lines.
6557 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6558 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6560 58. Messages for configuration errors now include the name of the main
6561 configuration file - useful now that there may be more than one file in a
6562 list (.included file names were always shown).
6564 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6565 for those rare installations that do not start the daemon as root or run it
6566 setuid root. I've cut out the call to initgroups() if the daemon is not
6569 60. The Exim user and group can now be bound into the binary as text strings
6570 that are looked up at the start of Exim's processing.
6572 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6574 62. Added $mailstore_basename variable.
6576 63. Installed patch to sieve.c from Michael Haardt.
6578 64. When Exim failed to open the panic log after failing to open the main log,
6579 the original message it was trying to log was written to stderr and debug
6580 output, but if they were not available (the usual case in production), it
6581 was lost. Now it is written to syslog before the two lines that record the
6582 failures to open the logs.
6584 65. Users' Exim filters run in subprocesses under the user's uid. It is
6585 possible for a "deliver" command or an alias in a "personal" command to
6586 provoke an address rewrite. If logging of address rewriting is configured,
6587 this fails because the process is not running as root or exim. There may be
6588 a better way of dealing with this, but for the moment (because 4.30 needs
6589 to be released), I have disabled address rewrite logging when running a
6590 filter in a non-root, non-exim process.
6596 1. The buildconfig auxiliary program wasn't quoting the value set for
6597 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6598 not defined. This bug was masked in 4.22 by the effect that was fixed in
6601 2. Some messages that were rejected after a message id was allocated were
6602 shown as "incomplete" by exigrep. It no longer does this for messages that
6603 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6605 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6606 have allow_domain_literals set, the ID did not get logged in the <= line.
6607 Domain literals are now always recognized in Message-ID: header lines.
6609 4. The first argument for a ${extract expansion item is the key name or field
6610 number. Leading and trailing spaces in this item were not being ignored,
6611 causing some misleading effects.
6613 5. When deliver_drop_privilege was set, single queue runner processes started
6614 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6615 same command in the process it spins off) were not dropping privilege.
6617 6. When the daemon running as "exim" started a queue runner, it always
6618 re-executed Exim in the spun-off process. This is a waste of effort when
6619 deliver_drop_privilege is set. The new process now just calls the
6620 queue-runner function directly.
6626 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6629 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6630 ignored. Though the use of -oP was forcing the writing of a pid file, it
6631 was always written to the default place.
6633 3. If the message "no IP address found for host xxxx" is generated during
6634 incoming verification, it is now followed by identification of the incoming
6635 connection (so you can more easily find what provoked it).
6637 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6639 5. Added some features to "harden" Exim a bit more against certain attacks:
6641 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6642 be put in Local/Makefile. This is like the never_users runtime option,
6643 but it cannot be overridden. The default setting is "root".
6645 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6646 prefix string with which any file named in a -C command line option
6649 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6650 is retained for -C and -D only if the caller of Exim is root. Without
6651 it, the exim user may also use -C and -D and retain privilege.
6653 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6654 command line option is disabled.
6656 6. Macro names set by the -D option must start with an upper case letter, just
6657 like macro names defined in the configuration file.
6659 7. Added "dereference=" facility to LDAP.
6661 8. Two instances of the typo "uknown" in the source files are fixed.
6663 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6664 the Configure-Makefile script screwed up while processing it.
6666 10. Incorporated PCRE 4.4.
6668 11. The SMTP synchronization check was not operating right at the start of an
6669 SMTP session. For example, it could not catch a HELO sent before the client
6670 waited for the greeting. There is now a check for outstanding input at the
6671 point when the greeting is written. Because of the duplex, asynchronous
6672 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6673 way, but not yet received, when the check is performed.
6675 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6676 on TCP/IP sockets, because this apparently causes some broken clients to
6679 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6680 unchanged) from the Cygwin maintainer.
6682 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6683 was supported instead of testing for mbx. Effectively a typo.
6685 15. The spa authenticator server code was not checking that the input it
6686 received was valid base64.
6688 16. The debug output line for the "set" modifier in ACLs was not showing the
6689 name of the variable that was being set.
6691 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6693 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6694 lookup on its own. Something else was needed to trigger the lookup. For
6695 example, a match in host_lookup or the need for a host name in a host list.
6696 Now, if $sender_host_name is referenced and the host name has not yet been
6697 looked up, a lookup is performed. If the lookup fails, the variable remains
6698 empty, and $host_lookup_failed is set to "1".
6700 19. Added "eqi" as a case-independent comparison operator.
6702 20. The saslauthd authentication condition could segfault if neither service
6703 nor realm was specified.
6705 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6706 error message that was logged was misleading, and incoming SMTP
6707 connections were dropped. The message is now more accurate, and temporary
6708 errors are given to SMTP connections.
6710 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6711 (except RSET) with a 421 error, until QUIT is received. However, it was
6712 failing to send a response to QUIT.
6714 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6715 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6716 ACL is now run after the helo_try_verify_hosts code.
6718 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6719 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6720 case-independent, so other case variants are also recognized). Apparently
6721 some systems use these upper case variants.
6723 25. If more than two messages were waiting for the same host, and a transport
6724 filter was specified for the transport, Exim sent two messages over the
6725 same TCP/IP connection, and then failed with "socket operation on non-
6726 socket" when it tried to send the third.
6728 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6730 27. The extern definition of crypt16() in expand.c was not being excluded when
6731 the OS had its own crypt16() function.
6733 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6734 as a preferred synonym for return_size_limit, both as an option and as an
6737 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6739 30. Changed the default configuration ACL to relax the local part checking rule
6740 for addresses that are not in any local domains. For these addresses,
6741 slashes and pipe symbols are allowed within local parts, but the sequence
6742 /../ is explicitly forbidden.
6744 31. SPA server authentication was not clearing the challenge buffer before
6747 32. log_message in a "warn" ACL statement was writing to the reject log as
6748 well as to the main log, which contradicts the documentation and doesn't
6749 seem right (because no rejection is happening). So I have stopped it.
6751 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6752 However, I am unable to do any testing of this.
6754 34. Fixed an infelicity in the appendfile transport. When checking directories
6755 for a mailbox, to see if any needed to be created, it was accidentally
6756 using path names with one or more superfluous leading slashes; tracing
6757 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6759 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6760 contents are added to the log line that is written for every discarded
6761 recipient. (Previously a log_message setting was ignored.)
6763 36. The ${quote: operator now quotes the string if it is empty.
6765 37. The install script runs exim in order to find its version number. If for
6766 some reason other than non-existence or emptiness, which it checks, it
6767 could not run './exim', it was installing it with an empty version number,
6768 i.e. as "exim-". This error state is now caught, and the installation is
6771 38. An argument was missing from the function that creates an error message
6772 when Exim fails to connect to the socket for saslauthd authentication.
6773 This could cause Exim to crash, or give a corrupted message.
6775 39. Added isip, isip4, and isip6 to ${if conditions.
6777 40. The ACL variables $acl_xx are now saved with the message, and can be
6778 accessed later in routers, transports, and filters.
6780 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6781 strings in the file are not string-expanded.
6783 42. If a MAIL command specified a SIZE value that was too large to fit into an
6784 int variable, the check against message_size_limit failed. Such values are
6785 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6786 day this will have to be increased, but I don't think I want to be around
6787 when emails are that large.
6794 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6795 iconv() is not standard in FreeBSD.
6797 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6798 IPv6 enabled. The observed symptom was a segmentation fault on return from
6799 the function os_common_find_running_interfaces() in src/os.c.
6801 3. In the check_special_case() function in daemon.c I had used "errno" as an
6802 argument name, which causes warnings on some systems. This was basically a
6803 typo, since it was named "eno" in the comments!
6805 4. The code that waits for the clock to tick (at a resolution of some fraction
6806 of a second) so as to ensure message-id uniqueness was always waiting for
6807 at least one whole tick, when it could have waited for less. [This is
6808 almost certainly not relevant at current processor speeds, where it is
6809 unlikely to ever wait at all. But we try to future-proof.]
6811 5. The function that sleeps for a time interval that includes fractions of a
6812 second contained a race. It did not block SIGALRM between setting the
6813 timer, and suspending (a couple of lines later). If the interval was short
6814 and the sigsuspend() was delayed until after it had expired, the suspension
6815 never ended. On busy systems this could lead to processes getting stuck for
6818 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6819 process, before it forks any delivery processes. The open lookup caching
6820 mechanism meant that the open file or database connection was passed into
6821 the delivery process. The problem was that delivery processes always tidy
6822 up cached lookup data. This could cause a problem for the next delivery
6823 process started by the queue runner, because the external queue runner
6824 process does not know about the closure. So the next delivery process
6825 still has data in the lookup cache. In the case of a file lookup, there was
6826 no problem because closing a file descriptor in a subprocess doesn't affect
6827 the parent. However, if the lookup was caching a connection to a database,
6828 the connection was closed, and the second delivery process was likely to
6829 see errors such as "PGSQL: query failed: server closed the connection
6830 unexpectedly". The problem has been fixed by closing all cached lookups
6831 in a queue runner before running a delivery process.
6833 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6834 seem to have the "const" qualifier which it has on other OS. I've
6837 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6838 *of the same type* (client or server) with the same public name that an
6839 error should be diagnosed.
6841 9. When Exim looked up a host name for an IP address, but failed to find the
6842 original IP address when looking up the host name (a safety check), it
6843 output the message "<ip address> does not match any IP for NULL", which was
6844 confusing, to say the least. The bug was that the host name should have
6845 appeared instead of "NULL".
6847 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6848 user that is built into the binary, and the -C or -D options is used, root
6849 privilege is dropped before the configuration file is read. In addition,
6850 logging is switched to stderr instead of the normal log files. If the
6851 configuration then re-defines the Exim user, the unprivileged environment
6852 is probably not what is expected, so Exim logs a panic warning message (but
6855 However, if deliver_drop_privilege is set, the unprivileged state may well
6856 be exactly what is intended, so the warning has been cut out in that case,
6857 and Exim is allowed to try to write to its normal log files.
6863 1. smtp_return_error_details was not giving details for temporary sender
6864 or receiver verification errors.
6866 2. Diagnose a configuration error if two authenticators have the same public
6869 3. Exim used not to create the message log file for a message until the first
6870 delivery attempt. This could be confusing when incoming messages were held
6871 for policy or load reasons. The message log file is now created at the time
6872 the message is received, and an initial "Received" line is written to it.
6874 4. The automatically generated man page for command line options had a minor
6875 bug that caused no ill effects; however, a more serious problem was that
6876 the procedure for building the man page automatically didn't always
6877 operate. Consequently, release 4.20 contains an out-of-date version. This
6878 shouldn't happen again.
6880 5. When building Exim with embedded Perl support, the script that builds the
6881 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6882 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6884 6. The freeze_tell option was not being used for messages that were frozen on
6885 arrival, either by an ACL or by local_scan().
6887 7. Added the smtp_incomplete_transaction log selector.
6889 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6890 was accepting AUTH without a new EHLO.
6892 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6893 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6896 10. Logging of TCP/IP connections (when configured) now happens in the main
6897 daemon process instead of the child process, so that the TCP/IP connection
6898 count is more accurate (but it can never be perfect).
6900 11. The use of "drop" in a nested ACL was not being handled correctly in the
6901 outer ACL. Now, if condition failure induced by the nested "drop" causes
6902 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6903 or "require"), the connection is dropped.
6905 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6906 that yield "discard" can now be used with an "accept" or a "discard" verb,
6907 but an error is generated for any others (because I can't see a useful way
6908 to define what should happen).
6910 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6911 are now processed in the same way as lines in the Exim configuration file.
6912 In particular, continuation lines are supported.
6914 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6916 15. Added -ti meaning -t -i.
6918 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6919 domains, and warn by logging if others are found.
6921 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6922 returned from GIFCONF when Exim is trying to find the list of interfaces on
6923 a host. The code in os.c has been modified to copy each ifreq to an aligned
6924 structure in all cases.
6926 Also, in some cases, the returned ifreq's were being copied to a 'struct
6927 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6928 means the last couple of bytes of an IPv6 address could be chopped if the
6929 ifreq contained only a normal sockaddr (14 bytes storage).
6931 18. Named domain lists were not supported in the hosts_treat_as_local option.
6932 An entry such as +xxxx was not recognized, and was treated as a literal
6935 19. Ensure that header lines added by a DATA ACL are included in the reject log
6936 if the ACL subsequently rejects the message.
6938 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6939 MD5 (which is deprecated).
6941 21. When testing a filter file using -bf, Exim was writing a message when it
6942 took the sender from a "From " line in the message, but it was not doing so
6943 when it took $return_path from a Return-Path: header line. It now does.
6945 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6946 with a valid header line field name (a series of printing characters
6947 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6949 23. Changed "disc" in the source to "disk" to conform to the documentation and
6950 the book and for uniformity.
6952 24. Ignore Sendmail's -Ooption=value command line item.
6954 25. When execve() failed while trying to run a command in a pipe transport,
6955 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6956 could be confused with a return value of 69 from the command itself. This
6957 has been changed to 127, the value the shell returns if it is asked to run
6958 a non-existent command. The wording for the related log line suggests a
6959 non-existent command as the problem.
6961 26. If received_header_text expands to an empty string, do not add a Received:
6962 header line to the message. (Well, it adds a token one on the spool, but
6963 marks it "old" so that it doesn't get used or transmitted.)
6965 27. Installed eximstats 1.28 (addition of -nt option).
6967 28. There was no check for failure on the call to getsockname() in the daemon
6968 code. This can fail if there is a shortage of resources on the system, with
6969 ENOMEM, for example. A temporary error is now given on failure.
6971 29. Contrary to the C standard, it seems that in some environments, the
6972 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6973 program. Exim now does this explicitly; it affects the formatting of
6974 timestamps using strftime().
6976 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6977 complaints. I've now initialized all the variables, to avoid this.
6979 32. Header lines added by a system filter were not being "seen" during
6980 transport-time rewrites.
6982 33. The info_callback() function passed to OpenSSL is set up with type void
6983 (*)(SSL *, int, int), as described somewhere. However, when calling the
6984 function (actually a macro) that sets it up, the type void(*)() is
6985 expected. I've put in a cast to prevent warnings from picky compilers.
6987 34. If a DNS black list lookup found a CNAME record, but there were no A
6988 records associated with the domain it pointed at, Exim crashed.
6990 35. If a DNS black list lookup returned more than one A record, Exim ignored
6991 all but the first. It now scans all returned addresses if a particular IP
6992 value is being sought. In this situation, the contents of the
6993 $dnslist_value variable are a list of all the addresses, separated by a
6996 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6997 to accept a host name and all its aliases if the forward lookup for any of
6998 them yielded the IP address of the incoming connection. Now it accepts only
6999 those names whose forward lookup yields the correct IP address. Any other
7000 names are discarded. This closes a loophole whereby a rogue DNS
7001 administrator could create reverse DNS records to break through a
7002 wildcarded host restriction in an ACL.
7004 37. If a user filter or a system filter that ran in a subprocess used any of
7005 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7006 the wrong values were passed to the pipe command ($thisaddress had the
7007 value of $0, $0 had the value of $1, etc). This bug was introduced by
7008 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7010 38. Improved the line breaking for long SMTP error messages from ACLs.
7011 Previously, if there was no break point between 40 and 75 characters, Exim
7012 left the rest of the message alone. Two changes have been made: (a) I've
7013 reduced the minimum length to 35 characters; (b) if it can't find a break
7014 point between 35 and 75 characters, it looks ahead and uses the first one
7015 that it finds. This may give the occasional overlong line, but at least the
7016 remaining text gets split now.
7018 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7019 file descriptors might be low, and that setting 1000 would always raise it.
7020 It turns out that in some environments, the limit is already over 1000 and
7021 that lowering it causes trouble. So now Exim takes care not to decrease it.
7023 40. When delivering a message, the value of $return_path is set to $sender_
7024 address at the start of routing (routers may change the value). By an
7025 oversight, this default was not being set up when an address was tested by
7026 -bt or -bv, which affected the outcome if any router or filter referred to
7029 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7030 log only when "message" or "log_message" are set. However, if one of the
7031 conditions was an address verification, or a call to a nested ACL, the
7032 messages generated by the underlying test were being passed through. This
7033 no longer happens. The underlying message is available in $acl_verify_
7034 message for both "message" and "log_message" expansions, so it can be
7035 passed through if needed.
7037 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7038 new expansion $bh_ to give the encoded byte string without charset
7039 translation. Translation happens only if iconv() is available; HAVE_ICONV
7040 indicates this at build time. HEADERS_CHARSET gives the charset to
7041 translate to; headers_charset can change it in the configuration, and
7042 "headers charset" can change it in an individual filter file.
7044 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7045 that creates RFC 2047 encoded "words" labels them as that charset instead
7046 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7047 expansion operator; (ii) when Exim creates a From: line for a local
7048 message; (iii) when a header line is rewritten to include a "phrase" part.
7050 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7051 buggy, causing it to skip the first lines of messages whose message ID
7052 ended in 'D'. This would not have bitten before Exim release 4.14, because
7053 message IDs were unlikely to end in 'D' before then. The effect was to have
7054 incorrect size information for certain domains.
7056 45. #include "config.h" was missing at the start of the crypt16.c module. This
7057 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7060 46. If there was a timeout during a "random" callout check, Exim treated it as
7061 a failure of the random address, and carried on sending RSET and the real
7062 address. If the delay was just some slowness somewhere, the response to the
7063 original RCPT would be taken as a response to RSET and so on, causing
7064 mayhem of various kinds.
7066 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7067 when I implemented it. It didn't allow for the fact that some option values
7068 may legitimately be negative (e.g. size_addition), and it didn't even do
7069 the right test for positive values.
7071 48. Domain names in DNS records are case-independent. Exim always looks them up
7072 in lower case. Some resolvers return domain names in exactly the case they
7073 appear in the zone file, that is, they may contain uppercase letters. Not
7074 all resolvers do this - some return always lower case. Exim was treating a
7075 change of case by a resolver as a change of domain, similar to a widening
7076 of a domain abbreviation. This triggered its re-routing code and so it was
7077 trying to route what was effectively the same domain again. This normally
7078 caused routing to fail (because the router wouldn't handle the domain
7079 twice). Now Exim checks for this case specially, and just changes the
7080 casing of the domain that it ultimately uses when it transmits the message
7083 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7086 50. If a filter generated a file delivery with a non-absolute name (possible if
7087 no home directory exists for the router), the forbid_file option was not
7090 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7091 the existing equality matching.
7093 52. Exim was using ints instead of ino_t variables in some places where it was
7094 dealing with inode numbers.
7096 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7097 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7098 called TMPDIR, and if it finds it is different, it changes its value.
7100 54. The smtp_printf() function is now made available to local_scan() so
7101 additional output lines can be written before returning. There is also an
7102 smtp_fflush() function to enable the detection of a dropped connection.
7103 The variables smtp_input and smtp_batched_input are exported to
7106 55. Changed the default runtime configuration: the message "Unknown user"
7107 has been removed from the ACL, and instead placed on the localuser router,
7108 using the cannot_route_message feature. This means that any verification
7109 failures that generate their own messages won't get overridden. Similarly,
7110 the "Unrouteable address" message that was in the ACL for unverifiable
7111 relay addresses has also been removed.
7113 56. Added hosts_avoid_esmtp to the smtp transport.
7115 57. The exicyclog script was not checking for the esoteric option
7116 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7117 will work only if exicyclog is run under the appropriate euid.
7119 58. Following a discussion on the list, the rules by which Exim recognises line
7120 endings on incoming messages have been changed. The -dropcr and drop_cr
7121 options are now no-ops, retained only for backwards compatibility. The
7122 following line terminators are recognized: LF CRLF CR. However, special
7123 processing applies to CR:
7125 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7126 nor a local message in the state where . is a terminator.
7128 (ii) If a bare CR is encountered in a header line, an extra space is added
7129 after the line terminator so as not to end the header. The reasoning
7130 behind this is that bare CRs in header lines are most likely either
7131 to be mistakes, or people trying to play silly games.
7133 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7134 was being incorrectly given as 18 bytes larger than it should have been.
7135 This is a VOB (very old bug).
7137 60. This may never have affected anything current, but just in case it has:
7138 When the local host is found other than at the start of a list of hosts,
7139 the local host, those with the same MX, and any that follow, are discarded.
7140 When the list in question was part of a longer list of hosts, the following
7141 hosts (not currently being processed) were also being discarded. This no
7142 longer happens. I'm not sure if this situation could ever has previously
7145 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7148 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7149 header. This is something that is recommended in a new Internet Draft, and
7150 is something that is documented as being done by Sendmail. There are two
7151 possible values. For messages generated by the autoreply transport, Exim
7154 Auto-Submitted: auto-replied
7156 whereas for all other generated messages (e.g. bounces) it adds
7158 Auto-Submitted: auto-generated
7160 63. The "personal" condition in filters now includes a test for the
7161 Auto-Submitted: header. If it contains the string "auto-" the message it
7162 not considered personal.
7164 64. Added rcpt_include_affixes as a generic transport option.
7166 65. Added queue_only_override (default true).
7168 66. Added the syslog_duplication option.
7170 67. If what should have been the first header line of a message consisted of
7171 a space followed by a colon, Exim was mis-interpreting it as a header line.
7172 It isn't of course - it is syntactically invalid and should therefore be
7173 treated as the start of the message body. The misbehaviour could have
7174 caused a number of strange effects, including loss of data in subsequent
7175 header lines, and spool format errors.
7177 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7178 client host had authenticated. This control can now be exercised by an ACL
7179 for more flexibility.
7181 69. By default, callouts do not happen when testing with -bh. There is now a
7182 variant, -bhc, which does actually run the callout code, including
7183 consulting and updating the callout cache.
7185 70. Added support for saslauthd authentication, courtesy of Alexander
7188 71. If statvfs() failed on the spool or log directories while checking their
7189 size for availability, Exim confusingly gave the error "space shortage".
7190 Furthermore, in debugging mode it crashed with a floating point exception.
7191 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7192 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7193 serious problem, Exim now writes to the main and panic logs when this
7194 happens, with details of the failure. It then refuses to accept the
7195 incoming message, giving the message "spool directory problem" or "log
7196 directory problem" with a 421 code for SMTP messages.
7198 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7199 0, 1, and 2 exist, because some OS complain for execs without them (see
7200 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7201 descriptors. However, the code omitted to check that the open succeeded,
7202 causing mysterious errors if for some reason the permissions on /dev/null
7203 got screwed. Now Exim writes a message to the main and panic logs, and
7204 bombs out if it can't open /dev/null.
7206 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7207 interact so that it is all more flexible. It is supposed to remain
7208 backwards compatible. Also added extra_local_interfaces.
7210 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7211 to bomb out with an assertion failure - to the client this appears as a
7212 connection drop. This problem occurs in the part of the code that was taken
7213 from the Samba project. Fortunately, the assertion is in a very simple
7214 function, so I have fixed this by reproducing the function inline in the
7215 one place where it is called, and arranging for authentication to fail
7216 instead of killing the process with assert().
7218 75. The SPA client code was not working when the server requested OEM rather
7219 than Unicode encoding.
7221 76. Added code to make require_files with a specific uid setting more usable in
7222 the case where statting the file as root fails - usually a non-root-mounted
7223 NFS file system. When this happens and the failure is EACCES, Exim now
7224 forks a subprocess and does the per-uid checking as the relevant uid.
7226 77. Added process_log_path.
7228 78. If log_file_path was not explicitly set, a setting of check_log_space or
7229 check_log_inodes was ignored.
7231 79. If a space check for the spool or log partitions fails, the incident is now
7232 logged. Of course, in the latter case the data may get lost...
7234 80. Added the %p formatting code to string_format() so that it can be used to
7235 print addresses in debug_print(). Adjusted all the address printing in the
7236 debugging in store.c to use %p rather than %d.
7238 81. There was a concern that a line of code in smtp_in.c could overflow a
7239 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7240 initially expressed, the concern was not well-founded, because trailing
7241 spaces are removed early. However, if the trailing spaces were followed by
7242 a NULL, they did not get removed, so the overflow was possible. Two fixes
7245 (a) I re-wrote the offending code in a cleaner fashion.
7246 (b) If an incoming SMTP command contains a NULL character, it is rejected
7249 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7250 runs initgroups(), so that if the Exim user is in any additional groups,
7251 they will be used during message reception.
7257 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob